WO2025233902A1

WO2025233902A1 - Method and device for controlling the spatial position of a surgical needle

Info

Publication number: WO2025233902A1
Application number: PCT/IB2025/054893
Authority: WO
Inventors: Federica FERRAGUTI; Cristian Secchi; Stefano Puliatti; Giampaolo Bianchi
Original assignee: Universita Degli Studi di Modena e Reggio Emilia
Current assignee: Universita Degli Studi di Modena e Reggio Emilia
Priority date: 2024-05-10
Filing date: 2025-05-09
Publication date: 2025-11-13
Anticipated expiration: 2026-11-10

Abstract

Method for controlling the spatial position of a surgical needle, comprising the steps of receiving a first three-dimensional model representative of an anatomical portion including a target point internally within a patient's body and anatomical parts in a surrounding area encompassing at least organs and blood vessels; receiving in the first three-dimensional model at least one straight-line target trajectory for the surgical needle extended at least from a representative point on the patient's skin to the target point; receiving a second three-dimensional model of the anatomical portion including the target point and anatomical parts in a surrounding area, wherein this second model is generated from an ultrasound scan of the patient and is referenced in the coordinate system of a collaborative handling device of the surgical needle; anchoring via a registration algorithm the first three-dimensional model comprising the at least one target trajectory to the second three-dimensional model so that the anatomical parts in the first model are positionally referenced to the coordinate system of the second model and are overlaid aligned with the corresponding anatomical parts in that second model; providing to a controller of the surgical needle handling device a command representing the angular position of the surgical needle based on the target trajectory.

Description

METHOD AND DEVICE FOR CONTROLLING THE SPATIAL POSITION OF A SURGICAL NEEDLE

DESCRIPTION

TECHNICAL FIELD

The present invention pertains to the field of methods for controlling the spatial position of a surgical needle before it is inserted into the patient's skin. The present invention also relates to a collaborative handling device capable of implementing the control method according to the present invention.

STATE OF THE ART

In the field of surgical techniques, such as those aimed at removing stones that form inside a patient's kidney, techniques that involve inserting a surgical needle into the patient's skin to reach the renal cavities where the stones to be removed are located are widely used. This surgical needle is hollow so that a guide wire can be inserted inside it to dilate the pathway and allow the insertion of surgical instruments suitable for fragmenting and extracting the stone. However, the techniques currently adopted, such as percutaneous nephrolithotomy (PCNL), require a great deal of experience on the part of the operator to ensure that when the surgical needle is inserted, adjacent vascular structures are not damaged, thus avoiding possible pre- and post-operative bleeding. Additionally, during the insertion of the surgical needle, it is crucial to avoid contact with anatomical structures or organs such as the pleura, colon, liver, and spleen.

Therefore, there is a constant need to develop solutions capable of minimizing the risks of the surgical procedure for the patient, supporting the surgeon during the intervention without thereby making the final outcome dependent on the surgeon's experience. Another requirement is to reduce the learning curve for residents or new surgeons in correctly and precisely learning and applying the methodology.

OBJECT AND SUMMARY OF THE INVENTION

The present invention aims to meet at least part of the aforementioned needs, and this objective is achieved through a method for controlling the spatial position of a surgical needle according to claim 1.

