WO2007036586A2 - Apparatus and method for obtaining information relating to cerebral haemodynamics - Google Patents

Abstract

The invention relates to an apparatus and method for obtaining information relating to cerebral haemodynamics. The inventive method comprises the following steps consisting in applying an excitation signal (Se) to the head and capturing two output signals (S1) and a second output signal (S2) with different cerebral blood flow or volume dependencies and different scalp blood flow and volume dependencies. The aforementioned signals are processed in order to obtain a result which reflects cerebral blood flow with minimum corruption by the blood flow in the scalp.

Description

APPARATUS AND METHOD OF OBTAINING INFORMATION RELATING TO THE

BRAIN HEMODINAMICS

TECHNICAL SECTOR OF THE INVENTION

The present invention is attached to the biomedical instrumentation sector and, more specifically, to the sector of measurement of features related to cerebral hemodynamics.

STATE OF THE TECHNIQUE The brain tissue is especially susceptible to lack of oxygen and nutrients: the absence of cerebral blood supply for a few minutes can lead to irreversible damage to the brain and, if blood supply is not restored immediately, it can cause subject's death

This shows the importance of the measurement of cerebral blood flow (FSC) in clinical practice to determine if its magnitude is within the normal limits, so that if a decrease in the patient is detected in the patient

FSC below these limits, appropriate therapies can be applied immediately before their consequences are manifested.

Among others, the most frequent causes by which the FSC can decrease below its normal limits, in addition to cardiac arrest, are:

1) In the first place, due to a sharp decrease in blood pressure, which may occur, among other causes, due to alterations in nerve regulation of the latter, - as a result of anesthesia in operations surgical; or for loss of large amounts of blood.

2) Secondly, due to an increase in the value of the intracranial pressure (ICP) above its limit value, which hinders or can even prevent the entry of blood into the interior space of the cranial structure.

Although alterations in blood pressure can be easily identified through its measurement, the elevation of the ICP above its normal values is usually manifested directly through the physiological alterations and neurological damage produced as a result of the decrease in FSC.

The most common causes of ICP elevation are those associated with head injuries, which in fact constitute the first cause of death in the West in subjects under 45 years of age (see Fearnside MR, Simpson DA. Epidemiology. In: Reilly P, Bullock R., editors, Head Injury, Honduras: Chapman & Hall Medical, 1997: 3-23). Other common causes are hydrocephalus, thrombosis, hemorrhages, tumors, cerebral ischemia following cardiac arrest, meningitis, etc.

Specifically, the cause-effect relationship between head trauma and ICP elevation can be understood taking into account that the inside of the skull, considered as a rigid and inextensible structure in the adult, is occupied by three elements; the brain, blood and cerebrospinal fluid. According to the accepted Monroe-Kelly doctrine, the sum of the volumes occupied by each of these three elements in the same individual must stay constant at all times of time, so an increase in volume of any of them has to be compensated by a decrease in the volume of at least one of the remaining. As a consequence of the trauma, and depending on its severity, an edema that increases the volume occupied by the brain usually occurs. This increase in volume, in a first stage, is compensated by the expulsion of cerebrospinal fluid into the spinal subarachnoid space and, simultaneously, by the emptying of the blood contained in the venous bed. If the edema becomes more acute, this process continues until, at a certain point, these compensation resources are exhausted. Thereafter, any small increase in the volume occupied by any of the three elements causes significant increases in the ICP as a result of the loss of elastance of the set. Occasionally, due to the difficulty of identifying the lesions caused by head trauma, even seemingly mild cases can evolve into serious complicated situations within a few hours.

The continuous evaluation of the FSC is of special relevance for the correct treatment of these patients. The optimal FSC evaluation device should meet the following requirements: quantitative assessment; high spatial resolution; continuous measurement; minimal or no influence on the normal functioning of the brain; non-invasive; reduced cost; laptop; and of application in the clinical scope. In addition, it would be convenient that it could be carried out by non-highly specialized personnel and be applicable to type of patient Currently, there is a wide range of methods to evaluate FSC directly, although none of them satisfies all the above characteristics. These techniques include: 1) Tomographic techniques, such as computerized axial tomography (CT) with contrast agent, positron emission tomography (PET), certain nuclear magnetic resonance imaging (NMR) techniques, and photon emission tomography single (SPECT), which, although excellent, have a high cost and require the transfer of the patient to a specific place where the measure is carried out, so, in addition to the damage that can be caused in critical states, it prevents that the monitoring of said state is constant.

2) The Transcranial Doppler, which allows the blood velocity to be recorded in the main arteries but, in addition to requiring highly specialized personnel, the relationship of the measurement with the FSC is dependent on the thickness of the vessel measured and is not applicable in all the subjects

3) The technique of dilution or washing of radioactive isotopes (eg ¹³³ Xe), whereby the patient is injected into the carotid or a radioactive solution is inhaled and the transit time of the isotopes through the cerebrovascular system is recorded.

In addition to the above, there are two methods to assess FSC indirectly, which, in reference to desirable requirements, complement existing techniques, both being non-invasive, relatively inexpensive, and easily applied in clinical settings: reoencephalography (REG) and near infrared spectroscopy (NIRS)

Although there are notable differences between the two methods, the REG and the NIRS are based on the same general physical principle. In them, a certain amount of energy (or excitation signal) is applied to the subject's head from the outer surface of the scalp (with scalp - in English: "scalp") means not only the part covered by hair but also the other parts of the fabric that covers the upper part of the head, including the forehead). This energy or signal is partially dispersed and / or absorbed differently by each of the biological substances and tissues that make up the head. Due to this, the value of the physical quantities associated with said energy or signal that can be collected at any point of a subject's head depends on the concentration and distribution of said biological substances and tissues. In both methods, the process consists in measuring on the surface of the scalp and during the application of said energy or signal, the value of a physical quantity associated with said energy and linked to a physiological quantity of interest, thus obtaining an indirect estimate of said physiological magnitude. However, both in the REG and in the NIRS, especially because the energy is applied from the outer surface of the scalp, the value of the measured physical magnitude not only depends on the value and distribution of said physiological variable in the brain, but also one's own in the scalp. Therefore, the measure taken in this way It is inherently contaminated by information from the scalp. Due to this limitation, the use of the REG and the NIRS in the assessment of the FSC is not extended in the practical clinic. More specifically, the type of energy used in the REG is an electric energy, whose distribution in the subject's head depends especially on the electrical conductivity and spatial arrangement of the tissues that compose it and particularly on the blood volume contained in each of them, while in the NIRS, the applied energy is an electromagnetic energy in the form of infrared light of certain wavelengths, the absorption of which is extremely dependent on the concentration of certain chromophore molecules and, particularly, on the concentration of hemoglobin, which is also related to the blood volume contained in the biological tissues that make up the head. Reoencephalography. Reoencephalography, also called transcranial plethysmography, is a non-invasive technique that emerged in 1950, through which FSC is indirectly evaluated through an electrical measurement. The pulsatile nature of the FSC is responsible for the cerebral blood volume (VSC), understood as the amount of blood contained in the endocranial space, varying pulsatically synchronously with the heartbeat. In mathematical terms, and considering the FSC as the incoming flow to the endocranial space, the evolution in time of the VSC is directly related to the integral of the FSC extended over time. The measure of the variations Synchronous with heart beat of the VSC can be performed indirectly and not quantitatively, and therefore also those of the FSC, taking advantage of the fact that the electrical characteristics of the blood are different from those of the rest of the tissues and substances that make up the head. In particular, the electrical conductivity of the blood is markedly higher than that of the brain tissue, so that the electrical conductivity of the assembly changes with varying the proportion with which both tissue and blood are mixed. In this way, the electrical conductivity of the head varies analogously to how the VSC does as a result of the pulsatile character of the FSC. The measurement of this global change in the electrical characteristics of the conductive medium that is the head of the subject is carried out non-invasively by measuring the electrical impedance of said head by means of electrodes attached to the surface of the scalp. Research on the ability of this measurement technique to evaluate the FSC began in 1950 and, sequentially, two main variants of the technique appeared depending on the electrode configuration used to perform the measurement: the so-called REG I or bipolar REG, in the that the impedance was measured by injecting an electric current between two electrodes located on the surface of the scalp and measuring the potential difference between them; and the so-called REG II or tetrapolar REG, in which the impedance measurement was performed by injecting a certain current between two electrodes located on the surface of the scalp and measuring the voltage difference between two additional electrodes located on the same surface. However, the detractors of the technique claimed that, due to the low electrical conductivity of the skull, the injected electrical current circulated mainly through the scalp tissue being minimal that crossed the skull, so the impedance variations that were they captured by reoencephalography were not really a consequence of the VSC variations, but rather those of the scalp blood volume (VSCC) associated with the pulsatile variations of the scalp blood flow (FSCC). From 1975, a consensus seems to be reached on the origin of the reoencephalographic signal, admitting that REG I reflects almost entirely the variations of the VSCC, while the REG II contains information on both the variations of the VSCC and those of the VSC , however, no consensus was reached regarding the proportion with which both are mixed (see Pérez JJ, Pebble E, Barcia ^' JA., "Quantification of intracranial contribution to rheoencephalography by a numerical model of the head"; Clinical Neurophysiology, 111 ( 2000); 1306-1314). In fact, recent studies confirm that the proportion with which the information derived from the VSC and VSCC variations is mixed strongly depends on the physical constitution of the subject, so there is no universal electrode position in which it can a REG II be performed without being contaminated by VSCC variations (see Pérez JJ, Pebble E, Barcia JA; "Influence of the scalp thickness on the intracranial contribution to rheoencephalography"; Physics in Medicine and Biology, 49 (2004); 4383 -4394). Currently, as a result of the doubts raised in the scientific community about the ability of the REG to reflect the pulsatility of the FSC, the technique is abandoned in the West in the clinical setting. WO-A-03/059164 describes a device and method for estimating FSC from impedance variations measured with electrodes located on the surface of the scalp. To decrease the influence of the skull bones in the FSC measurements, the electrodes are located in areas devoid of bone (auditory or nasal channels) or in areas where said skull bones are less thick.

