Medical assistive robot, surgical system and method for operating a medical assistive robot and / or a surgical system

Abstract

The present invention relates to a medical assistance robot (10) for predictive instrument provision for a surgeon (12), comprising at least two positioning units (14), each configured to receive and move at least one medical instrument (16), a procedure step recognition unit (18) configured to determine at least one procedure step parameter that characterizes a current step of several steps in which the surgeon (12) is currently performing a procedure on a patient (20), a forecasting unit (22) configured to generate a demand forecast for medical instruments (16) based on the procedure step parameter, and a control unit (24) configured to control at least one of the positioning units (14) based on the demand forecast.to include at least one medical instrument (26) as required and to make it available to the surgeon (12).

Description

[0001] The present invention relates to a medical assistance robot, a surgical system and a method for operating a medical assistance robot and / or a surgical system.

[0002] Robotic systems are becoming increasingly important in surgical procedures. Particularly in minimally invasive surgery (MIS), robotic systems are used primarily for complex procedures. However, such systems regularly require time-consuming setup and dismantling, which consumes valuable operating room time. For this reason, among others, the cost-benefit ratio of such robotic systems is frequently questioned.

[0003] An alternative application for a robotic system in an operating room is that of so-called "scrub nurse robots" (SNRs), which aim to support the surgical team. These are intended to take over the tasks of surgical assistants and, for example, hand medical instruments to the surgeon during a procedure.

[0004] The invention is based on the understanding that previously known SNR systems cannot provide medical instruments with sufficient predictive capability. Consequently, they cannot react in real time or within a reasonable timeframe to the dynamic demands of an operation. However, efficient and safe handover and retrieval of instruments – comparable to the movements of experienced surgical assistants – is crucial for the acceptance of such systems.

[0005] Based on the prior art, the invention aims to achieve an efficient provision of medical instruments to a surgeon.

[0006] The object is solved according to the invention by a medical assistance robot, a surgical system and a method for operating a medical assistance robot and / or a surgical system as described herein and defined in the claims.

[0007] The present invention provides for a medical assistance robot for the predictive provision of instruments to a surgeon. This robot comprises at least two positioning units, each configured to pick up and move at least one medical instrument, and a procedure step recognition unit configured to determine at least one procedure step parameter that characterizes a current step of several steps in which the surgeon is currently performing a procedure on a patient.Furthermore, the medical assistance robot includes a forecasting unit configured to generate a demand forecast for medical instruments based on the procedure step parameter, and a control unit configured to control at least one of the positioning units based on the demand forecast in order to pick up at least one demand-forecasted medical instrument and make it available to the surgeon.

[0008] Furthermore, the present invention provides for the provision of a surgical system comprising a medical assistance robot according to the invention.

[0009] Furthermore, the present invention provides a method for operating a medical assistance robot and / or a surgical system according to the invention.

[0010] The features according to the invention enable the efficient provision of medical instruments to a surgeon. By supporting the surgeon during the surgical procedure, their cognitive and physical strain is reduced, allowing them to focus more intently on the operation. The provision of the medical instruments is handled by the assistive robot. The use of multiple positioning units makes it possible to keep several demand-predicted tools readily available simultaneously. This allows for a response to short-term changes in the need for a medical instrument. For example, the predictive unit could define alternative demand-predicted medical instruments that might be required depending on the course of the procedure. By using multiple positioning units, demand-predicted medical instruments can thus be kept ready for alternative procedure scenarios.

[0011] Furthermore, incorrect provisioning of medical instruments can be avoided. This can compensate, in particular, for inaccurate demand forecasts by the forecasting unit. Generally, this unit can only determine the probability of a future need for a specific medical instrument. If the demand with the highest probability does not occur, another instrument with a lower probability of need can be stored by another positioning unit. Therefore, it is not necessary to rely on the demand with the highest probability occurring. Rather, the probability that the actually required medical instrument was previously stored by one of the positioning units can be increased.

[0012] Another advantage is that the robotic assistant supports the smooth flow of the procedure by simultaneously receiving a used medical instrument from the surgeon and seamlessly handing over a predicted instrument as needed. This optimizes the surgical workflow and the surgeon's interaction with the robotic assistant. The parallel operation of at least two positioning units also allows for the preparation of another medical instrument while one is being provided, thus minimizing downtime. This is particularly important when a rapid sequence of procedure steps is required, as it keeps waiting times for instrument provision to a minimum. It also allows for a flexible and rapid response to emerging situations during the procedure, such as an emergency requiring the stopping of bleeding.

[0013] In general, the assistive robot according to the invention enables intuitive interaction between the surgeon and the assistive robot and optimizes the entire process flow.

[0014] A "medical assistance robot" can be understood as a robotic system in a medical environment. An example of a medical assistance robot is a so-called "scrub nurse robot." This robot is intended to replicate the work of a surgical nurse and support or assist a surgeon during a procedure on a patient. The surgical nurse's tasks include, for example, observing the procedure and proactively selecting medical instruments based on these observations. When a need arises for an instrument, the surgical nurse no longer needs to select it from a storage area containing a supply of medical instruments. Instead, she selects it proactively and has it ready for the surgeon. The present invention is based on the idea of ​​replicating this process using a medical assistance robot.In this sense, a medical assistive robot can be understood as a robotic system that can be used in an operating room, particularly in close proximity to a surgeon performing a procedure. This can mean that the surgeon is within arm's reach of the assistive robot or a section of it, for example, a distal section of one of the positioning units in a specific position, during the procedure.

[0015] The medical robot, and in particular its positioning units, can be designed to meet the hygienic standards of an operating room, including sterile surfaces and a design that allows for easy cleaning. For example, the robot and / or parts of it, such as the positioning units, can be sterilizable. Alternatively, the robot and / or parts of it, such as the positioning units, can be prepared for a procedure using a sterile covering.

[0016] Additionally, a "medical assistance robot" can be specifically designed to provide medical instruments to a surgeon and / or hand them over to him. More generally, a medical assistance robot can be a robot that supports a surgeon during a procedure.

[0017] In this sense, "instrument preparation" can be understood as the selection, picking up, moving, holding, and / or handing over of a medical instrument. Specifically, "instrument preparation" can also include retrieving, returning, and putting away a medical instrument, particularly one that has been used. Generally speaking, it can refer to the entire process from selecting the instrument to handing it over to the surgeon.

[0018] The positioning unit can comprise various mechanical, electrical, and / or electronic components that enable the movement and / or positioning of a medical instrument. For example, the positioning unit allows the medical instrument to be moved and / or positioned within a space. Positioning can also refer to alignment; the positioning unit can, for instance, rotate the instrument.