According to a preferred embodiment of the present invention, a method for controlling the spatial position of a surgical needle is presented, which allows positioning the surgical needle before performing the surgical procedure in an angular position that can reach a predefined target point, e.g., a renal calyx where a stone to be removed is located, minimizing the risk of contact and damage to the organs and blood vessels surrounding the target point. Specifically, through the use of a surgical needle handling device, it is possible to move the surgical needle and lock it in a predefined position, allowing the surgeon to push the surgical needle axially without manually manipulating it, thereby reducing the risk of complications during the surgical procedure related to the surgeon's professional experience. To achieve this result, this method can be implemented by an electronic control unit of a computer in the hospital's operating room, and includes the step of receiving a first three-dimensional model representing an anatomical portion comprising a target point inside the patient's body, e.g., a point where a renal stone is located, and anatomical parts in a zone surrounding that target point. In particular, these anatomical parts include at least organs and blood vessels. For example, this model can be created from a computed tomography (CT) scan performed on the patient days before the surgical procedure. Additionally, the method includes the further step of receiving in the first three-dimensional model at least one target straight-line trajectory for the surgical needle, in which, for example, this trajectory passes through the organ containing the stones, e.g., through a renal calyx, to reach the stones themselves. According to one aspect of the present invention, this target trajectory extends from a point representing the patient's skin to the target point, preferably in a way that avoids interference with the organs and blood vessels between the skin and the target point, e.g., to avoid bleeding. This target trajectory can thus be followed by the surgeon in use, specifically by pushing the surgical needle axially into the patient's body without contacting organs and blood vessels, thereby minimizing the risk of dangerous bleeding for the patient. According to another aspect of the present invention, the method includes the further step of receiving a second three-dimensional model of the patient's anatomical portion comprising the target point and anatomical parts in a zone surrounding that target point, in which this second model is generated from an ultrasound scan of the patient and is referenced in the coordinate system of a collaborative surgical needle handling device. For example, this model is generated on the day of the surgical procedure, in which the patient is laid on the operating table after receiving anaesthesia, and through an ultrasound probe connected, for instance, to a robotic arm, it is possible to acquire a scan of the area of interest, which in turn is processed to generate the three-dimensional ultrasound model. It should also be noted that the first and second three-dimensional models are separate at the time of the ultrasound scan, so the method includes the further step of anchoring, through a registration algorithm, the first three- dimensional model comprising at least one target trajectory to the second three-dimensional model so that the anatomical parts in the first model are aligned and overlapped with the corresponding anatomical parts in the second model. In this way, the target trajectories are referenced in the same coordinate system of the surgical needle handling device to which the patient's position is in turn referenced. The method includes the further step of providing a command to a controller of the surgical needle handling device, representing the angular position of the surgical needle based on the target trajectory, so that when the handling device reaches this position and the surgical needle is carried by it, the surgeon can then push the surgical needle into the patient's body percutaneously to reach the target point.

DESCRIPTION OF THE DRAWINGS

The constructional and functional features of the method for controlling the spatial position of a surgical needle can be better understood from the detailed description that follows, in which reference is made to the attached figures representing a preferred and non-limiting embodiment, where:

- Fig.l shows a diagram comprising the steps to create the first three-dimensional model with classified anatomical parts according to a preferred embodiment of the present invention;

- Fig.2 shows an additional three-dimensional model representing an organ containing stones according to the preferred embodiment of Fig.l;

- Fig.3 shows a view of the three-dimensional model of Fig.2, including regions that identify points through which a respective suggested straight-line trajectory for the surgical needle passes;

- Fig.4 shows a view of the three-dimensional model of Fig.2, including a plurality of suggested straight-line trajectories for the surgical needle;

- Fig.5 shows a view of an additional three-dimensional model representing an abdominal area, including the organ with stones, a suggested trajectory, and a target trajectory, spaced apart based on an interference condition calculated between the suggested trajectory and the anatomical parts surrounding the organ with stones;

- Fig.6 shows a schematic view of a surgical needle handling device on which a support for a surgical needle and an ultrasound probe is fixed.

DETAILED DESCRIPTION OF THE INVENTION

According to a preferred embodiment of the present invention, a method for controlling the spatial position of a surgical needle is presented for performing surgical procedures, such as the removal of stones in a patient's organ, and in particular for positioning the surgical needle at a point on the patient's skin so that, when the desired position of the surgical needle is locked and the surgical needle is inserted into the skin to reach the cavity where the stone is located, the risk of the surgical needle contacting bones or other organs that could be damaged is minimized. Preferably, this control method is particularly suitable for procedures like percutaneous nephrolithotomy (PCNL), but it can also be used for the removal of stones in other organs, e.g., gallstones, for urinary drainage in case of obstruction of the renal excretory pathway, or for the treatment of small renal neoplasms, e.g., cryotherapy procedures. Specifically, to implement this control method, some preliminary steps are necessary. In particular, this control method involves the preliminary step of acquiring images, preferably tomographic, of the patient's anatomical portion, including the organ in which the stone is located, e.g., the patient is placed in a prone or supine position on a table that moves horizontally within a ring structure containing the X-ray tube emitting X-rays. Alternatively, these images can also be acquired through magnetic resonance. Once acquired, the tomographic images are further processed through a trained imaging classification algorithm to classify the scanned anatomical parts, and more specifically to classify the anatomical parts of the organ in which the stones are present and automatically identify the stone itself. This trained algorithm performs segmentation of the acquired tomographic scan and classifies the segmented image portions. Since these scans are three-dimensional, a 3D U-net type imaging classification algorithm (deep learning) can be used, implemented by an electronic control unit preferably connected for data exchange with the computed tomography (CT) device to receive the acquired tomographic scans and process them through the imaging classification algorithm to provide a classified image output. For example, this control unit can be on board a computer in the radiology lab of a hospital, with a screen connected to the computer to show the processed tomographic images to medical staff using the trained imaging classification algorithm. Preferably, in the case of acquiring tomographic scans of the patient's kidneys in which stones are present, through such algorithm it is possible to segment and classify the scanned anatomical parts by processing acquired tomographic images of the type without contrast medium, also called ‘basal’, or with contrast medium, including those called ‘delayed’.

