SYSTEM AND METHOD FOR DETECTING AN OPERATING STATE OR TREATMENT PROGRESS IN BLOOD TREATMENT

Description

Technical field

[0001] The invention relates to a blood treatment system and a computer-implemented method for detecting an operating state or treatment course or a deviation from an ideal or complication-free operating state or treatment course in a blood treatment, in particular in an extracorporeal blood treatment. background

[0002] In cases of chronic kidney failure, various methods of mechanical blood treatment are used to remove waste products and reduce fluid loss. In hemodialysis, diffusive transport of substances across the semipermeable membrane of a filter predominates, while in hemofiltration, convective transport of substances across the filter membrane occurs. Hemodiafiltration combines both methods. When the term "dialysis" is used below, it encompasses both pure dialysis treatment and hemodiafiltration.

[0003] Another blood treatment procedure for removing unwanted substances from the blood is therapeutic apheresis, in which whole blood and plasma are separated.

[0004] The known devices for carrying out the aforementioned blood treatment procedures have an extracorporeal blood circulation system and a blood treatment unit. The extracorporeal blood circulation system comprises an arterial blood line that supplies blood drawn from the patient to a blood treatment unit and a venous blood line that carries treated blood away from the blood treatment unit.

[0005] Blood to be treated is taken from the patient via a patient access point, and treated blood is also returned to the patient via the same access point.

[0006] In dialysis, the blood treatment unit is designed as a filter unit with a blood chamber connected to the dialysis fluid chamber via a semipermeable membrane. The dialysis fluid chamber is connected to the dialysis fluid system. The dialysis fluid system includes a dialysis fluid supply line that carries fresh dialysate to the dialysis fluid chamber and a dialysis fluid discharge line leading from the dialysis fluid chamber.

[0007] To monitor the complication-free course of dialysis treatment, a dialysis machine is typically equipped with a variety of sensors. For example, conductivity sensors are installed in both the dialysate supply and discharge lines to measure the electrical conductivity of the dialysate. A comparison of the conductivity values ​​measured by the two sensors allows conclusions to be drawn about the purification performance (clearance).

[0008] The extracorporeal blood circulation typically includes blood temperature sensors and blood pressure sensors in the arterial and venous blood vessels.

[0009] Automated analysis of sensor signals typically relies on exceeding or falling below threshold values ​​that are set to monitor the sensor signals. Setting these threshold values ​​requires an experienced operator and is often prone to errors.

[0010] The execution of blood treatment by the attending medical personnel relies largely on established routines. Interpreting the complex sensor signals and their trends can be challenging for the user, especially in atypical situations. The possibility of an incorrect user response resulting from such an error poses a risk to the patient.

[0011] Document US2011105979A1 describes healthcare devices and procedures that capture subjective and objective measurements of a patient's health status and treatment. These measurements are analyzed (e.g., by correlating them with each other and / or with standards) and reports are generated to support the ongoing diagnosis and treatment of the patient, as well as to assist physicians, nurses, and other healthcare professionals in decision-making, monitoring adherence to treatment protocols, facilitating regulatory compliance, billing, and so on. A processor generates patient queries related to treatments performed by the dialysis machine. The queries relate, at least in part, to subjective topics, such as…the patient's state of mental health and well-being, quality of life, extent of pain, and assessment of therapy success.

[0012] Document US2011087499A1 describes a system, a procedure, and a computer-readable medium for planning and optimizing dialysis treatments. The system may include a processor programmed to perform operations, including receiving a dialysis treatment parameter set and retrieving a predetermined list of dialyzers and corresponding cost and performance data for each dialyzer, determining the cost impact based on both clinical impact and cost value.

[0013] It is therefore the object of the present invention to solve at least one of the problems stated and to provide a computer-aided system that enables reliable detection of an operating state or treatment course of a blood treatment. Summary

[0014] This problem is solved by a blood treatment system according to claim 1, a computer-implemented method according to claim 6, a computer program product according to claim 11, and a computer program according to claim 127. Advantageous embodiments are specified in the dependent claims.

[0015] One embodiment relates to a computer system for detecting an operating state, a treatment progress, or a deviation from an ideal or complication-free operating state of a blood treatment machine or a treatment progress of a blood treatment. Examples of a blood treatment machine are a plasmapheresis machine, a hemodialysis machine, or a hemofiltration machine.

