RU2546099C1 - Method for visualising cutaneous blood flow variations in extremities - Google Patents

Abstract

FIELD: medicine.

SUBSTANCE: thermal imaging camera is used to determine skin temperature and its time variations. Temperature variations determined in each point of a thermal image of the extremity are broken down into spectral components with using mathematical wavelet transformation. Each spectral component frequency f is shifted to a previous moment of time by an interval Δt determined by formula $$\Delta t_{\text{PHASE}}(f_i,z)=z/\left(2\cdot \sqrt{\frac{\pi \cdot \lambda }{c\cdot \rho }\cdot f_i}\right),$$

wherein z is a thickness of biotissue layer, λ is a skin conductivity coefficient, c is a specific thermal capacity of skin, ρ is a skin density, f_i is a frequency of the i-th spectral component. Amplitude of each spectral component is multiplied by a coefficient determined by formula $$C_{AMP}(f_i,z)=\exp \left(z\cdot \sqrt{\frac{\pi \cdot c\cdot \rho }{\lambda }\cdot f_i}\right).$$

That is followed by reverse wavelet transformation of spectral components in each point of the thermal image to produce a resultant data array representing the blood flow variations distribution.

EFFECT: enlarging the examined surface area of the object and providing more accurate measurement of the peripheral blood parameters by temperature methods by using a new model of temperature signal propagation in biological tissue and visualising the spatial blood flow variations.

2 cl, 6 dwg, 1 tbl

Description

The invention relates to medicine, namely to functional diagnostics, and can be used to determine fluctuations in cutaneous blood flow in the limbs using the results of temperature measurements on the skin surface.

Blood flow in the extremities has long been an object of increased interest of physiologists and clinicians, due to its pronounced variability, both at rest and during stress tests (Burton, AC A study of the adjustment of peripheral vascular tone to the requirement of the Regulation of body temperature / AC Burton, RM Taylor // Am. J. Physiol.- 1940. - Vol. 129, - P. 566-577; Burton, AC Range and variability of the blood flow in the human fingers and the vasomotor Regulation of body temperature / AC Burton // Am. J. Physiol. - 1939. - Vol. 127. - No. 3. - P. 437-453; Shitzer, A. Simultaneous measurements of finger-tip temperatures and blood perfusion rates in a cold environment / A. Shitzer, A. Stroschein, MW Sharp, RR Gonzalez, KB Pandolf // J ournal of thermal biology. - 1997, - Vol.22, - No. 3, - P. 159-167).

Technical field

The most common methods for controlling skin blood flow fluctuations are laser Doppler flowmetry (LDF) and photoplethysmography (PPG) (Wright CI, Kroner CI and Draijer R. 2006, Non-invasive methods and stimuli for evaluating the skin's microcirculation Journal of pharmacological and toxicological methods 54, 1-25). The measurement of blood flow fluctuations by LDF and FPG methods has a number of limitations associated primarily with the influence of the contact of biological tissue with the sensor and the heterogeneity of the distribution of blood flow in the probed area. In this regard, non-contact methods should be an advantage, with the help of which it is possible to collect information on the distribution of blood flow in the studied tissue region and map the amplitude values of the oscillations, i.e. visualize blood flow fluctuations.

A well-known method for studying skin blood flow using laser Doppler imaging (LDV) (Serov A., Steinacher B., Lasser T. Full-field laser Doppler perfusion imaging and monitoring with an intelligent CMOS camera // Optics Express. - 2005. - T. 13. - No. 10. - S. 3681-3689) and the speckle-contrast method LASCA (Briers, JD, & Webster, S. (1996). Laser speckle contrast analysis (LASCA): a nonscanning, full-field technique for monitoring capillary blood flow Journal of biomedical optics, 1 (2), 174-179). When using these methods of visualization of blood flow, difficulties arise in determining the concentration of blood (and consequently, blood flow) because it depends on the amplitude of the laser radiation reflected or scattered by biological tissue. If there is a change in the structure of the tissue, for example, a transition from the skin to the nail, then an increase in the signal amplitude can be observed due to multiple scattering of radiation by the nail plate, while the structural optical properties of the nail are individual for each subject and there is ambiguity in determining the concentration of blood under the nail plate (Tuchin V .V. Optics of biological tissues: methods of light scattering in medical diagnostics, trans. From English V.L. Derbova, Moscow: FIZMATLIT - 2013, 811 pp.).