[0019] In some embodiments, the positioning unit is also equipped with a control and / or sensor unit configured to regulate a movement sequence, acquire position data, and / or coordinate interaction with the positioning unit's environment. Furthermore, as described below, the control and / or sensor unit may be configured to select and, in particular, identify a medical instrument.

[0020] Specifically, the positioning unit can be designed as a robotic arm configured to pick up, move, and deliver medical instruments. The robotic arm can, in particular, be a multi-joint robotic arm with, for example, at least six degrees of freedom. These can include, for instance, three rotational degrees of freedom about a longitudinal axis of a segment of the robotic arm and / or three pivotal degrees of freedom perpendicular to the longitudinal axis of the segment. The degrees of freedom can be defined, for example, by joints. A segment can extend from one joint to the next.

[0021] The positioning unit can include a gripping mechanism and / or a holding device for the medical instrument, enabling a secure and reliable connection of the medical instrument to the positioning unit. The gripping mechanism and / or holding device can allow for mechanical gripping of the medical instrument. Alternatively, a medical instrument could also be magnetically attached or receptive.

[0022] The positioning unit's mobility can be ensured by multiple joints and / or actuators that offer various degrees of freedom, such as rotation and translation. Furthermore, sensors and control components can be integrated into the positioning unit to ensure precise movement sequences and the detection and avoidance of collisions. A positioning unit can therefore be configured to retrieve medical instruments from a tray, move them into the desired position, and, if necessary, hand them to the surgeon. Likewise, the positioning unit can be configured to receive a medical instrument from a surgeon.

[0023] The positioning unit can be compact, especially if multiple units are required. These units must be able to move freely and without collision in close proximity to the surgeon. Therefore, each positioning unit can have a maximum length of 1.0 m, or more specifically 1.5 m, and / or a maximum weight of 50 kg.

[0024] In general, the multiple positioning units can be of identical construction. It is conceivable that the multiple positioning units are largely identical in construction and / or differ, for example, in their gripping mechanism. The corresponding gripping mechanism can be designed for different groups of medical instruments. These groups can differ, for example, in the size and / or weight of the medical instruments.

[0025] A medical instrument can include a tool, device, implant, and / or apparatus specifically designed for use in medical applications. For example, a medical instrument may include surgical instruments such as a scalpel, forceps, tweezers, scissors, drill, needle holder, sensor, clamp, and / or the like. These may be used during surgical procedures. Alternatively, medical instruments may include diagnostic and / or therapeutic tools such as endoscopes, catheters, probes, and / or tissue sampling instruments.

[0026] The procedure can be, for example, a therapeutic, diagnostic, and / or surgical measure. Specifically, the procedure might involve a surgical intervention. It can comprise a complex sequence of actions that can be divided into several different steps to be performed by the surgeon. Using the procedure step parameter, the procedure step recognition unit can provide a quantitative value that allows the surgeon to determine which of the several steps the surgeon is currently in during the procedure being performed on the patient. This provides a comparative value against which the progress of the procedure and / or the sequence of actions can be assessed and / or determined. The term "step" in the procedure can also refer to a phase of the procedure.

[0027] If the procedure involves tumor removal, for example, it can be greatly simplified to include one step making an incision, another step removing the tumor, and a third step suturing the incision. The procedure step recognition unit can be configured to detect which of these steps the surgeon is currently performing.

[0028] Furthermore, the procedure step recognition unit can include a camera and / or camera system configured to generate image data. The procedure step recognition unit can then be configured to determine at least one procedure step parameter based on the image data. This ensures reliable and efficient step recognition. The procedure step recognition unit can include an image recognition unit, which can be based on machine learning and / or a neural network. The camera can be mounted on the surgical robot. Alternatively or additionally, it is advantageous for the surgical system, such as the endoscope or operating light, to include the camera and / or camera system.Using the latter configuration, for example, a patient bed on which the patient lies during the procedure and a waiting area where the surgeon stays during the procedure can be monitored.

[0029] The camera system can include a stereo camera, or the camera itself can be designed as a stereo camera. Accordingly, it can capture location, positions, and / or movements in space.

[0030] In each of the aforementioned steps of the procedure, a different medical instrument may be required, or there may be a need for different instruments. This need can be predicted using the forecasting unit. A need can be stored in the demand forecast. Accordingly, the demand forecast can contain an individual probability of need, particularly derived through the application of artificial intelligence, for various medical instruments, describing the likelihood of a future need. A medical instrument assigned a high probability of need can be described as a "demand-forecasted" medical instrument.In other words, a demand-forecasted medical instrument can be a medical instrument that is needed with a probability of, for example, at least 30%, in particular at least 45%, preferably at least 60%, in a step following, in particular immediately following, the current step.

[0031] The forecasting unit can be configured to update the stored probabilities depending on the procedure step parameter. Thus, the probabilities can be continuously adjusted as the procedure progresses.

[0032] The forecasting unit can, for example, be configured to detect the progress of a procedure based on the current step, which can be described by the procedure step parameter. This can mean that the forecasting unit is configured to predict a future step of the procedure based on the procedure step parameter. Previously determined procedure step parameters can also be considered for this prediction. In other words, the forecasting unit can be configured to perform a prediction, in particular a step prediction and / or a demand prediction, based on several procedure step parameters, each of which characterizes a current and / or a previous step of the procedure. The future step(s) of the procedure can be associated with the need for a specific medical instrument.Accordingly, the forecasting unit can be configured to generate the demand forecast based on the identified future step.

[0033] The forecasting unit can be specifically configured to generate a demand forecast for a particular procedure. This specific procedure can be communicated to the assistance robot, for example, via an interface. Accordingly, the assistance robot can be suitable for supporting various procedures, but a specific procedure can be selected to direct the forecasting unit to that particular procedure.

[0034] If, as in the tumor removal example above, the surgeon is currently performing an incision to open a patient's abdominal cavity, this can be detected by the procedure step recognition unit, and a corresponding procedure step parameter can be generated. Based on this, the prognostic unit can recognize that the incision will be followed by tumor removal. Specific medical instruments are required for this procedure. Accordingly, the prognostic unit can generate a demand forecast and assign a probability of need to each medical instrument. Based on this, the instruments predicted to be needed—that is, those that are most likely to be required—can then be picked up by the positioning units.

[0035] The prediction unit can be implemented on a single processing unit together with the procedure step detection unit. Furthermore, the prediction unit can also include the camera and / or camera system described in connection with the procedure step detection unit. Accordingly, the prediction unit can access the same image data as the procedure step detection unit.