[0016] The computer system includes a first interface adapted to receive an operating parameter monitoring signal, wherein the operating parameter monitoring signal is a time-dependent signal from a sensor for monitoring an operating parameter of the blood treatment machine or a signal derived from the time-dependent signal of the operating parameter.

[0017] The creation of a derived signal involves extracting a temporal segment or window from the time-dependent signal of the operating parameter, for example, through a windowing operation, particularly the application of a weighted window function. Creating a derived signal can further involve extracting features from the time-dependent signal of the operating parameter or subjecting the time-dependent signal of the operating parameter to a filtering operation. The filtering operation can be linear or non-linear. Examples of linear filtering operations include a Fourier transform or a wavelet transform.

[0018] The operating parameter may include, for example, a blood flow rate, a blood temperature, a blood hematocrit in an extracorporeal blood circuit, a dialysate temperature in hemodialysis, a dialysate conductivity in hemodialysis, a blood pressure in an extracorporeal blood circuit, in particular an arterial blood pressure, a venous blood pressure, or a combination of arterial and venous blood pressure, or the power consumption of an actuator, in particular a blood pump. An operating parameter monitoring signal within the meaning of this disclosure may be formed by combining several operating parameter monitoring signals.

[0019] The computer system further includes a second interface adapted to receive a user response signal, where a user response signal represents a user input signal regarding the operating status or treatment progress, or deviation from an ideal or complication-free treatment progress or operating status, of a blood treatment performed with the blood treatment machine. The first and second interfaces can be implemented as a single interface.

[0020] The operating condition or treatment course, or the deviation from an ideal or complication-free operating condition or treatment course, can relate to the patient, the blood treatment machine, or the patient-blood treatment machine system. Examples include disconnection of the venous needle from the patient's access point, clotting of the dialyzer membrane, recirculation of the treated blood in the arterial part of the extracorporeal circulation, or a hypertensive crisis.

[0021] The user response signal is one of the following: an adjustment of a blood flow rate or ultrafiltration rate, the initiation of electrolyte administration, a manually triggered alarm or treatment stoppage, the detection of a hypertensive crisis, or the detection of a needle dislocation.

[0022] The user response can occur during ongoing treatment or based on an evaluation of historical treatment data, such as a historical operating parameter monitoring signal. The user response can be initiated by a patient, a practitioner, or an expert.

[0023] The user response signal can be timestamped to indicate the time of the user response or input in relation to the treatment process.

[0024] The user input lies within the time period or time window, or adjacent to it.

[0025] The computer system advantageously includes a storage unit configured to store a plurality of user response signals and operating parameter monitoring signals in relation to each other.

[0026] The computer system also includes an evaluation unit adapted to generate mapping or reference data that establishes a correlation between a specific user response and one or more corresponding operational parameter monitoring signals.

[0027] The mapping and reference data can be in the form of a function that maps an operating parameter monitoring signal to a corresponding user response.

[0028] The corresponding operating parameter monitoring signal can represent a typical operating parameter monitoring signal that matches a specific user action. The associated operating parameter monitoring signal can represent a variety of operating parameter monitoring signals that match a specific user action, or an averaging of a variety of operating parameter monitoring signals that match a specific user action. The averaging can be linear or non-linear.

[0029] The generation of the assignment and reference data can be carried out taking into account a machine-specific parameter of the blood treatment machine, for example in the sense that normalization takes place taking into account the machine-specific parameter.

[0030] Generating the mapping or reference data can be based on the application of a learning algorithm, such as a neural network, where the mapping between a specific user response and an associated operational parameter monitoring signal is learned from a large number of data tuples. Each data tuple comprises a specific user response and a corresponding operational parameter monitoring signal that was recorded at the time of the user response.

[0031] The evaluation unit is further adapted to compare a specific operating parameter monitoring signal, preferably a momentarily recorded operating parameter monitoring signal of an ongoing blood treatment, with the assignment or reference data in order to derive an operating state or treatment progress, or a deviation from an ideal or complication-free operating state or treatment progress of the blood treatment machine, in particular a momentary operating state or treatment progress, or a momentary deviation from an ideal or complication-free operating state of the blood treatment machine or the treatment progress of the ongoing blood treatment, or to derive a typical user reaction to this.

[0032] The specific operating parameter monitoring signal and the corresponding typical user response can be used in an iterative learning process to generate further and / or improved mapping and reference data.

[0033] The current operating state or the current deviation from an ideal or complication-free operating state may be linked to a typical or appropriate user response, for example, if the current operating state is an alarm state and the appropriate user response is an interruption of the blood treatment.