An alternative to laser methods for visualizing blood flow fluctuations can be methods of thermometry. The use of thermal imaging methods for blood flow analysis provides non-contact measurements, since they are based on registration of the object’s own radiation. Thermal imaging methods provide a high speed temperature measurement (50 or more measurements per second) and a wide area of coverage of the surface of the object. The difficulties in using thermal imaging methods to visualize blood flow fluctuations are mainly associated with the difference in the form of fluctuations in skin blood flow and skin temperature. At the same time, there is the problem of establishing an unambiguous correspondence between changes in skin temperature and blood flow, which forces researchers to consider temperature and blood flow fluctuations as independent fluctuations, between which there is some correlation.

Closest to the claimed solution is the "Method for registering blood microcirculation" (Podtaev S.Yu., Ershova A.I., Popov A.V., Morozov MK. Pat. No. 2390306, published on 05.27.2010, Bull. No. 15 ) In accordance with this method, the level of blood microcirculation is estimated by recording the temperature on the nail phalanx of the palm surface of the index finger of the patient. The temperature is continuously recorded using a contact temperature sensor for 10 minutes in the initial state (during the indicated duration, several periods of oscillations are recorded in the endothelial, neurogenic and myogenic ranges), then for 3 minutes during a breathing or cold test and for 10 minutes after samples. The conclusion about the state of blood flow regulation is based on the correlation established by the authors between skin temperature and blood flow fluctuations recorded by the Doppler flowmeter (Podtaev S. Wavelet-based correlations of skin temperature and blood flow oscillations / S. Podtaev, M. Morozov, P. Frick // Cardiovasc. Eng. - 2008. - Vol.8. - N3. - P.185-189). To obtain information about fluctuations in blood flow in the respiratory, myogenic, neurogenic and endothelial ranges, a wavelet analysis of temperature changes is used.

In this method, the conclusion about the regulation of blood flow is made on the basis of the existence of a correlation of temperature and blood flow, while the calculation of the correlation is carried out using the source signals without any changes. Since the signals of fluctuations in temperature and blood flow have different shapes, an adequate description of the fluctuations in blood flow requires a model to explain the differences in waveforms. As will be shown below, the use of model representations of the proposed method allows you to convert the temperature signal so that its shape is close to the form of fluctuations in blood flow.

However, in accordance with the indicated method, measurements are carried out only in the local area of the index finger and do not allow to assess the distribution of blood flow in other fingers and hands, which would open up the possibility of diagnosis, for example, disorders of the innervation of the ulnar, radial and median nerves, comparison of the symmetry of the distribution of blood flow on two opposite limbs etc.

The considered method for the analysis of blood flow parameters uses temperature measurements, and wavelet analysis of temperature fluctuations is used to process the experimental data, so the "Method for registering blood microcirculation" is taken as a prototype.

The objective of this solution is to visualize fluctuations in skin blood flow in the extremities by registering the dynamics of skin temperature distribution (dynamic thermograms) with a thermal imaging camera, mathematical processing of temperature data and converting thermograms into maps of the distribution of blood flow oscillations.

The technical result consists in increasing the studied surface area of the object and improving the accuracy of determining the parameters of peripheral blood flow by temperature methods through the use of a new model of the distribution of the temperature signal in biological tissue and visualization of spatial changes in blood flow fluctuations.