[0036] According to some embodiments, the forecasting unit comprises an AI model, which is based in particular on machine learning and / or a Markov process. In other words, the generation of the demand forecast can be based on a mathematical calculation procedure and / or an algorithm, for example, using artificial intelligence, machine learning, a Markov process, deep learning, and / or a neural network, in particular an open, closed, single-layer, and / or feedback neural network, and / or a combination thereof.

[0037] According to some embodiments, the prediction unit comprises a computer-based AI unit that is active in a learning phase, e.g., for data acquisition to derive or extend the AI ​​model, and / or an execution phase. During the learning phase, the AI ​​unit may have an annotation function configured to identify the instruments used during a variety of operations of a specific type and to record their frequency, duration, and sequence of use. Annotation can be performed both online, during the operation, and offline through video recording of the instruments used and / or the instrument handover. In the case of offline annotation, recorded data can be retrospectively manually evaluated, processed, and / or timestamped before being made available to the AI ​​unit for analysis.

[0038] According to some embodiments, the data and / or AI models can be taken over or supplemented via AI units of assistive robots from other operating rooms or clinics.

[0039] The AI ​​unit also includes, for example, an analysis functionality that calculates the probabilities for the use of a specific instrument and its subsequent instruments from the data collected during the learning phase. These probabilities can be calculated in an AI-supported model based, for example, on a stochastic process, in particular a Markov process. The model can represent the transition probabilities between individual instruments over the entire duration of an operation and / or over sequential time intervals. As already described, in some embodiments the probabilities are represented by Markov chains, whereby higher-order Markov chains can be used to account for dependencies on several preceding states. These states can describe previous steps.Alternatively or additionally, other machine learning methods, such as neural networks, especially convolutional neural networks (CNNs), can be used. In general, the AI ​​model of the forecasting unit can comprise a hybrid AI model that implements various machine learning methods.

[0040] During the execution phase, the AI ​​model can be configured to generate demand forecasts and / or optimize instrument provisioning based on previously determined probabilities. In some embodiments, the AI ​​system can remain self-learning during the execution phase. This can mean that demand forecasting can be improved during operation. For example, the frequency, duration, and / or sequence of use of medical instruments can continue to be recorded to continuously optimize the underlying model. Furthermore, the forecasting unit can be configured to identify usage-equivalent medical instruments that could serve as replacements for unavailable and / or defective medical instruments. These can then be provided as replacements for a demand-forecasted medical instrument.

[0041] The predictive unit, particularly the AI ​​model, can be configured to normalize the duration of a procedure and / or different procedures to make time-varying processes comparable. This can be achieved, for example, by applying dynamic time normalization, where the total duration of an operation is normalized to a uniform length. In this way, inhomogeneities in the probability distribution can be accounted for by analyzing defined phases of the operation, such as initiation, execution, and completion.

[0042] When using two or more medical instruments simultaneously, as is common in minimally invasive procedures, the AI ​​model can expand the number of states accordingly. Furthermore, specific adjustments can be made to ensure that the maximum order of the Markov chain is optimized depending on the number of available positioning units, particularly robotic arms.

[0043] Additionally, the forecasting unit can be configured to take into account manual requests for medical instruments from the operating room staff, allowing for short-term adjustments to the priorities set by the AI ​​model. For example, a surgeon can indicate a need for a specific medical instrument. This need can then be considered when generating the demand forecast, and / or the corresponding medical instrument can be provided with high priority.

[0044] The phrase "based on the demand forecast" can be understood to mean that the control unit is configured to activate at least one of the positioning units in such a way that a medical instrument for which the highest probability of need has been determined is picked up and made available to the surgeon. The control unit is therefore configured to activate the positioning units taking into account the demand for medical instruments.

[0045] The control unit can be configured to coordinate multiple positioning units in such a way that they perform parallel and / or independent movements. For example, the deployment of a new, demand-predicted instrument with one positioning unit can be performed simultaneously with the retrieval of a previously used medical instrument with another positioning unit.

[0046] In some embodiments, the control unit includes a safety unit designed to prevent collisions between the at least two positioning units. This safety unit can, for example, evaluate the image data from the camera and / or camera system and / or the navigated position of the robot arms to prevent the risk of a collision. The safety unit can also be designed to prevent collisions between one of the positioning units and the surgeon.

[0047] The provision of a medical instrument can be understood as presenting it to the surgeon in such a way that the surgeon can grasp and pick it up. The surgeon can therefore decide whether or not to pick up the instrument. In this sense, a provided instrument does not have to be picked up by the surgeon.

[0048] For example, if two positioning units each provide a medical instrument predicted to be needed, the surgeon can grasp and pick up one of the desired instruments and / or both instruments when the need actually arises.

[0049] In some embodiments, the control unit is configured to control the positioning units in such a way that each positioning unit moves the picked-up medical instrument from a storage position, where the instrument is kept, to a waiting position, where it is held ready for the surgeon to access. This enables efficient and / or error-resistant instrument retrieval. Consequently, the procedure flow is streamlined. For example, several medical instruments, predicted for future use, can be kept ready for access in the waiting position. When retrieval is required, these instruments do not first need to be selected from the storage position. This step can be performed upstream, allowing the retrieval process itself to be faster.Since at least two positioning units can be provided, each capable of holding a medical instrument readily accessible, the setup time can still be reduced. Incorrect predictions and / or the failure of the most likely prediction to materialize can be mitigated by keeping a selection of instruments predicted for demand readily available in the waiting position. This allows for flexible responses to various scenarios in the procedure flow and the consideration of alternative needs.

[0050] The storage position can be understood as a position in which the medical instrument is kept together with a selection of other medical instruments chosen before the procedure. The storage position can thus define a starting position in which the medical instrument is ready for pickup by one of the positioning units.

[0051] Furthermore, a storage position can be provided. A medical instrument that has been retrieved by the surgeon can be placed in the storage position. For example, the assistive robot can be configured to receive a used medical instrument from the surgeon and place it in the storage position and / or storage area. The storage position can coincide with a storage position and / or define a separate position.

[0052] A waiting position can be understood as a location where a medical instrument, predicted to be needed, is temporarily stored. From this waiting position, the medical instrument can then be retrieved when needed. If the need does not arise, it can be returned to its storage position and placed there. The waiting position can therefore define a location where a recorded, predicted-to-need medical instrument is kept readily accessible. In this sense, a selection of instruments can be kept in the waiting position that, according to the demand forecast, have a higher probability of being needed. These are, as already mentioned, pre-selected and do not need to be chosen from the storage position when a need arises. This is advantageous because a relatively large selection of medical instruments is stored in the storage position.In the waiting position, the number of readily available instruments may therefore be smaller than the number of medical instruments stored in the storage position.