[0034] The blood treatment machine can be configured to provide the user with a suggestion via a user interface to make the typical or appropriate user response.

[0035] The blood treatment machine can also be configured to automatically adjust the operating parameters in response to typical or appropriate user reactions, such as reducing the blood flow rate or interrupting the treatment.

[0036] One embodiment relates to a computer-implemented method for detecting an operating state of a blood treatment machine or a treatment progress of a blood treatment performed with the blood treatment machine, or a deviation from an ideal or complication-free operating state or treatment progress. The computer-implemented method comprises the following steps: Receiving an operating parameter monitoring signal, wherein the operating parameter monitoring signal is a time-dependent signal from a sensor for monitoring an operating parameter of the blood treatment machine or a signal derived from the time-dependent signal of the operating parameter; receiving a user response signal, wherein the user response signal is a signal of a user input with regard to the operating state or the course of treatment or the deviation from an ideal or complication-free operating state or course of treatment; storing a plurality of user response signals and operating parameter monitoring signals in relation to each other; generating mapping or reference data that establishes a mapping or correlation of a specific user response with an associated operating parameter monitoring signal; storing the mapping or reference data.Receiving a specific operating parameter monitoring signal at a specific time during blood treatment, and comparing that signal with the mapping or reference data to derive a current operating state or treatment progress, or a current deviation from an ideal or complication-free operating state or treatment progress at that specific time, or a typical user response thereto.

[0037] The same modifications can be applied to the computer-implemented method as to the described computer system. Brief description of the drawings

[0038] Figure 1 shows a dialysis system with a hemodialysis machine in an embodiment according to the present teaching. Figure 2 shows a blood treatment system in an embodiment according to the present teaching. Figure 3shows a flowchart of a computer-implemented method in an embodiment according to the present teaching. Figure 4 shows the mapping between user input and signal pattern using the example of a venous and an arterial pressure pattern. Detailed description of the drawings

[0039] Figure 1 shows a dialysis system with 200 with a dialysis machine 100 and a central monitoring unit 40.

[0040] The dialysis machine 20 has a blood treatment unit 2 (dialyzer) with a dialysis fluid chamber 4 in a dialysis fluid circuit 22, which is separated from a blood chamber 3 in a blood circuit by a semipermeable membrane 5.

[0041] Prepared dialysis fluid is supplied by a dialysis fluid source 16 and conveyed by a dialysis fluid pump 17 via the dialysis fluid supply line 15 into the dialysis fluid chamber 4. Used dialysis fluid is conveyed by the dialysis fluid pump 13 via the dialysis fluid discharge line 11 from the dialysis fluid chamber 4 to a drain 12. In an alternative embodiment, the used dialysis fluid can also be reconditioned.

[0042] In the dialysis fluid chamber 4, the dialysis fluid absorbs the uremic substances that pass from the blood in the blood chamber 3 into the dialysis fluid via the dialyzer membrane 5.

[0043] A balancing unit 14 ensures the balance between fresh and used dialysis fluid. Flow sensors, scales, or balancing chambers can be used for this purpose. In the latter case, the provision of a second dialysis fluid pump may be unnecessary.

[0044] In the optional configuration of a hemodiafiltration machine shown, an additional substitution line 18 is provided, through which filtered dialyzer fluid, pumped by a substitution pump 19, is added to the blood being treated as a substitute. The filter required for this is not shown for the sake of simplicity. The substitution fluid can be added downstream of the dialyzer (postdilution) or upstream of the dialyzer (predilution), as shown. A combination of these methods is possible.

[0045] The blood to be treated is taken from patient 7 via a patient access point and supplied by the blood pump 23 via the arterial blood line 8 of the extracorporeal blood circuit 6 to the blood chamber 3 of the dialyzer 2, where it is cleaned of unwanted blood components and returned to patient 7 via the venous blood line 9.

[0046] The dialysis machine has 100 different sensors for monitoring blood treatment. For example, a conductivity sensor 25 can be installed in the dialysis fluid supply line to determine the conductivity of the fresh dialysate. A conductivity sensor 24 can be installed in the dialysis fluid discharge line to determine the conductivity of the used dialysate. The signal from conductivity sensor 24 and the signal from conductivity sensor 25, or a combination of the signals from conductivity sensors 24 and 25, each form an operating parameter monitoring signal.