, вносящую временной сдвиг Δt_PHASE спектральных составляющих в зависимости от частоты f, и формулу $$C_{\text{AMP}}(f_i,z)=\exp \left(z\cdot \sqrt{\frac{\pi \cdot c\cdot \rho }{\lambda }\cdot f_i}\right)$$

, обеспечивающую преобразование амплитуд C_AMP спектральных составляющих, где z - толщина слоя биоткани, λ - коэффициент теплопроводности кожи, с - удельная теплоемкость кожи, ρ - плотность кожи, f_i -частота i-й спектральной составляющей, затем к преобразованным спектральным составляющим применяют обратную формулу спектрального синтеза, при этом полученный сигнал будет представлять собой колебания кровотока, а сопоставление определенного цвета каждому значению амплитуды колебаний кровотока в каждой точке термограммы позволит в результате визуализировать колебания кожного кровотока с помощью математической обработки динамических термограмм.The specified technical result is achieved by the fact that a method for visualizing fluctuations in skin blood flow in the extremities, including thermal imaging measurements of temperature fluctuations of the extremities and spectral analysis of temperature fluctuations, according to the solution, provides for recording thermograms of the subject's limbs for at least 20 minutes, after which temperature fluctuations are exposed at each point of the thermogram spectral analysis (for example, wavelet analysis), the result of which is the separation of individual spectral components guides, whose amplitude varies with time, then to the obtained spectral components employed formula $$\Delta t_{\text{PHASE}}(f_i,z)=z/\left(2\cdot \sqrt{\frac{\pi \cdot \lambda }{c\cdot \rho }\cdot f_i}\right)$$

introducing the time shift Δt _{PHASE of the} spectral components depending on the frequency f, and the formula $$C_{\text{Amp}}(f_i,z)=\exp \left(z\cdot \sqrt{\frac{\pi \cdot c\cdot \rho }{\lambda }\cdot f_i}\right)$$

that provides the conversion of the amplitudes C _{AMP of the} spectral components, where z is the thickness of the biological tissue layer, λ is the thermal conductivity of the skin, c is the specific heat of the skin, ρ is the skin density, f _{i is the} frequency of the i-th spectral component, then the inverse spectral components are applied a spectral synthesis formula, and the received signal will represent fluctuations in blood flow, and matching a certain color to each value of the amplitude of blood flow fluctuations at each point of the thermogram will allow to visualize fluctuations in skin blood flow using mathematical processing of dynamic thermograms.

The invention is illustrated by drawings. In FIG. 1a shows the time dependences of temperature (dashed line) and blood flow (the solid line is the envelope of the PPG, measured in volts), FIG. 1b - normalized values of blood flow fluctuations (scale on the left) and restored blood flow fluctuations using temperature data (scale on the right). In FIG. 2 - spectral components of temperature fluctuations (dashed lines) and blood flow oscillations (solid lines) constructed for frequencies 0.01, 0.03, 0.05 Hz; for each frequency, different values of the temperature signal delay times Δt ₁ , Δt ₂ , Δt ₃ are noted, decreasing with increasing frequency. FIG. 3 - frequency dependence of the delay of the spectral components of the temperature signal relative to the components of the fluctuations of blood flow. FIG. 4 - frequency dependence of the attenuation of the spectral components of the temperature signal S _T relative to the components of the fluctuations in blood flow S _BF . FIG. 5 is an example of visualization of blood flow fluctuations in the limbs, FIG. 6a - the initial signal of fluctuations in the temperature of the finger; 6b, c - restored signals of fluctuations in blood flow in the endothelial (Fig. 6b) and neurogenic (Fig. 6c) ranges.

SUMMARY OF THE INVENTION

Fluctuations in blood flow at rest are caused by both a reaction to a change in ambient temperature and spontaneous regulation of vascular tone, which causes low-frequency fluctuations in volumetric blood flow in the respiratory, myogenic, neurogenic and endothelial frequency ranges (Bernjak, A. Low-frequency blood flow oscillations in congestive heart failure and after β1-blockade treatment / A. Bernjak, PBM Clarkson, PVE McClintock, A. Stefanovska // Microvascular Research, -2008. - Vol. 76, - No. 3, - P. 224-232).