[0053] "Ready for access" can be understood to mean that access to the medical instrument is simplified. In this sense, access can be quick and efficient. However, in a ready-for-access position, the instrument does not necessarily have to be physically prepared. Therefore, a ready-for-access position can differ from a prepared position, in which the instrument is handed over to the surgeon. The control unit can be configured to control the positioning units of a medical instrument held in a ready-for-access position, such that each positioning unit moves the corresponding medical instrument from the waiting position to a prepared position, in which the medical instrument is handed over to the surgeon and / or prepared.

[0054] Alternatively or additionally, the instruments can be stored in a waiting position. For example, several medical instruments predicted to be needed can be kept readily accessible together in the waiting position. This can be achieved, for instance, by the assistance robot including a storage device in the waiting position, with the positioning units configured to place the medical instrument predicted to be needed and / or picked up onto the storage device. The storage device could, for example, be a tray located near the surgeon or the surgical field.

[0055] According to some embodiments, the storage device can be moved from the waiting position to the staging position. In this case, the medical instrument is not staged in the waiting position. Rather, several medical instruments can be placed on the storage device in the staging position for the surgeon to pick up.

[0056] The time from a surgeon's request to the robot's delivery can be short if the waiting position during the procedure is significantly closer to the surgeon than the storage position. In other words, the distance to be covered for delivery can be reduced. In a first step, the forecasting unit can generate the demand forecast; in a second step, the control unit can activate at least one of the positioning units so that a predicted medical instrument is picked up and moved to the waiting position; and in a third step, the delivery can then take place. In this third step, the medical instrument only needs to be moved a short distance for delivery.

[0057] The movement between the receiving and waiting positions can be done quickly, but the movement between the waiting position and the provisioning position is slower because it involves human-machine interaction, i.e., handing the instrument to the surgeon.

[0058] According to some embodiments, the control unit is configured to perform a step in selecting at least one demand-forecasted medical instrument from several medical instruments in a retrieval area, where a supply of medical instruments is stored. The assistive robot can thus almost completely replicate the work of surgical assistants. The entire process, from instrument selection to instrument delivery, can be carried out by the assistive robot. Within the retrieval area, the several medical instruments that together define the supply can each be arranged in a storage position. The selection process can include the recognition of the demand-forecasted medical instrument.The supply of medical instruments can be stored, for example, in a sterile container and, in particular, arranged on an instrument tray inside the sterile container. A camera can be positioned in the recording area, designed to image the stored instruments and generate corresponding image data. Based on this image data, a demand-predicted instrument can be identified using image recognition. Alternatively, the positioning units can each be equipped with a corresponding camera.

[0059] According to further embodiments, the surgical system comprises a storage unit located within the receiving area, configured to hold the supply of medical instruments and to determine the status of the stored medical instruments. The storage unit is configured to transmit this status information to the medical assistive robot, in particular its control unit. According to further embodiments, the storage unit defines the receiving area. The status information may, for example, relate to the presence of a specific instrument. For instance, if a specific instrument is in stock, an identifier for that instrument may be stored on a data storage device of the storage unit. In some embodiments, the position of each stored instrument, particularly within the receiving area, may also be stored on the data storage device.The location can also be status information. The storage unit can include the sterile container. Accordingly, the storage unit can include a camera for detecting the stock of medical instruments and / or for determining the location of the stored medical instruments.

[0060] Furthermore, the assistive robot can include a communication system designed to transmit information about the stored instruments and / or their location from the storage unit to the control unit. This allows the stored instruments to be targeted accordingly.

[0061] Furthermore, the control unit can be configured to move at least one medical instrument, as predicted in terms of demand, from the receiving area to a waiting area using at least one of the positioning units, in order to keep it readily available. In particular, a preselection of readily available medical instruments, based on the demand forecast, is kept in the waiting area. Thus, several medical instruments can be kept readily available based on the demand forecast. These can be provided more quickly than an instrument that first has to be selected from the exception area. Moreover, the medical instrument can be moved part of the way from the receiving area to the staging position even before it is needed.Furthermore, since the pre-selection of readily available medical instruments is kept open, the overall probability of quickly providing the instrument that is actually needed can be increased. It is not necessary to rely on a single probability of need. Instead, the pre-selection can be adjusted according to the demand forecast. For example, instruments for which a probability of need has been determined to exceed a certain threshold can be included in the pre-selection and moved to the waiting area accordingly. The threshold could, for example, be 50%. In addition, medical devices can be dynamically moved from the waiting area to the receiving area, and especially moved back, if the probability of need decreases.

[0062] According to some embodiments, the control unit is configured to execute the following steps: moving at least one of the readily available medical instruments, particularly from the waiting area, to a transfer area by means of at least one of the positioning units, based on a control command from the surgeon; and handing the readily available medical instrument to the surgeon within the transfer area. By keeping the instruments predicted to be needed as readily available instruments in the waiting area until they are moved to the transfer area, a collision with the surgeon can be avoided. The surgeon can therefore focus on the steps to be performed during the procedure. Only when a new instrument is needed can it be moved to the transfer area so that the surgeon can pick it up.The waiting area can be located outside the surgeon's manipulation area, and / or the transfer area can be located within the surgeon's manipulation area. This area can be understood as the space within which the surgeon moves, particularly with their arms, during a procedural step performed on the patient. The instrument ready for use is moved into the transfer area following a corresponding control command. Using this control command, the surgeon can communicate their need for a specific instrument to the control unit. In this respect, the assistive robot can be configured to interact with the surgeon.

[0063] According to some embodiments, the control unit is configured to move the demand-predicted medical instrument from the receiving area to the waiting area at a delivery speed, and the ready-to-access medical instrument from the waiting area to the transfer area at a transfer speed. The transfer speed is lower than the delivery speed. This approach takes advantage of the fact that the risk of collision with the surgeon is low when moving the instrument into the waiting area. Accordingly, the instrument can be moved quickly, thus reducing delivery times. Conversely, a slow delivery speed may be advisable when moving the instrument into the transfer area to avoid a collision with the surgeon.However, since the medical instruments predicted for demand can be temporarily stored in the waiting area as readily accessible instruments according to the preselection, the distance over which the corresponding instrument has to be moved for provision in the handover area can be minimized.

[0064] To facilitate communication between the surgeon and the surgical robot, the robot can include a user interface configured to receive a control command from a user, particularly the surgeon, and to transmit at least one control signal to the control unit to indicate a need for a medical instrument. The surgeon can then register their need via the user interface and thereby initiate the handover or provision of the instrument. This makes interaction with the surgical robot intuitive and mirrors the interaction with surgical assistants. The user interface can also be configured to receive and process voice commands, allowing the surgeon to communicate with the robot in the same way as with surgical assistants. Operation is therefore highly intuitive. The control unit can be configured to control the positioning units based on the control signal.