WO-A-95/35060 describes a method and apparatus for estimating the prognosis of damages affecting pathological processes in animals, including humans, by detecting changes in the electrical impedance of the affected organ.

Near infrared spectroscopy

Near infrared spectroscopy (NIRS) is an optical technique, whose application in monitoring cerebral hemodynamics is of relatively recent appearance, whereby changes in the blood of the amounts of oxyhemoglobin, deoxyhemoglobin and other proteins associated with the oxygen transport The relationship between these amounts is related, among others, to the blood oxidative state, the cellular metabolism and the FSC. Specifically, the physical principle of operation of the NIRS is based on two optical characteristics of biological tissues: on the one hand, said tissues are relatively translucent for radiation of lengths near the infrared wavelength (650 nm at 1100 nm) and, on the other, the spectral absorbance of oxyhemoglobin, deoxyhemoglobin and other chromophobic proteins of the blood differs substantially between them (see Owen-Reece H, Smith M, Elwell CE, Goldstone JC; ⁿ Near infrared spectroscopy "; British Journal of Anaesthesia, 82 (1999); 418-426). The NIRS is a non-invasive technique by which a beam of infrared light of certain wavelengths and usually generated by laser diodes , the subject's scalp surface is affected. Part of the emitted photons pass through the scalp, the skull and reach the brain tissue, being dispersed and absorbed in its path by the molecules that make up each and every one of these tissues Simultaneously, a high sensitivity photodetector (photomultiplier or high sensitivity silicon photodiode) located in a position close to the emitter and on the surface of l Scalp, captures the radiation that emerges through the surface of the scalp in that position. The intensity of the infrared radiation collected in the photodetector depends, among other factors, on the amount of hemoglobin that the light beam finds in its mean path and on its oxidative state, which provides information to evaluate the cerebral hemodynamic state of the subject being able to In addition, relate the measure to the regional FSC (see Obrig H, Villringer A: "Beyond the visible — Imaging the human brain with light"; Journal of Cerebral Blood Flow & Metabolism, 23 (2003); 1-18). At present, the NIRS has a greater field of application in neonates than in adults, since in First the cartilaginous character of the skull allows a greater penetration of light. In adults, however, due to the opacity of the skull, the photodetector not only receives information on the blood perfusion of the brain tissue, but also from the scalp itself, mixing both in unknown proportions, depending on this proportion of the subject and the location of the light emitter and receiver. This replicates the same problem of the REG described above. To avoid the influence on the NIRS of extracranial tissue, some manufacturers instead of using a single photodetector, have two photodetectors on the surface of the scalp to subsequently process the signal collected in both and cancel, even if only partially obtained, the information from the scalp tissue.

Lastly, and in relation to the NIRS, it should be noted that as the measure provided by the photodetector is especially sensitive to the amount of hemoglobin and its oxidative state contained in the tissues, and since this amount is proportional, among others, to the volume of blood contained in these tissues, the signal supplied by the photodetector varies in time in a pulsatile manner synchronously with the heartbeat, similar to how the VSC and the VSCC do.

WO-A-94/27493, US-A-4223680, US-A-5057695, US-A-5139025 and US-A-5482034 describe different systems based on NIRS technology or the like.

US-A-5490505 describes an apparatus for eliminating biological signal artifacts through processing. In summary, the serious consequences derived from an erroneous measurement of the FSC value due to the contamination of the measure by the FSCC, justify the need for a method and apparatus that allow to evaluate, directly or indirectly, FSC parameters with an influence FSCC reduced. Although there is a wide range of commercial devices of excellent performance to assess the patient's FSC, there is still a need for a device that provides information about the FSC in a non-invasive way, preferably at a low cost, which preferably allows continuous monitoring, and that preferably it is portable and easy to use. The two techniques, REG and NIRS, which already exist that allow such a measure to be carried out with these characteristics, present the insurmountable inconvenience of providing inherently mixed information derived from brain tissue with that coming from the scalp. Each of these registers can therefore be considered constituted by a component of intracranial origin plus another component of extracranial origin.

The aim of the present invention is, from the information provided by the REG, by the NIRS or any other technique that is based on the same general physical principle described above or similar principles, to devise a method and apparatus by which, capturing two or more regs of REG, NIRS or similar, the information corresponding to brain tissue is extracted from any of them, by substantially separating its intra and extracranial components.

DESCRIPTION OF THE INVENTION It is known to use statistical tools to separate components from different sources, present, in different proportions, in a plurality of technically captured signals, the problem is called "blind source separation" (BSS: "Blind Source Separation"). The Pérez JJ document: "Plethysmographic Components in Evoked Potentials Synchronized with the Heartbeat. Identification by Enhanced Rheoencephalography Techniques"; Doctoral thesis. Polytechnic university of Valencia; Valencia 2003 (Spain); Chapter 3.3, pages 114-117 proposes three methods, namely: a method known as "principal component analysis"; - a method known as "independent component analysis"; and a method of separation that would be based on differences between the process of blood perfusion of the scalp (with an abrupt increase in the volume of arterial blood upon arrival of the blood pulse, which is possible due to the distensibility of the arteries) and the process of cerebral blood perfusion (conditioned by the fact that the cerebral arteries cannot expand substantially, because the brain tissue is surrounded by the skull). However, the document does not explain how this theoretical possibility of taking advantage of morphological differences between the intra and extracranial components of the reencephalographic records could be implemented. The present invention is based on a practical use of said morphological differences, which will be discussed in more detail. in connection with the description of a preferred embodiment of the invention and with reference to the drawings.

Basically, the inventive concept lies in taking advantage of the fact that, as has been proven, the scalp blood flow (FSCC) shows a greater variability in time between beats than cerebral blood flow (FSC).

A first aspect of the invention relates to an apparatus for obtaining information related to cerebral hemodynamics of a human or animal subject. The apparatus comprises: means of application to the head (preferably, on the scalp - in English, "scalp" -, including the forehead) of the subject, of at least one excitation signal (Se) (which can be any signal, for example, a constant, sinusoidal signal, etc .; the means of application may consist of a signal generator and a terminal, electrode type, infrared light emitting element, etc., which can be used to apply the signal to head) ; pick-up means configured for simultaneous capture, at least two different positions of the subject's head, of a first output signal (Sl) and a second output signal (S2). When the apparatus is used, the pick-up means are selected and / or positioned so that both the first output signal (Sl) and the second output signal (S2) are a function of, that is, dependent on the signal of excitation (Se) and physicochemical characteristics of the scalp and / or the subject's brain, so that the relationship between said signals of output (Sl, S2) and said excitation signal (Se) depends on the FSCC and / or FSC of the subject, or on other physiological variables related to them, such as the scalp blood volume (VSCC) and / or the cerebral blood volume (VSC) of the subject. The selection of the pickup positions will be described later, in relation to a description of a method according to another aspect of the invention, and which can be performed with the apparatus of the invention. The apparatus of the invention further comprises calculation means configured to calculate the value of a function.

(F) of a difference (S2-KS1) between said second output signal (S2) and a fraction (K) of said first output signal (Sl), said function (F) being selected so as to represent an indication of blood flow and the calculation means being configured to calculate said fraction (K) so that the variability of the value of said function (F) is substantially minimal over a preselected time interval, according to a selected variability criterion .

Any conventional criteria can be used to determine (or, rather, "estimate") the variability of the value of the function (F) in the preselected time interval, for example: - Subtracting the value of the function (F) from its value average within said time interval and calculating the effective value of the result.

- Calculating the variance of the value of the function (F) within said time interval or calculating any of its higher order statistical moments. - Calculating the harmonic distortion of the value of the function (F) within said time interval.

- Subtracting from the value of the function (F) its average value within said time interval and calculating, in general, any dispersion or energy estimator of the result.

The value of the function (F) over the interval can be provided as an "estimate" or "reflection" of the FSC; the difference (S2-KS1) between said second output signal can also be provided as "output signal", and once the fraction (K) has been calculated

(S2) and said fraction (K) of said first output signal

(Sl); this output signal could reflect, for example, the VSC, or the variation in the blood content of certain chromophores proteins.

The function (F) may be the temporary derivative of said difference between said second output signal (S2) and said fraction (K) of said first output signal

(Sl). This option may be appropriate when the application and feedback means are such that the output signals have values that at each moment depend substantially on the VSCC and / or the VSC.

Alternatively, the function (F) may be a function whose value is (directly) proportional to said difference between said second output signal (S2) and said fraction (K) of said first output signal (Sl). This option may be appropriate when the output signals, at all times, have a value that depends substantially on the FSCC and / or the FSC. The apparatus may comprise means for selecting said time interval so that it corresponds to (is that is, to include) a significant part of a subject's cardiac cycle, for example, so as to include the time of the arrival of a blood pulse wave at the subject's head. The application means and the capture means may be configured to apply the excitation signal (Se) in at least one excitation position in the subject's head, to capture the first output signal (Sl) in at least a first position closer to said excitation position, and to capture said second output signal (S2) in at least a second position further away from said excitation position. In this way, it can be achieved that a first output signal depends very much on the FSCC (or VSCC) but almost nothing on the FSC (VSC) (for example, in line with what happens with a REG I signal) and that the second Output signal depends on both FSC (or VSC) and FSCC (or VSCC) (in line with what happens with a REG II signal).