To restore blood flow fluctuations according to the results of temperature measurements, it is necessary to determine the relationship between the two signals. The relationship between temperature fluctuations of the extremities and skin blood flow fluctuations was established as a result of our own research. The studies measured fluctuations in the temperature of the skin of the fingers and blood flow of a group of subjects who were under normal conditions at rest for 20 minutes, with a sampling frequency of 2 Hz. The blood flow signal was estimated by the envelope of the photoplethysmographic (PPG) signal recorded using a KL-79102 reflective sensor (Taiwan). The temperature was determined using a ThermaCam SC 3000 Flir Systems thermal imager (Sweden) with a temperature sensitivity of 0.02 ° C.

Signal registration within 20 minutes is necessary to obtain data on endothelial vibrations, the largest period of which is 200 seconds. Thus, over a period of 20 min = 1200 s, it will be possible to observe 6 complete periods of oscillation.

The group of subjects was 31 people: 21 men and 10 women aged 20-35 years. The measurements were carried out after the adaptation of the subjects to laboratory conditions for 10-15 minutes. Before measurements, subjects did not consume tonic or alcoholic beverages. All subjects were non-smokers. The measurements were carried out in the position of the test subjects sitting, hands were placed on a table with a surface made of a material having a low heat capacity (polystyrene). The index finger was located on top of the FPG sensor. Near the FPG sensor, the average temperature of the distal phalanx of the finger was recorded. The type of recorded signals is shown in FIG. 1a.

In accordance with the model used, blood flow oscillations generate a heat wave propagating from a certain depth to the skin surface. The wave is characterized by attenuation and a finite propagation velocity. Therefore, between the signals of fluctuations in temperature and blood flow, a connection can be established by analyzing fluctuations at individual frequencies. In this case, the frequency dependence of the delay and attenuation of the temperature signal relative to the blood flow signal will characterize the properties of the skin.

To study the relationship between the spectral components of temperature and blood flow fluctuations, a wavelet analysis with a Morlet basis was used. The use of the Morlet basis makes it possible to unambiguously compare the scale of the basis with the frequency of harmonic oscillations. The resulting spectral components had the form shown in FIG. 2. Based on these data, the time shift of the spectral components of the temperature relative to the components of the bloodstream and the analysis of the attenuation of the spectral components of the temperature were analyzed. As a result, the frequency dependences of the temporal shift of the spectral components of the temperature averaged over the group of subjects were constructed. 3 and attenuation of the spectral components of FIG. four.

The analytically reduced frequency dependences of the time shift and attenuation are described by the expressions

,(1) $$\Delta t_{\text{PHASE}}(f_i,z)=z/\left(2\cdot \sqrt{\frac{\pi \cdot \lambda }{c\cdot \rho }\cdot f_i}\right)$$

,(one)

,(2) $$C_{\text{Amp}}(f_i,z)=\exp \left(z\cdot \sqrt{\frac{\pi \cdot c\cdot \rho }{\lambda }\cdot f_i}\right)$$

, (2)

where z is the thickness of the biological tissue layer, λ is the coefficient of thermal conductivity of the skin, s is the specific heat of the skin, ρ is the density of the skin, f _i is the frequency of the i-th spectral component.

The spectral component of blood flow S _BF can be obtained by converting the spectral component of the temperature S _T according to the formula

(3) $$S_{BF}(f_i,t_j)=C_{AMP}(f_i,z)\cdot S_T\left(f_i,t_j+\Delta t_{PHASE}(f_i,z)\right)$$

(3)

Carrying out the synthesis of the spectral components of S _BF (for example, using the inverse wavelet transform formula), we obtain a signal considered as blood flow oscillations recovered from temperature data. In FIG. Dependencies are shown in Fig. 1b, demonstrating that the shape of the blood flow oscillations recovered from the temperature data (dashed line) is close to the shape of the oscillations recorded by the blood flow sensor (solid line).