[0065] The current step and the corresponding procedure step parameter can be reliably and securely identified if the medical assistance robot also includes a communication interface configured to receive usage information from at least one medical instrument used in the procedure. The procedure step recognition unit can be configured to determine the procedure step parameter based on this usage information. This usage information can include, for example, an instrument parameter from active devices, instruments, units, and / or the like, such as a light source, an RF device, a camera, an insufflator, and / or a pump. Based on this usage information, the active use of the instrument and / or device can be recorded, and conclusions can be drawn about the current step of the procedure.This enables precise determination of the current procedure step based on the actual use of the devices during the operation. For example, a light source power profile can be measured, a pump power profile can be determined, and / or the function of RF-assisted tissue cutting or coagulation can be determined, and the step can be identified based on this usage information. Furthermore, the procedure step recognition unit can be configured to determine the step based on usage information and image data from the camera (especially endoscopic and / or external) and / or associated camera system. If an AI model is used for step recognition, it can, for example, take usage information and image data as input parameters.For example, an AI model can be used to detect the steps and / or phases of a procedure, the input data and / or database of which includes image data and usage information, such as instrument parameters.

[0066] According to some embodiments, the surgical system further comprises a patient table on which the patient lies during the procedure, a waiting area where the surgeon remains during the procedure, a receiving area where a supply of medical instruments is stored, and a holding area where at least one medical instrument is kept readily accessible. The holding area is located closer to the patient table than the receiving area. As described above, the readily available medical instrument only needs to be moved a short distance when required. The step of selecting an instrument from the supply is eliminated and proactively performed. Instead, a pre-selection is made, which reduces the time required for retrieval when needed.

[0067] Furthermore, the surgical system can include an instrument monitoring unit designed to process usage information from multiple medical instruments and transmit this information to the procedure step recognition unit. The instrument monitoring unit can function as a hub, bundling the lines and / or other components of several supply units and medical instruments. Usage information can then be centrally collected and made available easily and efficiently.

[0068] The present invention is described below by way of example with reference to the accompanying figures. The drawing, the description, and the claims contain numerous features in combination. A person skilled in the art will expediently consider the features individually and use them meaningfully in combination within the scope of the claims.

[0069] If more than one instance of a particular object exists, only one of them may be identified with a reference symbol in the figures and description. The description of this instance can then be applied to the other instances of the object. If objects are named using numerical terms, such as first, second, third object, etc., these serve to identify and / or classify objects. Thus, for example, a first object and a third object, but not a second object, may be included. However, numerical terms could also indicate a number and / or sequence of objects.

[0070] They show: Fig. 1. A schematic representation of a surgical system; Fig. 2 a schematic representation of another embodiment of a surgical system; Fig. 3 a schematic representation of an assistance robot; Fig. 4. Another schematic representation of the surgical system; Fig. 5a a schematic state diagram of a process of predictive instrument provisioning; Fig. 5b another schematic state diagram of the process of predictive instrument provisioning; Fig. 5c another schematic state diagram of the process of predictive instrument provisioning; Fig. 5d another schematic state diagram of the process of predictive instrument provisioning, Fig. 6. A schematic diagram illustrating the time normalization of procedures of different types; and Fig. 7 a schematic flowchart.

[0071] The Fig. Figure 1 shows a schematic representation of a surgical system 60, comprising a medical assistance robot 10 and a patient table 54 on which a patient 20 lies during a procedure. In this case, the procedure is a tumor removal from the intestine and comprises several steps to be performed on the patient. The procedure is performed by a surgeon 12, who is assisted by an anesthesiologist 13.

[0072] Surgeon 12 is assisted by robot 10 during the procedure. Robot 10 is configured to proactively provide surgeon 12 with medical instruments 16. To do this, it detects the current step of the procedure and, based on this detection, generates a demand forecast containing the probability of needing various medical instruments 16. Based on this demand forecast, medical instruments 16 are selected and proactively provided to surgeon 12. These are selected, for example, using instrument usage or instrument transition probabilities, the determination of which is based on an AI model. In this way, appropriate medical instruments 16 can be determined for each step of the procedure. The surgeon can request a specific instrument 16 using a voice command, and it will then be provided.

[0073] To shorten the time required to provide instruments after a request has been submitted, the assistance robot 10 includes several positioning units 14, each configured to pick up and move a medical instrument 16. In addition, medical instruments 26 predicted to be needed are kept readily accessible in a waiting position 32, according to the demand forecast. The waiting position 32 is located closer to the surgeon 12 than a storage position 30, which contains a supply 38 of medical instruments 16. Accordingly, several medical instruments 26 predicted to be needed are stored there (see also Fig. 2) are arranged near the surgeon and held there by a positioning unit 14. These are stored there as readily accessible medical instruments 42 in a waiting position and can be quickly provided. After the surgeon 12 requests them, they no longer need to be moved from storage position 30 to the surgeon 12. The delivery time can be reduced accordingly. Furthermore, several instruments 26 (see also Fig. 2) Since the instruments 16 are kept readily available, requests for different instruments 16 can be considered. Several options for readily available instruments 42 are therefore provided, allowing for a response to different requests. This increases the probability that the instrument 16 for which there is actually a need will be kept readily available as an instrument 42. In this respect, redundancy is introduced.

[0074] The assistive robot 10 comprises three positioning units 14, configured as robot arms, a procedure step recognition unit 18, a prediction unit 22, a control unit 24, a user interface 48, and a communication interface 50. Furthermore, the surgical system 60, specifically the assistive robot 10 and / or the procedure step recognition unit 18, includes three cameras 34, each configured as a stereo camera. The surgeon 12 is filmed by the cameras 34 during the procedure. Based on the resulting image data and / or the image data from an endoscopic camera used by the surgeon, the procedure step recognition unit 18 can determine a procedure step parameter that characterizes the current step among several steps of the procedure. The procedure step recognition unit 18 may include image recognition algorithms, particularly those based on artificial intelligence.Several well-known image recognition methods can be used.

[0075] Furthermore, the procedure step recognition unit 18 is configured to take into account usage information of at least one medical instrument 52 used in the procedure when determining the procedure step parameter. The usage information is determined by means of an instrument monitoring unit 62 of the surgical system 60. This unit is configured to process usage information of several medical instruments 52 and to transmit the usage information to the procedure step recognition unit 18. The assistance robot 10 includes a communication interface 50 for receiving the usage information. The procedure according to Fig. The instruments used (52) are an endoscope in the surgeon's hand and a breathing mask for ventilating the patient. It is also conceivable to consider only the usage information for medical instruments (52) used by the surgeon himself. For example, usage information for the endoscope according to... Fig. 1 and an HF coagulation device (not shown). The user information for the breathing mask is not used according to this example. Corresponding supply units (not shown by the endoscope) operate these instruments 52. Performance data of these supply units are used as user information and are centrally recorded and processed by the instrument monitoring unit 62. The instrument monitoring unit 62 accordingly comprises several lines, two of which are only partially shown.