For example, a first trained imaging classification algorithm of the 3D U-net type (deep learning) is used to process ‘basal’ tomographic images to segment the parenchymal edges of the kidneys and renal stones, e.g., by applying a threshold value of the Hounsfield scale to the tomographic image, but also the bones and skin of the patient, e.g., by applying another threshold value of the Hounsfield scale. Furthermore, a second trained imaging classification algorithm of the 3D U-net type (deep learning) can be used to estimate a classification of the anatomical parts of interest on the ‘delayed’ tomographic scans, such as the parenchyma, ureter, and renal calyces. Training of the first and second classification algorithms includes the step of providing a set of ‘basal’ and ‘delayed’ computed tomographic scans, respectively, to each algorithm, followed by the manual classification of each input scan. Specifically, during training, each ‘basal’ type computed tomographic scan is visually analysed by an operator, who manually classifies the scan after identifying and delineating the regions corresponding to the parenchymal edges, renal stones, bones, and skin. Similarly, each ‘delayed’ type computed tomographic scan is visually analysed by the operator, who manually classifies each scan after identifying and delineating the regions corresponding to the calyces, ureter, and parenchyma.

For example, to achieve the result shown in a preferred embodiment in Fig.l, 125 ‘basal’ tomographic scans were input to the first algorithm, with 100 used for the training phase by the operator and 25 used in the test phase for validation of classification accuracy after training. The second algorithm received 60 ‘delayed’ tomographic scans as input, with 50 used for the training phase by the operator and 10 used in the test phase for validation of classification accuracy after training. It should also be noted that to accurately visualize the stones located inside the calyces, a ‘basal’ type computed tomographic scan must be performed. This scan, executed without the use of contrast medium, allows better visualization of the stones, which appear in the images with a different density than the surrounding renal tissue. The ‘delayed’ type computed tomographic scan, on the other hand, is performed following an injection of a liquid contrast medium and is useful for evaluating the contrast opacification of the renal excretory pathway. Therefore, at the end of this classification phase, a first and a second three-dimensional image IM1 and IM2 are obtained, in which different anatomical parts stand out visually, therefore to know the relative position of all classified anatomical parts, it is necessary to overlay the classified “basal” CT scan with the classified “delayed” CT scan so that the borders of an organ classified in the first image IM1 are aligned with the borders of the corresponding organ classified in the second image IM2, e.g. based on minimizing a parameter representative of the orientation error between the first and second images. The objective of this phase is therefore that of orienting one of the first and second images once superimposed on the other between the first and second images in order to guarantee that the position of the calculations previously detected kidneys is correctly defined inside the calyces, ensuring a precise and accurate representation of the classified anatomical parts in the first and second images. This phase is also known as 'registration' among experts in the field and to achieve this result, the classified tomographic images are processed using a registration algorithm, also known in the field, such as SlicerMorph, a toolbox of the 3D Slicer software. For example, the anatomical parts classified in the first image IM1 are referenced positionally relative to a first reference system, while the anatomical parts classified in the second image IM2 are referenced positionally relative to a second reference system. Using the registration algorithm, one of the first and second images can be positionally referenced within the reference system of the other, e.g., using known rotation matrices, to minimize a parameter representing the orientation error between the first and second images when oriented relative to a common reference system. After the classification of the “basal” and “delayed” CT images, a three-dimensional volume reconstruction algorithm needs to be applied to the classified images. For example, to achieve the 'registration' of the first image relative to the second image, it is possible to input into the registration algorithm a first classified image comprising the reconstructed volumes from the 'delayed' CT scan including an anatomical part of interest, e.g., the parenchyma with the calyces and ureter, and a second classified image comprising the reconstructed volumes from the 'basal' CT scan including the same anatomical part of interest, i.e., the parenchyma. Specifically, based on this registration algorithm, one of the first and second images is overlaid on the other so that the volume of the anatomical part of interest from the 'delayed' CT scan is oriented and overlaid with the