Using the described algorithm at each point of the thermogram allows you to visualize the distribution of blood flow fluctuations in the limbs. Examples of reconstructed maps of the distribution of fluctuations in blood flow of the upper extremities are shown in FIG. 5. In FIG. 6a shows an example of a temperature curve in the area of a finger. The endothelial blood flow fluctuations recovered from this curve are shown in FIG. 6b, and the neurogenic range in FIG. 6c.

To check the reliability of blood flow fluctuations recovered from temperature data, they were compared with the results of photoplethysmographic examination in the neurogenic and endothelial ranges for a group of subjects. The results of the comparison are shown in table 1.

Table 1. Average values of the correlation of blood flow fluctuations recovered from temperature data and fluctuations measured by the PPG-sensor blood flow. (Averaging was performed over a group of 31 subjects; in parentheses are the values of 1 and 3 quartiles of the distribution of correlation values).

Correlation of the spectral components of temperature and blood flow fluctuations Endothelial range
(0.005 - 0.02 Hz) Neurogenic range
(0.02 - 0.05 Hz) Endothelial + Neurogenic Range
(0.005-0.05 Hz) Source signals 0.19
(0.09, 0.28) 0.41
(0.22, 0.54) 0.16
(0.11, 0.18) After processing 0.71
(0.68, 0.75) 0.57
(0.51, 0.63) 0.67
(0.64, 0.74)

The data in the table show an increase in the correlation of vibrations in the studied ranges after processing from ~ 0.2 to ~ 0.7, which for physiological systems indicates a high degree of connection between the two processes. Thus, it has been demonstrated that fluctuations in blood flow restored using the proposed method for visualizing fluctuations in blood flow in the limb region do reflect changes recorded by the blood flow sensor. Using the algorithm used to convert temperature fluctuations into blood flow values at each point of the thermogram and matching each found value with a certain color allows you to visualize blood flow fluctuations, as shown in FIG. 5. Compared with photoplethysmographic data, the method of visualizing the distribution of blood flow fluctuations makes it possible to control hemodynamics simultaneously in several zones of the limb.

Claims (2)

1. A method for visualizing fluctuations in skin blood flow in the extremities, implemented by recording fluctuations in skin temperature in the area of the hand and subsequent mathematical wavelet analysis of temperature fluctuations, during which the dependence of the wavelet coefficients of temperature fluctuations on time is constructed, characterized in that temperature fluctuations are recorded thermal imaging method, calculating the dependence of the wavelet coefficients of temperature fluctuations on time, which are the spectral components of the count fucking temperature is carried out at each point of the thermal imaging image of the limb, then each spectral component of the frequency f is shifted to the previous time by the interval Δt defined by the formula $$\Delta t_{\text{PHASE}}(f_i,z)=z/\left(2\cdot \sqrt{\frac{\pi \cdot \lambda }{c\cdot \rho }\cdot f_i}\right)$$

where z is the thickness of the biological tissue layer, λ is the skin thermal conductivity coefficient, s is the specific heat of the skin, ρ is the skin density, f _i is the frequency of the i-th spectral component, the amplitude of each spectral component is multiplied by a coefficient determined by the formula $$C_{\text{Amp}}(f_i,z)=\exp \left(z\cdot \sqrt{\frac{\pi \cdot c\cdot \rho }{\lambda }\cdot f_i}\right)$$

, then perform the inverse wavelet transform of the spectral components at each point of the thermogram and get the resulting data array, which is the distribution of fluctuations of the skin blood flow. 2. The method according to claim 1, characterized in that each calculated value of the amplitude of the fluctuations in blood flow is associated with its color to visualize the fluctuations of the skin blood flow in the limbs.

Images