The calculation means may be configured to calculate the value of the function (F) from an average of the first output signal (Sl) and an average of the second output signal (S2) over a plurality of cardiac cycles .

The means for applying the excitation signal may comprise means for generating an electrical signal and at least two excitation electrodes applicable on both excitation positions.

The pick-up means may comprise at least two first pick-up electrodes applicable on first output positions, to capture said first output signal (Sl). Excitation electrodes and the first pick-up electrodes can be part of an electrode structure in which the first pick-up electrodes are located close to the excitation electrodes. This structure may comprise a patch, tape, elastic helmet or the like that incorporates the electrodes, whereby the electrodes are kept at a fixed distance between them.

The first pick-up electrodes may be located together with the excitation electrodes or be constituted by the excitation electrodes.

The first pick-up electrodes may be located at a distance less than 15 mm from the respective excitation electrodes.

On the other hand, the pick-up means can comprise at least two second pick-up electrodes, to capture said second output signal. These second pick-up electrodes can be part of an electrode structure (for example, a patch, a tape, an elastic helmet or the like in which the electrodes are located, with predetermined distances between them) in which each of the seconds pickup electrodes are further away from the nearest excitation electrode than the corresponding first pickup electrode. The second pick-up electrodes may, for example, be at a distance greater than 15 mm from the respective excitation electrodes (ie, each second pick-up electrode can be at a distance greater than 15 mm from the nearest excitation electrode). The pick-up means may be configured so that the first output signal is a signal of bipolar reoencephalography output (REG I) and so that the second output signal is a tetrapolar reoencephalography output signal (REG II).

Alternatively, the excitation means may comprise means for generating another type of signal, for example, an electromagnetic (Se) excitation signal, and the pick-up means will then comprise the corresponding sensors, for example, electromagnetic energy sensors configured to capture the first output signal (Sl) and the second output signal (S2). For example, the excitation signal may be an infrared radiation signal, in which case the apparatus could be based on NIRS technology.

The pick-up means may be configured such that said first output signal (Sl) depends substantially on the FSCC and does not depend substantially> _{χ on} the FSC, and said second output signal (S2) depends (for example, substantially) on the FSC and (for example, substantially) the FSCC. From these two signals in which the proportions between the influences of the FSC and the FSCC are different, a fairly reliable indication of the FSC can be obtained.

The apparatus can be an apparatus comprising several physically independent elements and even located in different places, in which case the apparatus could be considered as a distributed system. For example, the part that performs the calculations may be located in a physically remote place from the place of the pick-up of the output signals. Another aspect of the invention relates to a method of obtaining information related to hemodynamics. brain of a human or animal subject, comprising the steps of: applying to the head (preferably, on the scalp, including the forehead) of the subject, at least one excitation signal (Se) (for example, constant, sinusoidal, etc. ) ; to capture, during a capture period and in at least two different positions of the subject's head, a first output signal (Sl) and a second output signal (S2), being both the first output signal (Sl) and the second output signal (S2) a function, that is, dependent, on the excitation signal (Se) and physicochemical characteristics of the scalp and / or the subject's brain, so that the relationship between said output signals (Sl , S2) and said excitation signal

(It) depends on the FSCC and / or the subject's FSC, said first output signal (Sl) and said second output signal (S2) being captured at selected positions such that - said first output signal (Sl) depends on substantial form of the FSCC (or, at any time, the VSCC) and does not depend substantially on the FSC (or, at any time, the VSC) (that is, the desired thing is that the influence of the FSC or VSC on the first output signal be minimal), and - said second output signal (S2) depends on the FSC (or, at any time, the VSC) and the FSCC (or, at any time, the VSCC); calculating the value of a function (F) of a difference (S2-KS1) between said second output signal (S2) and a fraction (K) of said first output signal

(Sl), said function (F) being selected so that its value represents an indication of blood flow (that is, if the Sl and S2 signals reflect a blood volume at all times, the function F will be a temporary derivative; if Sl and S2 reflect a flow at any time, the function F will be a directly proportional function, etc.) and said fraction (K) being selected so that the variability of the value of said function (F) is substantially minimal over a preselected time interval, according to a selected variability criterion. As indicated above, any suitable criteria can be used to determine variability. The important thing is that the variability can be reduced so that the influence of the FSCC or VSCC is substantially eliminated, as can be seen from what has been explained above. Thus, the variation in the value of the function (F) over time reflects the variations in the FSC, with minimal (or at least substantially reduced) contamination by FSCC. As the output signal, the value of the function (F) or the value of the difference (S2-KS1) between said second output signal (S2) and said fraction (K) of said first output signal (Sl) can be taken , depending on what you want to see (for example, if you want to detect the variation of the VSC, the variation of the FSC or the variation of the blood content of certain chromophore proteins, etc.).

The first output signal (Sl) may at each moment have a value that depends substantially on the VSCC and that does not depend substantially on the VSC, and the second output signal (S2) may have a value that depends at each moment (for example, so substantially) of the VSC and (for example, substantially) of the VSCC.

In this case, the function (F) may be the temporary derivative of the difference between said second output signal (S2) and said fraction (K) of said first output signal (Sl).

In the event that the first output signal (Sl) has at each moment a value that depends substantially on the FSCC and that does not depend substantially on the FSC, and the second output signal (S2) has at each moment a value that depends ( for example, substantially) of the FSC and (for example, substantially) of the FSCC; The function (F) can be a function whose value is directly proportional to the difference between the second output signal (S2) and the fraction (K) of the first output signal (Sl).

The time interval may correspond to (ie, include) a significant part of a subject's cardiac cycle, for example, it may include the time of the arrival of a blood pulse wave to the subject's head.

The excitation signal (Se) can be applied in at least one excitation position on the subject's head, obtaining the first output signal (Sl) in at least a first position closer to said excitation position, and obtaining said second exit sign

(S2) in at least a second position further away from said excitation position.

The value of the function (F) can be calculated from an average of the first output signal (Sl) and an average of the second output signal (S2) over a plurality of cardiac cycles, for example, over a few tens or several hundred cardiac cycles.

The excitation signal can be an electrical signal, in which case the excitation signal can be applied with at least two excitation electrodes, applied on respective excitation positions. The first output signal (Sl) can be captured with at least two first pick-up electrodes applied on respective first output positions, which can be close to the excitation positions (for example, at a distance less than 15 mm from the respective excitation positions, that is, each first pick-up electrode may be located at a distance less than 15 mm from the nearest excitation electrode, or even identical to the excitation positions (in which case the first pick-up electrodes may comprise the electrodes of excitation, that is, the same electrodes can be used to apply the excitation signal to capture the first output signal.) The second output signal (S2) can be picked up with at least two second pick-up electrodes located in respective second output positions further away from the excitation positions than said first output positions (i.e., each second The pickup electrode may be located at a distance from the nearest excitation electrode, greater than the distance between the first corresponding pickup electrode and said closest excitation electrode), for example, at a distance greater than 15 mm from the respective excitation positions (i.e. each second pickup electrode may be at a distance greater than 15 from the nearest excitation electrode).

The first output signal may be a bipolar reoencephalography output signal (REG I) and the second output signal may be a tetrapolar reoencephalography output signal (REG II). This allows to use conventional equipment to capture the signals, and the only thing that should be added to these conventional equipment would be its configuration to handle the detected signals, something that can be done, for example, by installing appropriate software.

Alternatively, the excitation signal (Se) can be an electromagnetic signal, in which case the first output signal (Sl) and the second output signal (S2) can be captured with respective electromagnetic energy sensors. For example, the excitation signal may be an infrared radiation signal, in which case NIRS technology could be used.

The method of the invention can also be carried out separately from the capture of the output signals. That is, the method can be carried out remotely (in space and / or time) of the signal pick-up. In such a case, an operator would receive signals that can supposedly correspond to the first and second output signals discussed above. The method could be limited to calculating the value of a function (F) of a difference (S2-KS1) between said second output signal (S2) and a fraction (K) of said first output signal (Sl), said function (F) selected to represent an indication of blood flow and selecting said fraction (K) of so that the variability of the value of said function (F) is substantially minimal over a preselected time interval, in accordance with a selected variability criterion; and providing FSC-related data based on said difference (S2-KSl) between said second output signal (S2) and said fraction (K) of said first output signal (Sl). The function (F) may be the temporary derivative of said difference between said second output signal (S2) and said fraction (K) of said first output signal (Sl), or a function whose value is proportional to said difference between said second output signal (S2) and said fraction (K) of said first output signal (Sl); It all depends on what is considered as reflected in the output signals to which the method is applied (that is, it depends on whether the output signals are directly related to volume or flow).

The apparatus of the invention may be an apparatus configured to carry out the method according to the invention.

Another aspect of the invention relates to a computer program, which comprises a program code (or program code means) configured to carry out the method of the invention, when executed in a programmable electronic device (for example, on a PC or other computer equipment associated with means of obtaining the relevant output signals).

Although the inventive concept is applicable both in the scope of the REG and in the NIRS, for simplicity, this text explains the invention especially in its version applied to the REG; however, the same Explanation is equally applicable to NIRS and other similar or similar technologies, mutatis mutandis.