[0076] The user interface 48 includes a microphone configured to receive voice commands from the surgeon 12. The surgeon can therefore control the assistive robot 10 using voice commands. In this sense, the voice command is a control command. The user interface 48 is communicatively connected to the control unit 24. Based on the control command, the user interface generates a control signal to indicate a need for a medical instrument 16. The surgeon 12 can thus indicate their need and initiate the provision of the instrument.

[0077] The control unit 24, the prediction unit 22, and the procedure step recognition unit 18 are jointly implemented on a computing unit 58. This unit is located in an operating room where the procedure takes place. However, it is positioned externally to an assistive robot base 28. The positioning units 14 are articulatedly mounted to the assistive robot base 28. Furthermore, the base includes power electronics for moving the positioning units 14.

[0078] Forecasting unit 22 is configured to generate a demand forecast for medical instruments 16 based on the procedure step parameter, in particular using an AI model. As already described, it assigns, for example, a demand probability to each of the medical instruments 16, which are stored in the demand forecast.

[0079] The demand probabilities are continuously updated. This allows for dynamic adjustments to the procedure and its progress.

[0080] The predictive unit comprises a computer-based AI unit that operates in a learning phase and / or an execution phase. During the learning phase, the AI ​​unit features an annotation function designed to identify the instruments used during a variety of operations of a specific type and to record their frequency, duration, and sequence of use. The annotation could also be extended to a variety of operations of a specific type, using a particular surgical technique and / or for an individual surgeon. Annotation can be performed both online, during the operation, and offline through video recording of instrument handovers. When annotation is performed offline, recorded data can be retrospectively manually evaluated, processed, and / or timestamped before being made available to the AI ​​unit for analysis.

[0081] The AI ​​unit also includes an analysis function that calculates the probabilities for the use of a specific instrument and its subsequent instruments from the data collected during the learning phase. These probabilities can be calculated in an AI-supported model based, for example, on a stochastic process, particularly a Markov process. The model can represent the transition probabilities between individual instruments over the entire duration of an operation and / or across sequential time intervals.

[0082] During the learning phase, additional tools can be used to accurately map the procedure. For example, gloves with haptic sensors can be used. These can record the handling of instruments and other usage data, which is used to optimize the model. These tools are no longer used during the execution phase. Data collection from the instruments used can be performed concurrently with the execution phase to optimize the learning phase and / or the AI ​​models.

[0083] The probabilities are represented by Markov chains, with higher-order Markov chains being used to account for dependencies on multiple preceding states. The states describe previous steps.

[0084] It is also conceivable to use other machine learning methods, such as neural networks, especially convolutional neural networks (CNNs). In general, the AI ​​model of the forecasting unit can comprise a hybrid AI model that implements various machine learning methods.

[0085] In the execution phase, the AI ​​model is configured to generate demand forecasts and / or optimize instrument provisioning based on previously determined probabilities. The AI ​​system is designed to be self-learning during the execution phase. This means that demand forecasting is continuously improved during operation. For example, the frequency, duration, and / or sequence of use of medical instruments can be continuously recorded to optimize the underlying model. Furthermore, the forecasting unit can be configured to identify usage-equivalent medical instruments that could serve as replacements for unavailable and / or defective medical instruments. These replacements can then be provided as substitutes for a medical instrument identified in the forecast.

[0086] The procedure step recognition unit 18, in particular the AI ​​model, is configured to normalize the duration of the procedure and / or different procedures in order to make time-varying processes comparable. See, for example, the schematic diagram according to Fig. 6. The various procedures are plotted on a timeline labeled OPM (OP1, OP2, OP3, etc.). This can be achieved, for example, by applying dynamic time normalization, where the total duration of an operation is normalized to a uniform length. This allows for the consideration of inhomogeneities in the probability distribution by analyzing defined OP phases such as initiation, execution, and completion. This normalization thus enables the comparison of procedures of different durations, and the states and their state probabilities are presented consistently. To counteract potential inhomogeneities in the probability distribution of the Markov chain over the entire procedure duration, a differentiated analysis can be performed. This analysis considers defined phases of the procedure, such as procedure initiation (e.g.,The process encompasses the opening of the abdominal cavity or insertion of trocars in minimally invasive procedures, the execution of the procedure (e.g., gallbladder removal, open or laparoscopic), and the completion of the procedure (e.g., wound closure or wound care). The described standardization and phase analysis enables more precise modeling of procedure types and improves the quality of demand forecasting.

[0087] When two or more medical instruments are used simultaneously, as is common in minimally invasive procedures, the AI ​​model can expand the number of states accordingly. Furthermore, specific adjustments can be made to ensure that the maximum order of the Markov chain is optimized depending on the number of available positioning units, particularly robotic arms.

[0088] Additionally, forecasting unit 22 is configured to accept manual requests for medical instruments 16 from surgeon 12, thereby changing the priorities of the KL model in the short term. For example, surgeon 12 can indicate a need for a specific medical instrument 16. This need will be taken into account when generating the demand forecast, and the corresponding medical instrument 16 will be provided with high priority.

[0089] Based on the results of the forecasting unit 22, the control unit 24 controls the positioning units 14. It is configured to control at least one of the positioning units 14 based on the demand forecast in order to pick up at least one demand-forecasted medical instrument 26 and make it available to the surgeon 12. If the demand forecast indicates a demand probability of over 50% for a specific medical instrument 16, it is classified as a demand-forecasted medical instrument 26. Subsequently, it is picked up so that it can be made available to the surgeon 12 if needed.

[0090] As previously described, the instrument 26, which is projected to be needed, will first be moved to waiting position 32 to be kept there as a readily accessible medical instrument 42. It will therefore be moved closer to the surgeon 12 and kept there for quick deployment.

[0091] First, after the demand forecast has been generated, the predicted instrument 26 is selected from the stock 38 of medical instruments. The stock 38 is stored in a storage unit 64 of the surgical system 60. This unit defines a receiving area 36. Within the receiving area 36, ​​several medical instruments 16 are arranged in storage positions 30. Using image recognition and / or electromagnetic identification, the predicted instrument 26 is identified among these instruments 16 and, after selection, picked up by the positioning unit 14.