volume from the 'basal' CT scan, e.g., the parenchyma. Preferably, in addition to the 'delayed' CT scan, an 'arterial' tomographic scan and a 'venous' tomographic scan can be acquired when the contrast medium is injected, in a known manner, e.g., acquiring a scan at different time phases after the contrast medium injection, so that the arteries and veins can be identified and classified respectively in the same way indicated above. Once this phase is completed, it is possible to estimate the rotation matrix to orient and overlay the other anatomical parts of interest in order to provide an output of a complete first three-dimensional model Ml, particularly comprising the parenchyma, calyces, arteries, veins, ureter, and renal stones, for both the right and left kidney (Fig.l). Preferably, it is also possible to obtain the first three- dimensional model Ml comprising the classified anatomical parts using other imaging techniques, e.g., magnetic resonance. According to another aspect of the present invention, once the first three- dimensional model Ml is created, preferably from the computerized tomographic scan, it is possible to exploit this first three-dimensional model to identify a point on the skin of the patient on which to rest and insert the operating surgical needle in order to reach the anatomical part in which the stone is located. Additionally, this model can be used to calculate a target trajectory for the surgical needle to reach a target point, such as the fornix of the calyx, to remove the stone without contacting the surrounding anatomical parts and risking damage. Therefore, it is crucial not only to identify the surgical needle insertion point on the skin but also to determine its angular orientation to achieve the desired result. To accomplish this result, a first algorithm can be used to calculate a straight-line trajectory free from interference with organs and blood vessels, preferably to avoid haemorrhages, said algorithm receives the first three-dimensional model Ml as input and processes it to calculate a target insertion trajectory for the surgical needle. Specifically, the algorithm involves receiving the previously obtained three-dimensional model with classified anatomical parts and then removing volumes representing non-interest anatomical parts, such as bones and other organs, from this model. Thus, once the volumes of non-interest parts are removed, the first three-dimensional model retains volumes representing the anatomical parts of interest, potentially including the stone, such as the renal calyces. Subsequently, the algorithm involves discretizing the volumes of the anatomical parts of interest to define a point cloud representing the perimeter surface of these anatomical parts, resulting in a three-dimensional model C representing the organ with stones (Fig. 2). Preferably, this discretized surface represents at least one cavity of the organ with stones, such as a renal calyx. Additionally, for each point on this surface, a normal N is calculated, oriented away from the anatomical part under examination, i.e., externally to the cavity. According to an aspect of the present invention, for each point, a plurality of straight-line trajectories, such as lines, is generated passing through this point and inclined relative to the normal up to a predefined first maximum inclination, e.g., 30 degrees. Thus, the lines passing through each point enter the discretized volume of the anatomical part of interest until intersecting a point longitudinally opposite to the entry point of each line. In this way, the lines passing through each point enter the discretized volume of the anatomical part of interest until it intersects a point longitudinally opposite to the one through which each straight line entered. In this way, it is possible to calculate the intercept I on the opposite side to that on which the normal to the point crossed by each line is located. Therefore, a segment is defined between the point through which a line passes and the calculated intercept, and the line passing through this segment thus defines a straight trajectory T for the surgical needle within the anatomical part of interest, i.e., in the cavity. Consequently, for each point on the examined surface, a set of segments is obtained, each defined between the common entry point and its respective intercept on the opposite side of the point with the normal. Moreover, such algorithm includes the step of comparing the lengths of each segment for each point under examination and subsequently the step of selecting the segments longer than a predefined length. Specifically, the selected segments, particularly the lines passing through them, represent suggested straight-line trajectories for the surgical needle during the surgical operation. Each of these segments represents a straight path the surgical needle can follow within the anatomical part of interest, such as the renal calyces, which is particularly relevant from an operational perspective as it allows the surgical needle to reach the target point. It is important to note that in surgical procedures for the removal of renal stones, it is sufficient for the surgical needle