BRIEF DESCRIPTION OF THE FIGURES To complement the description and in order to help a better understanding of the features of the invention, in accordance with some preferred examples of practical implementation thereof, a set of such is attached as an integral part of said description. Drawings in which for illustrative purposes and not limitation, the following has been represented:

Figure 1 shows, schematically, an apparatus according to a preferred embodiment of the invention, applied to a subject. Figure 2. - Schematically shows a typical electrocardiogram of a subject, as well as the cerebral blood flow (FSC) and the corresponding scalp blood flow (FSCC).

Figure 3.- Shows, schematically, an apparatus according to an embodiment of the invention, applied to a subject.

Figure 4. - Shows possible positions of the excitation and pick-up electrodes, according to an exemplary embodiment of the invention. Figures 5-8 reflect results of an experiment in which the method according to a preferred embodiment of the invention was applied to a human subject.

Figure 9 schematically reflects some of the functional elements of the apparatus according to a preferred embodiment of the invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT

Fig. 1 represents a general view of a preferred embodiment of the invention, which is composed of an electronic apparatus or system 100, of four electrodes 106, 107, 108 and 109, and of the cables necessary to interconnect the system and the electrodes The figure also represents the use of the invention in a subject 110 to monitor and / or store, by means of a computer system 111, a plethysmographic signal 112 dependent on cerebral blood flow (FSC) and substantially independent of blood flow from the scalp (FSCC) .

The electronic system includes an electrical energy source 101 by means of which an electric current (which constitutes an excitation signal) is injected into the subject's head through the excitation electrodes 106 and 109, located on the surface of the scalp 104 of the subject 110. Likewise, the electronic system has a first electronic subsystem 102 which is responsible for obtaining a first plethysmographic signal from the measurement of the electrical potential between a first pair of electrodes created by the injected current. In order that the plethysmographic signal obtained by the electronic subsystem 102 is substantially or exclusively due to the FSCC, said first pair of pick-up electrodes are placed two by two in the immediate vicinity of the excitation electrodes 106 and 109 or coincide with them such and as shown in Fig. 1, wherein said first pair of pickup electrodes are formed by the electrodes 106 and 109 themselves. In the same way, a subsystem Electronic 103, similar to 102, is responsible for obtaining a second plethysmographic signal from the measurement of the electrical potential between a second pair of pick-up electrodes 107 and 108, also located on the surface of the scalp 104. In order that this Second plethysmographic signal obtained by the electronic subsystem 103 is specifically dependent on the FSC even if it is contaminated by the FSCC, the second pick-up electrodes 107 and 108 are relatively far from the excitation electrodes 106 and 109. On the other hand, the electronic system comprises a processor 105 configured to jointly analyze the plethysmographic signals obtained by the electronic subsystems 102 and 103 to, from them, substantially eliminate the FSCC dependent information from the signal obtained by the electronic subsystem 103, leaving mainly the FSC dependent information . The resulting plethysmographic signal is available as an output of the invention for its acquisition, treatment, registration and / or analysis in a computer system 111, which can be part of the apparatus of the invention or consist of a computer system (for example, a conventional PC ) independent of the apparatus of the invention.

Multiple variations of the invention described above can be conjugated to obtain similar output plethysmographic signals. To facilitate the understanding of the inventive concept on which these variations are based, the physical principle of operation of the invention will be described below. As a result of the excitation produced by the application of the electric power source 101 through of the excitation electrodes 106 and 109, a distribution of current lines is created in the subject's head that gives rise to an electric field whose potential is capable of being measured by the appropriate electronics. This distribution of current lines, and therefore the electrical potential associated with them, will vary over time as the electrical properties of the conductive medium that the subject's head changes, which happens periodically each time a blood pulse reaches the tissues of the head: the blood has an electrical conductivity greater than that of the tissues that make up the head, so an increase in the volume of blood contained in the head decreases its electrical resistance. In each cardiac pulse, therefore, the volume of blood that reaches the head decreases the electrical impedance of the head, reducing the measurable electrical potential from its surface. These variations in electrical impedance, quantifiable by the relationship between the measured electrical potential and the injected current, constitute a plethysmographic signal of the head specifically called reoencephalogram that has been previously described.

The state of the technique detailed in the previous section reflects that when the capture of the plethysmographic signal is carried out in positions close to or identical to the current injection positions, the impedance variations detected are due in their entirety to the variations of blood content of the scalp tissue, the contribution of cerebral blood pulse being minimal or nil. Thus, and using the terminology of the prior art, the plethysmographic signal obtained by subsystem 102 is a REG I _{7, provided} that the same pair of electrodes is used for current injection as for the measurement of electrical potential. This signal, therefore, contains information on the variations, beat to beat, of the scalp blood volume (VSCC), without the variations of cerebral blood volume

(VSC) have an effect on it.

On the other hand, and simultaneously with the capture of the plethysmographic signal corresponding to REG I, the subsystem 103 makes another additional measure of the electric potential created by the energy source 101, but using this time the second pick-up electrodes 107 and 108, both located in locations necessarily different from the excitation electrodes 106 and 109 (which, as indicated, also constitute in this case the first collection electrodes). The electrical potential thus measured, and expressed in values per unit of injected current, constitutes a second plethysmographic signal whose variations, beat to beat, are due, according to the state of the art, to the variations of the VSC and those of the VSCC mixed in unknown proportions because it is a REG II.

The embodiment of the invention according to Figure 1 presents to the processing element 105 two plethysmographic signals: one of which comes from subsystem 102 and provides information exclusively on the VSCC, while the other is supplied by subsystem 103 and contains information both VSC and VSCC, both mixed in unknown proportions. The processor 105 is therefore responsible for eliminating (at less, substantially) of the plethysmographic signal provided by the subsystem 103, the information about the VSCC, which is known from the output of the subsystem 102, leaving the processor 105 at its output a plethysmographic signal dependent exclusively (or, at least, mainly ) of the VSC (although in practice there may be a certain dependence on the VSCC; however, if said dependence is sufficiently reduced, this does not prevent the use of the signal obtained to reach conclusions on the development of the VSC and, therefore, the FSC).

The problem described is what is classically known as "blind separation of sources." This name includes those problems of signal processing in which from the observation of the arbitrary mixtures of a certain number of signal sources it is intended to know the information provided by each of said sources. The resolution of these problems is classically addressed from the perspective of statistics and it is necessary to take advantage of some statistical particularity that relates the original sources to proceed with the separation of these. Thus, for example, the so-called "independent component analysis" assumes that the signals provided by the original sources are statistically independent (see Hyvárinen A, Karhunen J ₁ Eye E. Independent component analysis. New York: Wiley, 2001). There is, however, or at least not known, any statistical property of the VSC and the VSCC that serves as a foothold to achieve separation of their mixtures. .

- 32 -

The invention described herein takes advantage of physiological differences instead of statistics between blood filling of the brain tissue and scalp tissue, as shown in Fig. 2. In this figure, the shape of Periodic wave of FSCC, FSC and electrocardiogram (ECG) for three cardiac cycles obtained experimentally. Specifically, curve 200 represents the ECG of a healthy subject that shows the electrical activity of the heart. In said curve, a relatively narrow and positive peak can be observed in the instant of time 203, which in electrocardiography is specifically called "R wave". During the time window occupied by said peak, ventricular contraction occurs, which corresponds to the impulse of the blood contained in the ventricles into the vascular system. In a resting subject, the heart functions as a pulsatile pump that injects in the arterial systemic circulatory system an average of 58 ml of blood in each heartbeat that, in permanent regime, crosses the capillaries in the form of continuous flow to, subsequently, return to the heart through the venous system. The elasticity of the arteries allows the storage of blood between beat and beat, draining continuously through the capillaries. In an electrical analogy, the process is similar to the charging of a capacitor by repeated deltas of current Dirac (blood stroke associated with each beat), then discharged through a constant current source (venous return). Thus, at each instant of time such as 203 of Fig. 2, a certain amount of blood is propelled to the arterial circulatory system, which advances along it in the form of a blood pulse wave. At that time, the FSCC represented in curve 202 is practically null and, approximately 100 milliseconds later, at time 204, the blood pulse reaches the arteries of the scalp, which is reflected in the FSCC in the form of an abrupt peak. During the time window that covers this peak, the arteries of the scalp suddenly fill with blood. Immediately afterwards, as a consequence of the mechanical inertia of the blood and the elasticity of the arteries, the direction of blood flow is slightly reversed, the aortic valve closing and producing an oscillation around the null value that can be observed in curve 202 in the moments after 204, similar to the phenomenon of water hammer in dynamic fluid.