[0092] The instrument 26, as previously described, is then moved to the waiting position 32 and held there as a readily accessible medical instrument 42. The waiting position 32 is located in a waiting area 40, in which several readily accessible instruments 42 are kept for later use. The surgical system 60 comprises these areas, namely the waiting area 40 and the receiving area 36. It also includes a patient area 56, in which the surgeon 12 remains during the procedure. This area includes an intervention area 66, within which an incision is made in the patient 20. The waiting area 40 is located closer to the patient area 56 than the receiving area 36.

[0093] The multiple readily available instruments 42 are part of a preselection 44 of readily available medical instruments 42, which are arranged in the waiting area 40. These are the instruments 16 that are assigned a high probability of being needed according to the demand forecast. Here, 50% is defined as the threshold for classification as a demand-forecasted instrument 26, which is to be kept readily available as an instrument 42. Other thresholds are also conceivable, particularly in connection with the number of positioning units 14. The more positioning units 14 are provided, the lower the threshold can be and the higher the probability that the instrument used next by the surgeon is located in the waiting area.

[0094] If the surgeon then signals a need using the control command or voice command, the corresponding instrument is made available. For this purpose, the relevant instrument 42, which is ready for access, is moved from waiting area 40 to a transfer area 46. There, the instrument is handed over to the surgeon.

[0095] There is an increased risk of collision when moving into the transfer area 46, as the waiting area 56 includes the transfer area 46. The waiting area 40 and the receiving area 36 are located outside the waiting area 56, which is why there is a lower risk of collision there. Accordingly, the delivery speed into the transfer area 46 can be lower relative to the delivery speed from the receiving area 36 to the waiting area 40.

[0096] It is also possible that the surgeon may remove a medical instrument from or return it to the waiting area; this approach to the waiting area is detected by the monitoring unit, and movements of the positioning units to or from the waiting area are stopped and released again when the surgeon's interaction in the waiting area is complete.

[0097] Furthermore, the assistance robot 10 is equipped to detect and counteract potential collisions with operating room personnel and objects using sensors. This is achieved through optical monitoring via cameras 34 and light barriers (not shown).

[0098] The storage unit 64 comprises a sterile container (not shown in detail) and an instrument tray located inside the sterile container. The instruments 16 are arranged on the instrument tray. The storage unit also includes a camera located in the recording area 36, ​​which is configured to image the stored instruments 16 and generate corresponding image data. Based on this image data, the instrument 26 predicted to be needed is identified using image recognition. Alternatively, the positioning units 14, for example, can each have a corresponding camera. The camera can also determine the position of the instruments 16. This information is communicated to the control unit 24, thus facilitating the selection of the instrument 26 predicted to be needed.

[0099] The Fig. Figure 2 shows a schematic representation of another embodiment of a surgical system 60', comprising a medical assistive robot 10'. The surgical system 60' is very similar to the surgical system 60, which is why the differences are primarily discussed. In addition, the representation of some units is omitted and the Fig. 1 referred. In contrast to the assistance robot 10 according to Fig. 1 The assistance robot 10' has a storage device 68 (see also Fig. 3) The storage device 68 has a storage surface on which medical instruments can be placed. Furthermore, the storage device 68 is located in the waiting area. Accordingly, a medical instrument 26, as predicted for future use, can be placed on the storage device 68 to keep it readily available as an accessible medical instrument 42. The surgeon 12 can then retrieve a desired instrument from the storage device 68 as needed. Thus, the pre-selection 44, based on the predicted need, is kept readily accessible on this device.

[0100] Furthermore, the storage device 68 is horizontally movable. If necessary, it can therefore be moved from a waiting position to the transfer area 46 (see Fig. 1) be movable. Accordingly, it forms a positioning unit.

[0101] The Fig. Figure 3 shows a schematic representation of a section of the assistance robot 10'. More precisely, positioning units 14 are shown, which are attached to the assistance robot base 28. The positioning units 14 are designed as multi-jointed robot arms, each having a gripper 70 at its distal end for receiving a medical instrument.

[0102] In some embodiments, the computing unit 58 is according to Fig. 1 is arranged within the assistance robot base 28. It is also conceivable that at least one of the cameras 34 is arranged on the assistance robot base 28.

[0103] Furthermore, the storage device 68 can be seen, for a more detailed description of which refer to the Fig. Reference is made to 2. On it, medical instruments 26, as predicted for demand, are stored as readily accessible medical instruments 42. The storage device 68 can be moved into a transfer area 46. This makes the instrument requested by the surgeon 12 available. Accordingly, the transfer to the surgeon can take place via the storage device 68.

[0104] Fig. Figure 4 shows another schematic representation of the surgical system 60. This includes the assistance robot 10, the control unit 24, and the instrument monitoring unit 62. It also includes an AI unit 72, on which the prognostication unit and the procedure step recognition unit are implemented.

[0105] The assistance robot 10 has the positioning unit 10 with the grippers 70. The grippers 70 are equipped with sensors and configured to pick up medical instruments. The instrument monitoring unit 62 is equipped with several supply units 78 of the medical instruments 52 used in the procedure (see Fig. 1) connected to and communicating with the assistance robot 10. The surgical system 60 has a critical movement range 74, within which the acquisition area 36 and the waiting area 40 are located. Due to the rapid movements of the positioning units 14, there is a risk of collision between them within this range 74. To prevent a collision, the surgical system 60 has a monitoring unit 76 that optically monitors the critical movement range 74. If there is an immediate risk of collision between two positioning units 14, the control unit 24 is informed accordingly and the collision is avoided.

[0106] Furthermore, the surgical system has a transfer area 46. A delivery speed during a movement from the waiting area 40 to the transfer area 46 is lower than a delivery speed within the critical movement area 74.

[0107] The positioning units 14 and the storage device 68 ( Fig. 1, Fig. 2, Fig. 3 to Fig. 4) may also accept used medical instruments or other items from the surgeon 12.