to reach the fornix of the calyx by following a suggested straight-line trajectory without needing to be pushed deep into the calyx to reach the stone. In particular, the surgical needle is oriented toward the stone in order to reach it by following the suggested straight trajectory and the latter is further utilized to reach and fragment the stone into smaller pieces, for example, using a laser. Preferably, for each calculated segment a further plurality can be generated of straight lines inclined with respect to the respective segment up to a second maximum inclination, less than the first maximum inclination, in order to identify whether among these straight lines there is one which defines a further segment having a length between the entry point of the line and the respective intercept greater than the one previously calculated, i.e. the length of the segment. This way, a plurality of segments can be calculated for each point on the discretized three-dimensional model C. However, since the anatomical part of interest, such as the renal calyces, is irregularly shaped, the difference in length between segments can be significant, meaning that only a portion of them is relevant for the insertion of the surgical needle. Specifically, segments longer than a predefined length can be considered for each point, ensuring that the lines passing through these selected segments represent suggested straight trajectories for the surgical needle, for example, using a clustering algorithm. This approach allows for the identification of one or more regions Zl, Z2, also called 'clusters,' in which points on the discretized surface are located, each intersected by a suggested trajectory, i.e., a line passing through a segment whose length within the renal calyx, for example, exceeds the predefined threshold length (Fig. 3). Furthermore, for each obtained region or 'cluster,' the segment with the maximum length within the anatomical part of interest is subsequently calculated among those within the examined region such that for each examined region, a segment with the maximum length TS1, TS2 (and the corresponding point it passes through) is identified for the insertion of the surgical needle (Fig. 4). Preferably, given common surgical practice, the region or regions located on the lower posterior calyces can be defined as preferred for insertion, however the surgeon has the option in use to modify this predefined selection and choose one of the other calculated trajectories, each in a region different from the lower posterior calyces. Once the suggested straight trajectories are calculated using an algorithm that meets these predefined geometric and positional conditions, it is also possible to determine if there is any interference between each suggested straight trajectory and the surrounding anatomical parts of the area of interest in which the stone is located. To do this, the first three-dimensional model Ml is restored to its initial conditions before the trajectory calculation phase, that is comprising the volumes of both the anatomical part of interest and the surrounding classified parts, with the suggested trajectories defined on this model. In this way, by using an additional algorithm to calculate a straight trajectory free from interference with organs and blood vessels to avoid haemorrhages, it is possible to determine if the suggested trajectory presents any interference along the path between the anatomical part of interest and the patient's skin, in order to ensure that, in use, the surgical needle does not contact or damage another organ. Specifically, a target trajectory TT1 for the surgical needle can be calculated, representing the path the surgical needle must follow during the surgical procedure (Fig. 5). In particular, when no interference is detected, the suggested trajectory TS1 corresponds to the target trajectory TT1 for the surgical needle. On the other hand, when interference is detected, the method according to the present invention includes an additional step to calculate a target trajectory TT1 angularly displaced from the interfering anatomical part until no interference is present (Fig. 5), for example, using an algorithm that minimizes the angular distance between the initially proposed orientation and the new orientation. It should be noted that the target trajectory also passes through the point intersected by the suggested trajectory and is calculated to be displaced from the anatomical part with which the suggested trajectory has interference. It's also important to note that the target trajectory TT1 can be angularly displaced from the suggested trajectory TS1 up to a predefined threshold, beyond this threshold, the resulting target trajectory would be excessively inclined to the extent that the straight trajectory would not reach the stone to be removed, as intended by following the suggested trajectory TS1. On the other hand, by defining a threshold, the surgical needle in use can follow the calculated target trajectory TT1 without contacting the anatomical part that would be contacted if the suggested trajectory were followed, while still reaching the target point in the renal calyx.