However, the described arterial filling process is not extrapolated to cerebral blood circulation. Of the 58 ml of blood injected in a pulsatile manner in each heartbeat in the arterial systemic circulatory system, around 12 ml enter the cranial cavity, which represents about 20% of the systemic cardiac output for an organ that barely weighs 1300 gr. Although the pulse wave travels through the internal carotids and vertebral arteries penetrating the cranial cavity, the increase in the volume of the arterial bed associated with the increase in pressure cannot occur in the endocranial space in the same way that it occurs in the scalp, due to the rigidity of the enveloping structure that involves the skull. Thus, according to the Monroe-Kelly doctrine, the sum of the volumes of blood, tissue Cerebral and cerebrospinal fluid must be kept constant at all times of time. Therefore, the cerebral arteries cannot expand as do those of the scalp, so it is the arteries outside the skull, both the carotid and the vertebral arteries, that store the blood that, during the rest of the cardiac cycle, will enter the cranial cavity As a consequence, under normal conditions the blood flow entering the cranial cavity must be, and is, noticeably less pulsatile or, in other words, more constant between beats than that of similar peripheral arteries as shown in the FSC curve 201. It can be seen that, after the ventricular contraction of the instant 203 and around the instant 204 of the arrival of the pulse wave to the scalp, said pulse wave reaches the carotid and vertebral arteries, the curve 201 showing a step in the FSC due to the increase in blood pressure in these vessels, which contrasts with the abrupt peak of the FSCC 202 at such times. The profile that follows the FSC curve

201 after said step, which corresponds to the filling of the cerebral arteries, ^• is a practically linear decrease caused by the practically constant emptying of the carotid and vertebral arteries, which sustain the flow by means of the elastic energy stored in their walls. Therefore, the profile of the FSC curve, such as curve 201 shown in Fig. 2, is relatively constant, presenting in practice a relationship between the peak-peak amplitude and its average value throughout the cardiac cycle between 0.55 and 0.75, while blood flow in the Peripheral arteries such as those of the scalp have a profile similar to that of an under-absorbed sinus such as that of the FSCC 202 curve. This is precisely the physiological difference mentioned above that is exploited in this invention to eliminate information related to the scalp. : During the time interval indicated in window 205, which corresponds to the time of the range of a blood pulse wave at the subject's head, the instantaneous value of FSCC 202 varies intensely compared to that of FSC 201.

Returning again to Fig. 1, the signals provided by subsystems 102 and 103 are, however, directly proportional to the blood volume contained in the brain tissue and / or that of the scalp. For this reason, and since the physiological difference stated, by which the plethysmographic components are separated according to their origin, has been formulated in terms of blood flow and not blood volume, it must be taken into account the relationship between both functions: in a first approach, and assuming that the return to the venous system is carried out at a constant flow rate, it can be assumed that the arterial blood flow is the time derived from the blood volume contained in the tissue. Therefore, the signals provided by subsystems 102 and 103 are indirectly dependent on FSC and FSCC.

Fig. 9 schematically reflects how, by means of the electric power source 101 and the excitation electrodes (not illustrated), the excitation signal Se (t) is applied (which can be any type of suitable excitation signal, such and how which are usually used in reoencephalography) on the scalp 104 of the subject and how the two output signals provided by subsystems 102 and 103 are obtained, as described above. In accordance with this embodiment of the invention, the processor 105 performs the following operations, as illustrated in Fig. 9: subtract (105A) from the second plethysmographic signal provided by the second subsystem 103 (hereinafter referred to as: " the second output signal S2 (t) ") a fraction (K) of the plethysmographic signal provided by the first subsystem 102 (hereinafter referred to as:" the first output signal Sl (t) "), both simultaneously captured during a time window 205 containing the arrival of the blood pulse wave to the tissues of the head;

- obtains (105B) the derivative of the signal resulting from said subtraction;

- calculates (105C) an indicator of the variability (J) of the result of said derivative;

- and adjust (105D) the value of said fraction (K) until the value of said variability indicator (J) is minimal.

Once the fraction (K) of the signal 102 to be subtracted from the signal 103 is obtained so that the variability indicator (J) of the subtraction derivative is the minimum possible (J = J _rain ), calculate (105E ) and delivers as output the difference signal (S2 (t) -KSl (t)), for the value of K corresponding to J _min ) (with which the output reflects the VSC and its temporal variation) (or, alternatively, it gives as a result the value of function F, in which case the output reflects the FSC and its temporal variation).

As for the variability indicator referred to above, since it is applied to a derivative function, the result of said derivative has a null average value, so any variable related to the energy value of the derivative can be used as an indicator of variability . For example, the processor 105 can use the mean quadratic value, proportional to the statistical variance, as the variability indicator, although alternatively any other indicator directly or indirectly related to the energy obtained in both the temporal and spectral domain can be used Separation processing can be applied to all those physiological signals not directly related to the FSC but which, by their nature, maintain some indirect relationship with it. Thus, for example, the NIRS makes a measure related to the concentrations of molecules contained in the blood such as oxyhemoglobin, deoxyhemoglobin and other chromophores proteins. Therefore, if such concentration in blood is assumed constantly throughout a cardiac cycle, the concentration in the tissue varies beat to beat with the volume of blood contained in said tissue. Thus, the processing described above can be applied in the same manner as in plethysmographic signals to determine the fraction of the optical signal that is originated in the scalp, in order to cancel it and obtain a signal originated exclusively by the brain tissue. Virtually all of the preferred invention described herein admits variations in various aspects.

The application of the invention described on the subject 110 can be applied equally to an animal. The energy source 101 can be a source of alternating current, of constant amplitude and low enough that it does not cause physiological damage to the human or animal being. Depending on the object of study, its frequency can be chosen so that the current does not pass through the cell membrane or, alternatively, high enough that it crosses all or part of the cell membrane to obtain information about the intracellular fluid. Alternatively, the energy source 101 may be constituted by a constant voltage alternating voltage source that simultaneously performs the measurement of the circulating current through its terminals and supplies said information to the subsystems 102 and 103 for the computation of the plethysmographic signal. Optionally, the energy source 101 may be constituted by a current source, or alternatively by a source of arbitrary voltage, value and waveform that simultaneously measures the circulating current and supplies said information to subsystems 102 and 103 for the computation of the plethysmographic signal.

The excitation electrodes 106 and 109, as well as the pick-up electrodes 107 and 108 can be electroencephalographic electrodes of any type adhered or held to the surface of the scalp with products or accessory elements, respectively, suitable for this. Alternatively, any of said electrodes are conductive elements that maintain electrical contact with the surface of the scalp or, optionally, some of them or all of them are inserted into the scalp tissue. The positions occupied by the excitation electrodes 106 and 109 can be any pair of positions of the scalp, near or far between them, as long as said electrodes are not in physical contact with each other. The positions occupied by the second pick-up electrodes 107 and 108 can be chosen from among the points that make up the shortest imaginary arc that joins the excitation electrodes 106 and 109 and exceeds them both, without being necessary therefore that the two or either of the two pick-up electrodes 107 and 108 are between those of excitation 106 and 109. Optionally, the pick-up electrodes used by subsystem 102 may be different from those of excitation 106 and 109, using a third pair of electrodes, not included in Fig. 1, in which case the greatest distance between any of the excitation electrodes 106 and 109, and the closest of the pickup electrodes of said third pair is preferably less than 15 millimeters and, simultaneously, any of the second pick-up electrodes 107 and 108 is further away from any of the excitation electrodes 106 and 109 than any of the elements may be. ctrodos of capture of said third pair.

Optionally, the application of the electric power source can be performed using two or more electric power sources or, additionally or alternatively, more than two electrodes can be used for its application. The electronic subsystems 102 and 103 may be non-linear circuits that deliver to the processor 105 electrical, digitized or analog signals, directly proportional to the amplitude variations of the potential difference measured between their respective inputs per unit of circulating current.

The processor 105 may be a digital electronic system based on an element that executes a program, such as a computer, a microprocessor or a digital signal processor, which applies the separation procedure to the signals from subsystems 102 and 103 and which , in case these signals were analogical, prior to their processing it digitizes them. Alternatively, the processor 105 may be a digital electronic system implemented over programmable logic or in a specific application integrated circuit. The processor 105 may also be an analog or mixed electronic system that applies said method of separating the input signals. Said method of separating the input signals from the subsystems 102 and 103, and carried out by the processor 105, can be performed directly from the signals provided by said subsystems 102 and 103 or, alternatively, in order to reduce the signal to noise ratio of said signals, the separation procedure can be performed from the average of a selectable number of segments of said signals synchronized with the cardiac cycle of the subject or animal. The possible alternatives corresponding to the extraction of information of intracranial origin are those described above.

Additionally, the invention may have a cathode ray tube, plasma screen or other display element to present the information corresponding to the signal provided by the processor 105.

Although a preferred embodiment of the invention is as described hereinbefore, a more generic embodiment of the invention can be represented as shown in Fig. 3. In it, the source of energy 301 (for example, of electromagnetic energy, for example, in the near infrared field) is applied to the head of the subject 310 from the outer surface of the scalp 304 by an application device 306. The type of energy is selected so that its dispersion and / or absorption inside the head of subject 310 depends on at least one hemodynamic parameter of the scalp and brain tissue. Simultaneously, a first sensor 307, specific to the type of energy applied (for example, a sensor configured to detect electromagnetic radiation in the near infrared field) and located in the vicinity of device 306, collects information on said parameter of leather hemodynamics scalp and takes it to electronic subsystem 302 for conditioning the received signal. Also simultaneously, a second sensor 308 similar to the first sensor 307, located at a suitable distance, collects information of said parameter of cerebral hemodynamics mixed in unknown proportions with information of said parameter from leather hemodynamics scalp, to take it to a subsystem 303 that conditions the signal. Finally, the 305 processor:

- collects the conditioned signals from the two subsystems 302 and 303 during a time window that includes at least the range of a blood pulse wave to the subject's head;

- processes these signals so that their variation in time coincides with that of the blood flow of the tissue from which they inspect the hemodynamic parameter (for example, if what the sensors detect depends directly on the blood volume (VSC / VSCC), the process may include the step of obtaining the derivative of the difference between the signal of the subsystem 303 and a fraction of the signal of the subsystem 302; if what the sensors detect depends directly on the blood flow (FSC / FSCC), the process may be limited to obtaining the difference , as such, between the subsystem signal 303 and a fraction of the subsystem signal 302);

- determines the fraction of the processed signal from subsystem 302 that must be subtracted from the processed signal from subsystem 303 so that the variation in time of the result of the subtraction (or the derivative of the subtraction, depending on whether they are dependent on volumes or flows) is the minimum possible; - obtains the signal resulting from subtracting said fraction of the original signal from subsystem 302 from the original signal from subsystem 303; and presents at the output the resulting hemodynamic parameter of said last subtraction.