[0108] The Fig. Figures 5a to 5d each show a schematic state diagram of a predictive instrument deployment process, derived in particular from the transition probabilities of instrument use via the AI ​​model. Each state diagram corresponds to a step. According to some embodiments, different transition probabilities can exist within a single step. These can also be represented by the state diagram. Fig. 5a shows a state diagram according to step k, in Fig. 5b shows a state diagram according to step k+1, in Fig. 5c shows a state diagram according to step k+2 and in Fig. Figure 5d shows a state diagram according to step k+3. Xn denotes a medical instrument in each step (e.g., X1 corresponds to the first medical instrument). These instruments are in different states in the various steps. Box 80 defines an application state by the surgeon, box 82 an access-ready state in the waiting area, and box 84 a storage state in the receiving area. The numbered arrows indicate possible movement sequences. The following configurations exist in steps k to k+3: Step k: X2 is in use X1 is located in the waiting area X2 is located in the waiting area X4 is located in the recording area Step k+1: X2 is taken back in the handover area and transported to the waiting area. X3 is transported from the waiting area to the transition area for handover and application. X1 is transported from the waiting area to the reception area. X4 is transported from the recording area to the waiting area (The arrows (1; 2) indicate the possible sequence of movements) Step k+2: X3 is taken back in the handover area and transported to the waiting area. X4 is transported from the waiting area to the transition area for handover and application. X3 is being transported from the waiting area to the reception area. X1 is transported from the staging area to the waiting area (The arrows (1; 2; 3) indicate the possible sequence of movements) Step k+3: X4 is taken back in the handover area and transported to the waiting area. X2 is transported from the waiting area to the transition area for handover and application. X1 is transported from the waiting area to the reception area. X3 is transported from the recording area to the waiting area (The arrows (1; 2; 3) indicate the possible sequence of movements)

[0109] Fig. Figure 7 shows a schematic flowchart for AI- and robot-assisted instrument deployment. It includes the following steps: - Step 90: Provision of the stock of medical instruments, especially those with instrument coding; - Step 92: Learning phase; recording instrument changes during a procedure of a specific type with annotation; - Step 94: Analysis phase; training of the AI ​​model (e.g., probability transition matrix); - Step 96: predictive robot-assisted instrument provisioning based on the AI ​​model; and - Step 98: Optimization of the AI ​​model through data collection and analysis of further procedures of a specific type. Reference symbol list 10 assistance robots 12 Surgeon 13 Anesthesiologist 14 Positioning unit 16 medical instruments 18 Procedure Step Recognition Unit 20 patients 22 Forecast Unit 24 control unit 26 demand-forecasted medical instrument 28 Assistance robot base 30 storage positions 32 waiting positions 34 Camera 36 Recording area 38 stock 40 Waiting area 42 readily available medical instruments 44 Pre-selection 46 Transfer area 48 User interface 50 Communication interface 52. Medical instrument used in the procedure 54 patient beds 56 Lounge area 58 computing units 60 surgical system 62 Instrument monitoring unit 64 storage units 66 Intervention area 68 Storage device 70 grippers 72 AI units 74 critical range of motion 76 Monitoring unit 78 supply unit 80 boxes 82 boxes 84 boxes 90 steps 92 steps 94 steps 96 steps 98 steps

Claims

Medical assistance robot (10) for predictive instrument provision for a surgeon (12), comprising: - at least two positioning units (14), each configured to pick up and move at least one medical instrument (16); - a procedure step recognition unit (18) configured to determine at least one procedure step parameter that characterizes a current step of several steps in which the surgeon (12) is currently performing a procedure on a patient (20); - a forecasting unit (22) configured to generate a demand forecast for medical instruments (16) based on the procedure step parameter;and a control unit (24) configured to control at least one of the positioning units (14) based on the demand forecast, in order to receive at least one demand-forecasted medical instrument (26) and make it available to the surgeon (12). Medical assistance robot (10) according to claim 1, wherein the control unit (24) is configured to control the positioning units (14) such that the respective positioning unit (14) moves the picked-up medical instrument (16) from a storage position (30) in which the medical instrument (16) is kept to a waiting position (32) in which the medical instrument (16) is kept ready for access by the surgeon (12). Medical assistance robot (10) according to claim 2, wherein the waiting position (32) is arranged closer to the surgeon (12) than the storage position (30) during the performance of the procedure. Medical assistance robot (10) according to one of the preceding claims, wherein the control unit (24) is configured to perform the following step: - Selection of at least one demand-forecasted medical instrument (26) from several medical instruments (16) in a receiving area (36), wherein a supply (38) of medical instruments (16) is stored in the receiving area (36). Medical assistance robot (10) according to one of the preceding claims, wherein the control unit (24) is configured to perform the following step: - Moving at least one demand-forecasted medical instrument (26), in particular from the receiving area (36), into a waiting area (40) by means of at least one of the positioning units (14) in order to keep it available as an accessible medical instrument (42), wherein in particular a preselection (44) of accessible medical instruments (42) is kept available in the waiting area (40), which is based on the demand forecast. Medical assistance robot (10) according to claim 5, wherein the control unit (24) is configured to perform the following steps: - Moving at least one of the accessible medical instruments (42), in particular from the waiting area (40), into a transfer area (46) by means of at least one of the positioning units (14), based on a control command from the surgeon (12); and - Transferring the accessible medical instrument (42) to the surgeon (12) within the transfer area (46). Medical assistance robot (10) according to claims 5 and 6, wherein the control unit (24) is configured to move the demand-forecasted medical instrument (26) from the receiving area (36) to the waiting area (40) at a delivery speed and the access-ready medical instrument (42) from the waiting area (40) to the transfer area (46) at a transfer speed, wherein the transfer speed is less than the delivery speed. Medical assistance robot (10) according to one of the preceding claims, further comprising: - a user interface (48) configured to receive a control command from a user, in particular the surgeon (12), and to transmit at least one control signal to the control unit (24) to indicate a need for a medical instrument (16). Medical assistance robot (10) according to one of the preceding claims, further comprising: - a communication interface (50) configured to receive usage information of at least one medical instrument (52) used in the procedure, wherein the procedure step recognition unit (18) is configured to determine the procedure step parameter based on the usage information. Surgical system (60) comprising a medical assistance robot (10) according to any of the preceding claims. Surgical system (60) according to claim 10, further comprising: - a patient bed (54) on which the patient (20) lies during the performance of the procedure; - a waiting area (56) in which the surgeon (12) stays during the performance of the procedure; - a receiving area (36) in which a supply (38) of medical instruments (16) is kept; - a waiting area (40) in which at least one medical instrument (16) is kept ready for access, wherein the waiting area (40) is arranged closer to the receiving area (56) than the receiving area (36). Surgical system (60) according to one of claims 10 or 11, further comprising: - an instrument monitoring unit (62) which is configured to process usage information of several medical instruments (52) and to transmit the usage information to the procedure step recognition unit (18). Surgical system (60) according to one of claims 10 to 12, further comprising: - a storage unit (64) arranged in a receiving area (36) and configured to store a supply (38) of medical instruments (16) and to determine status information about the stored medical instruments (16), wherein the storage unit (64) is configured to transmit the status information to the medical assistance robot (10), in particular the control unit (24). Method for operating a medical assistance robot (10) according to one of claims 1 to 9 and / or a surgical system (60) according to one of claims 10 to 13 .

Images