Preferably, if interference is detected and multiple suggested trajectories are calculated, it is possible to select the one that does not interfere with the surrounding anatomical parts of the organ with stones. If the calculated target trajectory TT1, angularly displaced from the suggested trajectory TS1, still interferes with an anatomical part, it is possible to select another point from the available ones in the region or 'cluster,' corresponding to a respective suggested straight trajectory to follow, e.g., a suggested straight trajectory TS2. This way, the spatial angular position of the target trajectory TT1 can be stored along with the anatomical part of interest, e.g., the calyces, and the surrounding parts, so that the resulting three-dimensional model defines the straight trajectories that the surgical needle can follow in use without contacting surrounding organs and reaching the target point in the renal calyx. For example, the target trajectory TT1 is a line viewable on a screen displaying the first three-dimensional model Ml, which extends from the calyx to the patient's skin in order to identify the insertion point of the surgical needle on the skin. Consequently, for each previously calculated target trajectory, a corresponding point on the skin is defined in the three-dimensional model to provide an output of multiple points with the respective straight insertion trajectory of the surgical needle, which the surgeon can select based on their preferences. It should also be noted that the position of each point and the angular orientation of the corresponding target trajectory are known, as they are referenced to the coordinate system of the first three-dimensional model Ml generated by the tomographic scan. According to another aspect of the present invention, this control method involves positioning the surgical needle at a point on the patient's skin to be angularly positioned to follow the previously calculated target trajectory. In this way, when the surgical needle is in this spatial position, it can be locked angularly and pushed, following this trajectory, into the patient's body to reach the anatomical part of interest where the stone to be removed is located. To perform this phase, it is possible to use a handling device on which the operating surgical needle can be fixed. For example, as shown in a schematic view in Fig. 6, such a handling device can be a robotic arm BR configured to extend in use towards the portion of the patient's body that is to undergo the surgical operation so as to be able to position the surgical needle in the desired trajectory. To perform such phase the three-dimensional model comprising the previously calculated target trajectories is used. For example, the robotic arm can be rigidly fixed to the operating table in the surgical room in which the patient undergoes the surgical operation. To position the surgical needle in such a way that it can be inserted at a point on the skin and with an inclination based on the target trajectory TT1, the position of the patient's body must be referenced to a known coordinate system once the patient is placed on the operating table. Tracking the position of the patient's body also involves tracking the position of the anatomical parts in which the stone to be removed is located, as these may have a different position within the patient's body compared to their position when the tomographic scan was performed, i.e., the generation of the first three-dimensional model Ml. Therefore, at the end of the robotic arm longitudinally opposite to that fixed to the operating table or similar support, a holder for a surgical needle SA is fixed. Such support has a first end rigidly fixed to the robotic arm and a second end, longitudinally opposite to the first, on which it is rigidly fixed in use, preferably in a removable manner, an ultrasound probe SC which is used to scan the portion of the patient's body with the anatomical parts of interest in which the stone to be removed is located. Based on this constructional configuration, the ultrasound probe is positionally referenced to the coordinate system of the robotic arm. Furthermore, there are SM sensors on board the operating surgical needle support, e.g. inertial platforms, each of which is preferably arranged on board radial elements extending transversely from the surgical needle holder, so as to be able to detect a change in the position of the surgical needle holder with respect to the reference system of the robotic arm. These sensors are also connected for data exchange to a control unit connected for data exchange to the robotic arm and the ultrasound probe, enabling the control unit to process received signals to calculate the position of the surgical needle holder during use. Given that the ultrasound probe SC is rigidly fixed to the holder, the position of the ultrasound probe can be easily determined considering the dimensions of the holder and the spatial position of the ultrasound probe relative to these sensors.