Execution example of the invention. Although the patented method can be applied to techniques such as REG or NIRS, among others, the experiment carried out with the first of the techniques is detailed below. In this description, it is understood that any plethysmographic signal obtained in the head of a subject is composed of two components mixed in arbitrary proportions: a first one of these components, called extracranial, is caused by variations in blood content of the scalp tissue associated with the heartbeat, while a second component, called intracranial, is caused by variations in blood content of the brain tissue associated with the heartbeat. Eleven volunteers participated in the experiment, which were formally informed of their objectives. During the trial, the subject remained comfortably seated in a reclining seat, receiving specific instructions not to rest his head on the backrest and to remain with his eyes closed during the registration periods.

In the development of the experiment, a classic reencephalograph was developed and calibrated by the Department of Electronic Engineering of the Polytechnic University of Valencia (Spain). Said reoencephalograph formed part of the electronic system 100, including in it the source of electrical energy 101 and the electronic subsystems 102 and 103. The outputs of the classical reoencephalograph were the outputs of said subsystems, which were processed in a personal computer that made this experiment the processor function

105. Hereinafter, it will refer to the output of the subsystem electronic 102 as REG I, and at the output of electronic subsystem 103 as REG II.

To study the ability of the invention to obtain plethysmographic signals exclusively dependent on cerebral blood flow and independent of blood flow from the scalp, in each subject five different tests were performed, in which the current application electrodes 106 and 109 were always maintained. in the same position, varying between tests the positions occupied by the pick-up electrodes 107 and 108. According to the state of the art, the variation of the positions of the pick-up electrodes guarantees the variation of the proportions with which they are mixed the extra and intracranial component to form the plethysmographic signal of REG II that is extracted from subsystem 103.

Specifically, the positions occupied by the current application electrodes 106 and 109 were, in all tests, the electroencephalography positions C5 and C6 of the standard system 10-20, which correspond to positions 401 and 402 of Fig. 4. Between both positions, and always in the coronal plane, the eleven positions 403, 404, 405, 406, 407, 408, 409, 410, 411, 412 and 413 were located, of which the first ten were used to locate the pickup electrodes 107 and 108 in the different tests.

In order not to create a preferred conduction path of the injected current between the excitation electrodes located at 401 and 402, it was decided not to permanently place electrodes in the eleven intermediate positions. Instead, as measuring electrodes, they used two electrodecephalography electrodes of tripod type that, fastened with an elastic band, were placed in the positions where they were required in each of the five tests. Finally, to improve the signal-to-noise ratio of plethysmographic signals, the average of segments of said signals synchronized with the cardiac cycle was performed. For this, the electrocardiogram (ECG) was obtained from each subject, simultaneously with the capture of the plethysmographic signals in the so-called "lead I", for which a stainless steel electrode was placed on each wrist of the subject.

Registers After having placed the current application electrodes, that is, the excitation electrodes, in positions 401 and 402, an elastic silicone band was placed around the subject's head to hold the tripod electrodes. Next, and for the first of the tests, the tripod electrodes were placed in measuring positions 403 and 404 and the REG I, REG II and ECG of the subject captured on said electrode set were simultaneously recorded for 3 minutes . Next, the same process was repeated on the same subject by moving the pair of measuring electrodes to the pairs of positions 405 and 406; 407 and 408; 409 and 410; and, finally, 411 and 412. This process was repeated in each of the eleven voluntary subjects.

Digitization and preprocessing of data The signal pairs of REG I and REG II obtained in each of the tests, provided by the electronic subsystems 102 and 103 respectively, were taken to a personal computer that performed the tasks of the processor 105. First, said processor digitized the plethysmographic signals, together with the electrocardiographic signal, at a sampling rate of 500 samples per second.

On the other hand, and as mentioned above, in order to improve the signal-to-noise ratio of plethysmographic signals, synchronous segments were averaged with the heartbeat of said signals by the following procedure: the moments of ECG recording were detected in which the R wave (highest electric peak of the ECG representative of ventricular contraction) was produced; each plethysmographic signal was divided into signal segments, one for each cardiac cycle, extended from 100 ms before the R wave of the ECG to 700 ms after it, - and finally for each instant of time the values were averaged separately of all the signal segments obtained from each plethysmographic signal, each test and each subject. This procedure for improving the signal-to-noise ratio is classically used in the extraction of so-called evoked potentials.

Thus, in each trial and for each subject, two plethysmographic signals of REG I and REG II were simultaneously obtained, each of which is considered the result of the sum, in different and arbitrary proportions, of an extracranial component, reflecting the variations associated with the heart beat of the blood volume of scalp (VSCC), and other intracranial, reflecting the variations associated with the heartbeat of cerebral blood volume (VSC). Regarding these components, in this experimental study it is assumed that, for each subject, the morphology and phase with respect to the cardiac cycle of each component is identical regardless of the position of the electrodes used.

The nomenclature used hereinafter in the reference to the plethysmographic signals and their components is: R ₁ It) and R ₂ (t). - Averages of the segments extracted from the plethysmographic signals corresponding to the REG I and REG II registers, respectively, obtained in each trial and in each subject, adjusted to zero average.

C _JÍ (t) and C _Bi (t). - Intra and extracranial components ¹ , respectively, which together make up the average R¿ (t) register, which can be i = l or i = 2. ^c m (t) and ^C _In (t) .- Intra and extracranial components, normalized to unit variance, which make up any of the average R ₁ (t) and R ₂ (t) records. X [j] .- Discrete variable corresponding to the sampling of the continuous variable X (t), where the index "j" indicates the order of the sample, and the variable X (t) can be any of the above described (R ₁ (t), R ₂ (t), C _u (t), C _Ei (t), C ₁ Jt) and C _Bn (t)). By the above definitions, the average registers R ₁ (t) and R ₂ (t) can be expressed as

expresiones que, normalizando a varianza unidad las componentes intra y extracraneales proporcionan las siguientes ecuaciones R_x(t) = a_nC_En(t) + a_nC_In{t) (3 )

<hf*_><!)+ cι₂₂C_[n(t) i*⁾ donde los coeficientes a_±j y las componentes C_En(t) y C_1n (t) son desconocidas. De ambas ecuaciones, puede escribirse, restando a la Ec. (4) una fracción K de la Ec. (3),

expressions that, normalizing to unit variance the intra and extracranial components provide the following equations R _x (t) = a _n C _In (t) + a _n C _In {t) (3)

<hf * _> <!) + cι ₂₂ C _[n (t) i * ⁾ where the coefficients a _{± j} and the components C _En (t) and C _1n (t) are unknown. From both equations, one can write, subtracting a fraction K from Eq. (3) from Eq. (4),

Ka_nyθ_En(t)+ (a₂₂- Ka_nyθ_In(t) ⁽5⁾ donde K es una constante cualquiera.R ₂ (O-

Ka _n yθ _In (t) + (a ₂₂ - Ka _n yθ _In (t) ⁽ 5 ⁾ where K is any constant.

Given Eq. (5), the objective is to know the value of K, characteristic of each individual and of each electrode position, which cancels the term that multiplies the extracranial component C _Bn (t), which will be obtained, from the weighted subtraction of the average registers R _x (t) and R ₂ (t), the intracranial component multiplied by a constant.

Component Separation

Since the intracranial component is due to VSC variations and the extracranial component is due to VSCC variations, those derived from said volumes, and therefore from said components, are directly related to cerebral blood flow

(FSC) and with scalp blood flow (FSCC) respectively.-

As previously stated, the waveform of FSC 201 under normal conditions is minimally pulsatile compared to that of FSCC 202, which is a succession of impulse functions. Therefore, the signal resulting from the weighted difference of the plethysmographic signals expressed in Eq. (5) will show a more pulsatile derivative the higher the information content from the FSCC that is contained in said difference signal.

The separation procedure proposed in the present invention is based precisely on this criterion: the FSC resulting from the separation of the intra and extracranial components must be such that the pulsatility of the result should be the minimum possible. In other words, the value of the constant K of Eq. (5) that annuls the term that multiplies the extracranial component C _In (t), must minimize the variance of the derivative of the first term of equality.

For this, an indicator of variability J (K) is defined to minimize that, in case of sampled variables, it would be of the form

J (K) = E ^ R ₂ [Jj R ₂ [J - I] - K (R ₁ [J and R ₁ [J - I}) '} ⁽⁶⁾ where function E is called the "expected value", "mathematical hope" or "average value". Note that Eq. (6) of definition of function J (K) could be written in the case of continuous functions (not sampled) as

donde T sería el intervalo de tiempo de cálculo de variabilidad.

where T would be the time interval of variability calculation.

According to the above criteria, the K ^* value that minimizes the previous cost function transforms Eq. (5) into

R ₂ (O- K ^* R _x (t) = (a ₂₂ - K ^* a _n V _ln (t) < ⁸ >

Finally, the problem is greatly simplified if, as stated in the state of the art, it is admitted that the variations associated with the heartbeat of the signal Reglet plethysmography of REG I are caused exclusively by extracranial blood circulation, which cancels the constant at ₁₂ in Eq. (3) defining .R ₁ (U), so Eq. (8) remains in such case how

R ₂ (O- K ^* R _λ (t) = a ₂₂ C _In (t) ~ C _n (t) < ⁹ > where C ₁₂ (t) is the intracranial component collected in REG II The extracranial component of REG II C _B3 (t) will be, finally

Once the value of the constant K * has been determined for each subject and for each electrode position, the processor 105 finally obtains the desired plethysmographic signal, dependent on the FSC and independent of the FSCC, subtracting a fraction of the plethysmographic signal obtained by the subsystem 103 a fraction K * of the plethysmographic signal obtained by subsystem 102.

Assessment of the goodness of the method and apparatus of the invention.

The method and apparatus of the invention considers each of the R ₂ (t) average registers, typical of each subject and each electrode position, as the weighted sum of two components C ₁₂ (t) and C _E2 (t) . On the other hand, the experiment carried out obtained from each subject five average R ₂ (t) records from the plethysmographic signals obtained with the subsystem 103 in five pairs of different electrode locations, so that the weighting coefficients with which add both components vary in each trial (see Pérez JJ, Pebble E, Barcia JA, "Quantification of intracranial contribution to rheoencephalography by a numerical model of the head ", Clinical Neurophysiology 111 (2000); 1306-1314). To assess the ability of the method and apparatus of the invention to separate the intra and extracranial components, the following criteria were adopted:

Premise: For each subject, the components C _In (t) and C _In (t), whose weighted sum make up the average registers R ₁ (t) and R ₂ (t), are invariant, regardless of the location of the electrodes that is used and the period of time in which the registration is made.

Assessment criterion: The method and apparatus for obtaining plethysmographic signals dependent on the FSC and independent of the FSCC is considered the better the higher the degree of satisfaction of the previous premise. Quantification of criteria: The discrimination ability of the component C ₁₂ (t) of the method ^"and apparatus of the invention is quantified, according to the assumption by the average value of the assembly formed by the correlation coefficients between all pairs of these components of each subject. If the average value of the elements of the set of said correlation coefficients is significantly higher than that obtained from the average R ₂ (t) records, it is understood that the method appropriately separates the intracranial component from the extracranial one.

Statistical treatment: For the statistical treatment of the results, the correlation coefficients obtained are normalized by transformation. Fisher's Z Results Representation As an example, the results obtained in one of the subjects are shown in Figs. 5, 6, 7 and 8. Specifically, Fig. 5 and 6 show, respectively, the average registers R _x (t) and R ₂ (t) normalized to unit variance obtained in said subject. As can be seen, the morphology of the five average R ₁ (t) records obtained in said subject is practically identical, since they are taken in the same subject and with the same electrode position, being the minimum differences between them and due to the fact that said records have been obtained in different trials and, therefore, at different time intervals.

According to the conclusions derived from the clinical trials carried out by other authors and by the theoretical study previously published by the inventors

(See Pérez JJ, Pebble E, Barcia JA, "Quantification of intracranial cσntribution to rheoencephalography by a numerical! model of the head", Clinical Neurophysiology 111

(2000); 1306-1314), said R _x (t) records are caused by the variation in time, beat to beat, of the VSCC.

In Fig. 5, the instant of time 501 corresponds to the R wave of the ECG, that is, with the instant at which the ventricular contraction occurs. Shortly after, the instant of time 502 occurs, which corresponds to the moment in which the blood pulse wave, caused by ventricular contraction, reaches the arteries of the scalp. From the instant of time 502 to 503, the blood filling of the scalp arteries suddenly occurs, causing the abrupt decrease in impedance represented by the average R _x (t) registers of Fig. 5 (the 501, 502 and 503 instants in figures 6-8 they correspond to those in figure 5). As can be seen, the derivative of said registers would have a similar shape to that of the FSCC 202 curve shown above. This pattern of the average R ₁ (t) records was also seen in the records of the rest of the voluntary subjects who participated in the study.

Contrasting with the morphological identity of the family of curves described, in Fig. 6 the five average R ₂ (t) registers obtained in the same previous subject are shown, in which the instant of time 501 marks the R wave. The figure shows that the morphology of the average register R ₂ (t) depends strongly on the position of the second pick-up electrodes 107 and 108. This is due, as has been repeatedly mentioned in this document, to the weight with the that the intra and extracranial components are mixed to form the plethysmographic signals of REG II strongly depends on said electrode position. In fact, the morphological diversity of the average R ₂ (t) records constitutes robust evidence that the intra and extracranial components have different morphologies. Even more specifically, it can be seen that some of the average registers in Fig. 6, instead of showing an abrupt decrease in the impedance represented by said register between the instants of time 502 and 503, shows on the contrary an increase in the impedance, indicating that the effect on the measured impedance of the VSCC variations changes sign. Theoretical studies conducted by the authors of the present invention predict such behavior of REG II (see Pérez JJ, Pebble E ₁ Barcia JA, "Quantiflcation of intracranial contribution to rheoencephalography by a numerical model of the head", Clinical Neurophysiology 111 (2000); 1306-1314).

The results of the separation of extra and intracranial components are shown in Figs. 7 and 8 respectively. The components C _Sn (t) that make up the five average registers R ₂ (t) of the example subject are shown superimposed in Fig. 7. In said figure, time instant 501 indicates, as in Figure 5, the temporal position of the R wave; the instants 502 and 503 are identical to the instants 502 and 503 of Figure 5. The five curves show identical morphologies, due to the fact that the restrictions imposed on the processing assign to the extracranial component C _Eα (t) the morphology of the register average R _x (t) which, as previously mentioned, is virtually invariant for each subject having obtained their five records using the same electrode location.

In Fig. 8, finally, the five intracranial components obtained in the example subject are shown superimposed from their average R ₂ (t) registers shown above in Fig. 6. Obtaining these registers is the main objective of the invention of the method and apparatus set forth herein, and represent the variations of the VSC of said subject associated with the heartbeat. As already indicated in relation to Figure 5, the instant of time 501 corresponds to the R wave of the electrocardiogram of the subject, while the instants of time 502 and 503 are those that have already been described, in which he appreciated the sudden arrival of the blood pulse wave to the subject's scalp. As can be seen in this Fig. 8, the five registers represented have a high morphological similarity between them indicating, in the absence of statistical assessment, the goodness of the method and apparatus of the present invention. In addition, as can be seen, the family of resulting curves does not show any sudden change in the time window between time 502 and 503, which suggests that these curves are independent of the FSCC. Table 1 shows the results of the statistical analysis performed, presenting the average values and ranges for α = 0.05 of the correlation coefficients obtained between any pair of average records R ₁ (t) and R ₂ (t), and all pairs of components extracted from the average R ₂ (t) records of each subject.

As can be seen in this table, the values corresponding to the average register R ₁ (t) and the extracranial component C _E2 (t) coincide, since both are morphologically identical since the average register R ₁ (t) is considered to be an exclusive reflex of the VSCC variations. On the other hand, the correlation coefficients of the average R ₂ (t) registers have a relatively high average value, although extended over a very wide range, which suggests that, for each individual, the morphology of the register depends strongly on the position of the reading electrodes. This can be understood by considering the fact that the weights with which the intra and extracranial components that make up the average R ₂ (t) register vary with the electrode position. However, the average value of the correlation coefficients in the case of intracranial components C ₁₂ (t) extracted from the average R ₂ (t) records is greater than that of the average R ₂ (t) records themselves, being framed in a fork noticeably narrower than that of the average R ₂ (t) records. In addition, the statistical analysis indicates that there are significant differences between both variables with a confidence interval P <0.05. These facts show that, in accordance with the premise established for the assessment of the procedure, the proposed separation method, on which the apparatus of the invention is based, manages to extract from the original plethysmographic signals the component that is due to the variations, beat to beat, of the VSC.

Range p ^ (α = 0.05)

(t); ^C E, ₂ (t) 0 .9947 0 .9880 0.9977

, ₂ (t) 0 .9675 0 .7279 0.9965

R ₂ (t) 0 .9493 0 .1789 0.9981 Table 1. Average values and ranges for α = 0.05 of the correlation coefficients obtained between all pairs of records and components extracted from each subject.

Finally, it should be understood that, after the processor 105 determines the appropriate value of the constant K for the position used for current and measurement injection electrodes, said processor delivers at its output the plethysmographic signal resulting from directly subtracting the signal from the electronic subsystem 103 a fraction JC * of the plethysmographic signal from the electronic subsystem 102, without the need for any additional calculation as long as the positions of the injection and measurement electrodes remain unchanged. In this text, the word "comprises" and its variants (such as "understanding", etc.) should not be construed as excluding, that is, they do not exclude the possibility that what is described includes other elements, steps, etc.

On the other hand, the invention is not limited to the specific embodiments that have been described but also covers, for example, the variants that can be made by the average person skilled in the art (for example, in terms of the choice of materials, dimensions , components, configuration, etc.), within what follows from the claims.