According to another aspect of the present invention, the surgical needle holder SA includes, on the end where the ultrasound probe is fixed during use, a protruding portion PA transversely to the extension direction of the surgical needle holder. In particular, this protruding portion has a plurality of holes Fl, preferably three, arranged on the side opposite to the patient's body during use. These holes are inclined differently from each other relative to a longitudinal axis of the surgical needle holder, preferably ranging between 20 and 45 degrees, and are configured to converge towards a common hole F2 positioned facing the patient's body during use. Each hole on the protruding portion is sized to accommodate the surgical needle during use, which exits through the common hole F2. Advantageously, by keeping the surgical needle holder SA with the protruding portion fixed in a predefined position, the surgical needle can be angularly locked once inserted into one of the holes. In particular, the robotic arm BR is operated so that the surgical needle holder reaches a position where the axis of the hole engaged by the surgical needle is substantially aligned with the target trajectory TT1, thus, for the surgeon, it is sufficient to axially push the surgical needle into the patient's body to reach the anatomical part of interest where the stone to be removed is located. Since the position of the ultrasound probe SC in the robotic arm BR's coordinate system is referenced, this aspect can be utilized to track the position of the patient's body part on the operating table, particularly by referencing the position of the body and its relevant anatomical parts of interest to the robotic arm's coordinate system. To achieve this result, while the patient is on the operating table, the surgeon manually holds the ultrasound probe SC and performs a scan by placing the probe on the body part of interest, such as the kidney area. Specifically, since the spatial position of the ultrasound probe SC is referenced to the robotic arm's coordinate system, ultrasound scans acquired can be processed using an image registration algorithm, e.g., ImFusion, to generate a second three-dimensional ultrasound model M2 representing the scanned anatomical parts and referenced to the robotic arm's coordinate system, thus the position of the patient, particularly the position of the anatomical parts of interest are known. In particular, with the position of the ultrasound probe SC known relative to the robotic arm's coordinate system and known the position of the scanned anatomical parts relative to the probe's coordinate system, the position of these anatomical parts relative to the robotic arm's coordinate system can be calculated, e.g., using known rotation matrices. Therefore, the first three- dimensional model Ml, generated from the computerized tomographic scan, and the second three- dimensional model M2, generated from the ultrasound scan, are processed using the image registration algorithm so that the anatomical parts identified in the first three-dimensional model are superimposed in an oriented manner with their corresponding anatomical parts in the second three-dimensional model, i.e. the anatomical parts identified in the first three-dimensional model are positionally referenced to the coordinate system of the second three-dimensional model. As discussed earlier, for example, the anatomical parts identified in the first three-dimensional model are positionally referenced to a first coordinate system, while the anatomical parts identified in the second three-dimensional model are positionally referenced to a second coordinate system. In this way, by exploiting the registration algorithm it is possible to positionally refer the first three- dimensional model in the reference system of the first three-dimensional model so that the edges of the corresponding anatomical parts in the two models are aligned. For instance, this alignment can be achieved using known rotation matrices to minimize a parameter representing the orientation error between the first and second three-dimensional models relative to a common coordinate system, specifically that of the second three-dimensional model corresponding to the robotic arm. In this manner, it is possible to define on the ultrasound three-dimensional model the previously calculated surgical needle insertion target trajectories from the first three-dimensional model, i.e., the one created from the tomographic scan. According to a further aspect of the present invention, through this registration imaging algorithm it is possible to anchor the first three- dimensional model Ml, resulting from the tomographic scan, to the second three-dimensional model M2 resulting from the ultrasound scan, so that at the end of the registration phase also the spatial position of each of the target trajectories for the surgical needle is referenced with respect to the reference system of robotic arm. It should be noted that in this type of surgical procedure, the patient undergoes anaesthesia, preferably general anaesthesia, so once positioned on the operating table, the patient remains essentially immobile, thus, after performing the ultrasound scan, the position of the patient's body remains tracked throughout the course of the stone removal procedure. In this way, a third three-dimensional model M3 is generated, which is provided as output to the surgeon, for example, displayed on a screen connected to the computer where the electronic control unit processes the two three-dimensional models using the previously described image registration algorithm. The resulting three-dimensional model M3 then shows the target trajectories of the surgical needle which can be selected by the surgeon according to their preferences, e.g., on the three-dimensional model shown on the screen lines extending from each point on the skin to the renal calyx where the stone is located can be shown. When one of the surgical needle trajectories is selected by the surgeon, the electronic control unit receives an input signal representing the selected insertion target trajectory and processes this signal to calculate a spatial position that the surgical needle holder must reach to position the needle in the desired target trajectory, e.g., the target trajectory TT1 shown in Fig. 6. Therefore, the electronic control unit is programmed to generate a robotic arm actuation signal to move the surgical needle holder support to the representative position of the target trajectory. Specifically, when the surgical needle holder support SS reaches this position, the axis of one of the holes in the protruding portion is substantially aligned with the selected target trajectory TT1 previously chosen by the surgeon. The position of the surgical needle holder is thus angularly locked so that the surgeon can insert the surgical needle into the corresponding hole, which can then be manually pushed into the patient's body until the surgical needle tip contacts the patient's skin and reaches the target point percutaneously.