Claims

1. - Apparatus for obtaining information relating to cerebral hemodynamics of a human or animal subject, characterized in that it comprises: means of application (101, 106, 109; 301, 306) to the head (104; 304) of the subject, of at least one excitation signal (Se); pick-up means (102, 103, 106-109; 302, 303, 307, 308) configured for simultaneous collection, at least two different positions of the subject's head, of a first output signal (Sl) and a second output signal (S2); and calculation means (105, 305) configured to calculate the value of a function (F) of a difference (S2-KS1) between said second output signal (S2) and a fraction (K) of said first output signal ( Sl), said function (F) being selected so that it represents an indication of blood flow and the calculation means (105, 305) being configured to calculate said fraction (K) so that the variability of the value of said function (F ) is substantially minimal over a preselected time interval, according to a selected variability criterion. 2. Apparatus according to claim 1, characterized in that said function (F) is the temporary derivative of said difference between said second output signal (S2) and said fraction (K) of said first output signal (Sl). 3. Apparatus according to claim 1, characterized in that the function (F) is a function whose value is proportional to said difference between said second output signal (S2) and said fraction (K) of said first output signal (Sl) . 4. - Apparatus according to any of the preceding claims, characterized in that it comprises means for selecting said time interval so as to include a significant part of a cardiac cycle of the subject. 5. Apparatus according to claim 4, characterized in that it comprises means for selecting said time interval so as to include the time of the arrival of a blood pulse wave to the subject's head. 6. Apparatus according to any of the preceding claims, characterized in that the application means and the capture means are configured to apply the excitation signal (Se) in at least one excitation position in the subject's head, to capture the first output signal (Sl) in at least a first position closer to said excitation position, and to pick up said second output signal (S2) in at least a second position further away from said excitation position. 7. - Apparatus according to any of the preceding claims, characterized in that the calculation means (105) are configured to calculate the value of the function (F) from an average of the first signal of output (Sl) and an average of the second output signal (S2) over a plurality of cardiac cycles. 8. - Apparatus according to any of the preceding claims, characterized in that the means for applying the excitation signal comprise means (101) for generating an electrical signal. 9. Apparatus according to claim 8, characterized in that the means for applying the excitation signal additionally comprise at least two excitation electrodes (106, 109) applicable on individual excitation positions. 10. Apparatus according to claim 9, characterized in that the pick-up means comprise at least two first pick-up electrodes (106, 109) applicable on first output positions, to capture said first output signal (Sl). 11. Apparatus according to claim 10, characterized in that the excitation electrodes (106, 109) and the first pick-up electrodes (106, 109) are part of an electrode structure in which the first pick-up electrodes are located close to the excitation electrodes. 12. Apparatus according to claim 11, characterized in that the first collection electrodes (106, 109) are located together with the excitation electrodes (106, 109). 13. Apparatus according to claim 12, characterized in that the first collection electrodes (106, 109) are constituted by the excitation electrodes (106, 109). 14. Apparatus according to claim 11, characterized in that the first collection electrodes are at a distance of less than 15 mm from the respective excitation electrodes. 15. Apparatus according to any of claims 9-14, characterized in that the pick-up means comprise at least two second pick-up electrodes (107, 108) to capture said second output signal (S2). 16. Apparatus according to claim 15, characterized in that said second pick-up electrodes (107, 108) form part of an electrode structure in which the second pick-up electrodes (107, 108) are further away from the excitation electrodes ( 106, 109) corresponding, than said first pickup electrodes (106, 109). 17. Apparatus according to claim 16, characterized in that each second pick-up electrode (107, 108) is at a distance greater than 15 mm from the nearest excitation electrode. 18. Apparatus according to any of the preceding claims, characterized in that the collection means they are configured so that the first output signal is a bipolar reoencephalography output signal (REG I) and so that the second output signal is a tetrapolar reoencephalography output signal (REG II). 19. Apparatus according to any of claims 1-7, characterized in that the excitation means comprise means (301) for generating an electromagnetic excitation signal (Se) and because the collection means comprise electromagnetic energy sensors (307, 308). configured to capture the first output signal (Sl) and the second output signal (S2). 20. Apparatus according to claim 19, characterized in that said excitation signal is an infrared radiation signal. 21. Apparatus according to any of the preceding claims, characterized in that the collection means (102, 103, 106-109; 302, 303, 307, 308) are configured so that said first output signal (Sl) depends substantially. of the scalp blood flow and not substantially depend on cerebral blood flow, and said second output signal (S2) depends on the cerebral blood flow and the scalp blood flow. 22.- Method of obtaining information related to cerebral hemodynamics of a human or animal subject, which includes the steps of: apply to the subject's head, at least one excitation signal (Se); to capture, during a capture period and in at least two different positions of the subject's head, a first output signal (Sl) and a second output signal (S2), both the first output signal (Sl) and the second output signal (S2) being dependent on the excitation signal (Se) and physicochemical characteristics of the scalp and / or the subject's brain, so that the relationship between said output signals (Sl, S2) and said excitation signal (Se) depends on the blood flow of the scalp and / or the cerebral blood flow of the subject, said first output signal (Sl) and said second signal being output (S2) captured at selected positions so that said first output signal (Sl) depends substantially on the blood flow of the scalp and does not depend substantially on cerebral blood flow, and said second output signal (S2) depends on blood flow cerebral and blood flow of the scalp; calculate the value of a function (F) of a difference (S2-KS1) between said second output signal (S2) and a fraction (K) of said first output signal (Sl), said function (F) being selected so that it represents an indication of blood flow and said fraction (K) being selected so that the variability of the value of said function (F) is substantially minimal over a preselected time interval, in accordance with a selected variability criterion. - (J ₄ - 23.- Method according to ^'claim 22, wherein said first output signal (Sl) has at all times a value which substantially depends on the blood volume of the scalp and does not substantially depends on the cerebral blood volume, and said second Output signal (S2) has a value at all times that depends on the cerebral blood volume and the scalp blood volume; said function (F) being the temporary derivative of said difference between said second output signal (S2) and said fraction (K) of said first output signal (Sl). 24. Method according to claim 22, characterized in that said first output signal (Sl) has at each moment a value that depends substantially on the blood flow of the scalp and does not depend substantially on the cerebral blood flow, and said second signal Outlet (S2) has a value at all times that depends on cerebral blood flow and scalp blood flow; said function (F) being a function whose value is proportional to said difference between said second output signal (S2) and said fraction (K) of said first output signal (Sl). 25. Method according to any of claims 22-24, characterized in that said Time interval includes a significant part of a subject's cardiac cycle. 26.- Method according to claim 25, characterized in that said time interval includes the time of the arrival of a blood pulse wave to the subject's head. 27. Method according to any of claims 22-26, characterized in that the excitation signal (Se) is applied in at least one excitation position in the subject's head, obtaining the first output signal (Sl) in at least one first position closer to said excitation position, and said second output signal (S2) being obtained in at least a second position further away from said excitation position. 28.- Method according to any of claims 22-27, characterized in that the value of the function (F) is calculated from an average of the first output signal (Sl) and an average of the second output signal ( S2) over a plurality of cardiac cycles. 29. Method according to any of claims 22-28, characterized in that the excitation signal is an electrical signal. 30. A method according to claim 29, characterized in that the excitation signal is applied with at least two excitation electrodes (106, 109) applied on respective excitation positions. 31. Method according to claim 30, characterized in that the first output signal (Sl) is captured with at least two first pick-up electrodes (106, 109) applied on respective first output positions. 32.- Method according to claim 31, characterized in that the first output positions are close to the excitation positions. 33. Method according to claim 31, characterized in that the first output positions are identical to the excitation positions. 34. Method according to claim 33, characterized in that the first pick-up electrodes (106, 109) comprise the excitation electrodes. 35. Method according to claim 31, characterized in that the first output positions are at a distance of less than 15 mm from the respective excitation positions. 36.- Method according to any of claims 30-35, characterized in that the second output signal (S2) is captured with at least two second pick-up electrodes (107, 108) located in respective second output positions further away from the positions of excitation than said first exit positions. 37. Method according to claim 36, characterized in that the second output positions are at a distance greater than 15 from the respective nearest excitation positions. 38.- Method according to any of claims 22- 37, characterized in that the first output signal is a bipolar reoencephalography output signal and that the second output signal is a tetrapolar reoencephalography output signal. 39.- Method according to any of claims 22- 38, characterized in that the excitation signal (Se) is an electromagnetic signal and that the first output signal (Sl) and the second output signal (S2) are captured with respective electromagnetic energy sensors (307, 308). 40. A method according to claim 39, characterized in that the excitation signal is an infrared radiation signal. 41.- Signal analysis method, applicable to signals captured on the head of a subject comprising a first output signal (Sl) and a second output signal (S2) that both are dependent on the same excitation signal (Se ) and of physicochemical characteristics of the scalp and / or of the subject's brain, the method comprising the step of: - calculating the value of a function (F) of a difference (S2-KS1) between said second output signal (S2) and a fraction (K) of said first output signal (Sl), said function (F) being selected so that it represents an indication of blood flow and said fraction (K) being selected so that the variability in the value of said function (F) is substantially minimal over a range of preselected time, according to a selected variability criterion; providing data related to cerebral blood flow based on said difference (S2-KS1) between said second output signal (S2) and said fraction (K) of said first output signal (Sl). 42.- Method according to claim 41, characterized in that said function (F) is the temporary derivative of said difference between said second output signal (S2) and said fraction (K) of said first output signal (Sl). 43.- Method according to claim 41, characterized in that said function (F) a function whose value is proportional to said difference between said second output signal (S2) and said fraction (K) of said first output signal (Sl ). 44. - Apparatus according to any of claims 1-21, characterized in that it is configured to carry out the method according to any of claims 22-40. 45.- Computer program, characterized in that it comprises a program code configured to carry out the method according to any of claims 21-43, when executed in a programmable electronic device.