As described earlier, in the case of kidney stone removal, the needle is pushed to reach the renal calyx while avoiding collision with other organs and bones and simultaneously, the ultrasound probe SC must remain functional and in contact with the patient's body so that the surgeon can view the anatomical parts of interest during the surgical procedure through the ultrasound image, while the surgical needle advances into the patient's body. Once the surgical needle holder support SS with the protruding portion SA is positioned to place the surgical needle on the desired target trajectory TT1, the surgeon inserts the needle into the corresponding hole and pushes it until the needle tip contacts the patient's skin. At this point, the needle can be further pushed percutaneously to reach the renal calyx where the stones are located. To track the position of the advancing surgical needle inside the patient's body, a plurality of sensors, e.g., inertial platforms, can be placed on board such a needle, through which, the position of the surgical needle tip can be detected during use, allowing the tracking of when the surgical needle tip reaches the stone and thus stopping the insertion into the patient's body. For example, to achieve this result, the spatial position of each sensor relative to the coordinate system of the robotic arm can be referenced, therefore, knowing the geometry of the surgical needle and the distance of the sensors from the needle tip, the spatial position of the needle tip relative to the coordinate system of the robotic arm can be referenced and tracked during the surgical procedure.

Claims

1. Method for controlling the spatial position of a surgical needle, comprising the steps of:

- receiving a first three-dimensional model (Ml) representing an anatomical portion comprising a target point internally within a patient's body and anatomical parts in a surrounding area encompassing at least organs and blood vessels,

- receiving in the first three-dimensional model (Ml) at least one straight-line target trajectory (TT1) for the surgical needle extended at least from a representative point on the patient's skin to the target point,

- receiving a second three-dimensional model (M2) of the anatomical portion comprising the target point and anatomical parts in a surrounding area, wherein this second model is generated from an ultrasound scan of the patient and is referenced in the coordinate system of a collaborative handling device (BR) of the surgical needle,

- anchoring, via a registration algorithm, the first three-dimensional model (Ml) comprising the at least one target trajectory (TT1) to the second three-dimensional model (M2) so that the anatomical parts in the first model are positionally referenced to the coordinate system of the second model and are overlaid aligned with the corresponding anatomical parts in that second model,

- providing to a controller of the surgical needle handling device (BR) a command representing the angular position of the surgical needle based on the target trajectory (TT1).

2. Control method according to claim 1, wherein the relative position between the anatomical portion of the patient comprising the target point and the handling device (BR) of the surgical needle remains fixed between the step of receiving the second three-dimensional model (M2) and the anchoring step.

3. Control method according to claim 1 or 2, wherein the straight-line target trajectory (TT1) is free of interference with organs and blood vessels between the skin and the target point.

4. Control method according to claim 3, further comprising the step of receiving in the first three-dimensional model (Ml) a suggested straight-line trajectory (TS1) and applying to this suggested trajectory an algorithm for identifying a straight-line trajectory free of interference with organs and blood vessels, so as to calculate a condition of interference between this suggested trajectory and the organs and blood vessels in the surrounding area of the target point, and wherein the suggested straight-line trajectory (TS1) is the target trajectory (TT1) when no interference is found.

5. Control method according to claim 4, further comprising the step of, when an additional suggested trajectory (TS2) is received in the first three-dimensional model (Ml) and interference of this additional suggested trajectory with the organs and blood vessels in the surrounding area of the target point is found through the calculation algorithm of a straight-line trajectory free of interference, selecting between the two trajectories the one without interference.

6. Control method according to claim 4 or 5, wherein, to calculate a condition of interference between the suggested trajectory (TS1, TS2) and the organs and blood vessels surrounding the target point, such method comprises the further step of:

- receiving a third three-dimensional model (C) of an organ comprising the target point contained in the first three-dimensional model (Ml),

- generating the at least one suggested trajectory (TS1) through an algorithm for calculating straight- line trajectories, said algorithm satisfying a predefined geometric condition and a predefined position condition to generate the at least one suggested trajectory.

7. Control method according to claim 6, wherein the third three-dimensional model (C) comprises a three-dimensional grid of points representing a surface of at least one cavity of the organ comprising the target point, a normal (N) for each point of the three-dimensional grid, wherein such normal extends outwardly from the at least one cavity, and a plurality of lines passing through each point of the three-dimensional grid, wherein such lines are inclined relative to the normal so that each has an angle equal to or less than a predefined angle relative to the normal of the point they pass through, and wherein, to satisfy the geometric and position condition to generate the at least one suggested trajectory (TS1), such method further comprises the step of:

- for each line passing through a point, calculating the intercept (I) on the surface of the at least one cavity from the opposite side to where the normal (N) of that point is located,

- calculating, for each line passing through such a point, the length of the segment defined by each line from that point to the corresponding calculated intercept,

- selecting for each point the segments having a length greater than a predefined length, so that the lines passing through these selected segments are representative of the suggested straight-line trajectories,

- selecting for each point from all selected segments the one with the maximum length.

8. Control method according to claim 7, wherein, when a suggested straight-line trajectory (TS1) is calculated that meets the geometric condition and the predefined position condition, calculating a condition of interference with the surrounding anatomical parts of the organ comprising the target point through the algorithm for calculating a straight-line trajectory free of interference with the organs and blood vessels, and when interference is found, adjusting the angular orientation of the suggested trajectory until no interference is present.

9. Control method according to any of the preceding claims, wherein the organ comprising the target point is a kidney, the target point is a stone, the surface of the at least one cavity is representative of a renal calyx, and the surrounding area of the organ comprising the target point is an abdominal area.

10. Collaborative handling device (BR) of a surgical needle comprising: - a robotic arm positionally referenced to a first coordinate system,

- a support for a surgical needle (SA) rigidly attached to a movable end of the robotic arm,

- an ultrasound probe (SC) rigidly attached to the support for a surgical needle (SA) so as to be positionally referenced to the first coordinate system,

- a controller of the robotic arm connected in data exchange to an electronic control unit programmed to execute the control method according to any one of claims 1 to 8.