[0047] One operational parameter monitoring signal based on the combined measurement of the signals from conductivity sensors 24 and 25 is the dialysance. This relates to monitoring blood purification performance, or clearance, which is important in extracorporeal blood therapy. Clearance K is defined as the proportion of blood flow through the blood chamber that is completely cleared of toxins, particularly urea. In practice, dialysance is determined instead of clearance. This measures the permeability of the filter membrane to electrolytes contained in the dialysis fluid. To do this, the concentration of one or more electrolytes in the dialysis fluid is changed in the dialysis fluid circuit upstream of the dialysis fluid chamber 4 using a bolus generator (not shown), and the resulting change in concentration downstream of the dialysis fluid chamber is measured.The changing concentration of one or more electrolytes upstream and downstream of the dialysis fluid chamber affects the conductivity of the dialysis fluid and is measured by the conductivity sensors 25 upstream and 24 downstream of the dialysis fluid chamber 4. Such a method for determining clearance during ongoing treatment is called online clearance monitoring and is disclosed in the applicant's European patent EP 0 911 043, the disclosure of which is incorporated in its entirety into the present application. A commercial method based on this principle is marketed by Fresenius Medical Care Deutschland GmbH under the name OCM (Online Clearance Monitor).

[0048] Sensors in the extracorporeal blood circuit 6 can include an arterial pressure sensor 27 and an arterial temperature sensor 10 in the arterial blood line 8, and a venous pressure sensor 26 and a venous temperature sensor 21 in the venous blood line 9. A power consumption signal is provided by the blood pump 23. The signals from the venous pressure sensor 26, the arterial pressure sensor 27, the venous temperature sensor 21, the arterial temperature sensor 10, and the power consumption signal of the blood pump 23 are each time-dependent and each form an operating parameter monitoring signal. Additionally, a hematocrit sensor (not shown) can be provided in the arterial and / or venous blood line, each providing an operating parameter monitoring signal.

[0049] Operating parameter monitoring signals are transmitted to the central processing unit 20 during ongoing processing, recorded there, and transmitted to the central monitoring unit 40 after completion of the processing. Alternatively, operating parameter monitoring signals are transmitted continuously to the central monitoring unit 40 during ongoing processing. The central monitoring unit 40 may include a database.

[0050] The dialysis machine 100 also features an input unit 30, which allows the user to input information regarding the blood treatment or its progress, or a deviation from an ideal or complication-free treatment course or operation of the dialysis machine. The user response can involve a standard input during the ongoing blood treatment, triggering an alarm or treatment stoppage, or indicating the cause of an alarm. The user input is transmitted to the central processing unit 20 and processed there. During the ongoing treatment or after its completion, a signal containing the user input is transmitted to the central monitoring unit.

[0051] The user input and the corresponding assigned operating parameter monitoring signal are stored in the central monitoring unit and processed according to the conditions associated with Figure 3described data acquisition or learning phase 4001 and / or monitoring phase 4002.

[0052] Figure 2 schematically shows a blood treatment system according to the present teaching.

[0053] The blood treatment system 300 has a large number of blood treatment machines 201, ..., 20n, whose operating parameters are continuously monitored.

[0054] The blood treatment machines 201, ..., 20n each have at least one sensor 211 for monitoring an operating parameter of the blood treatment and for generating a time-dependent sensor signal or a signal derived from it.

[0055] Furthermore, the blood treatment machines 201, ..., 20n each have at least one input / output unit 212, which is adapted to allow user input with regard to the operating status or the course of treatment or the deviation from an ideal or complication-free course of treatment or operating status of a blood treatment carried out with the blood treatment machine.

[0056] Blood treatment machines 201, ..., 20n each have an interface 213 for communication with the central unit 301 and for the transmission of user inputs and sensor signals to the central unit 301.

[0057] The central unit 301 has an interface 312 for communication with the blood treatment machines 201, ..., 20n and for receiving user inputs and sensor signals from the blood treatment machines.

[0058] The central processing unit (CPU) 301 can be implemented as a standalone system or as a distributed system. The CPU can be part of a client / server architecture and / or run as part of a cloud network. The CPU 301 includes a storage unit 311, configured to store multiple user response signals and operational parameter monitoring signals in relation to each other, and an evaluation unit 313, adapted to generate mapping or reference data that correlates a specific user response with an associated operational parameter monitoring signal. The storage unit 311 is further adapted to store the mapping and reference data. The storage unit 311 can be configured as a database. The evaluation unit 313 has access to the mapping and reference data stored in the storage unit 311.

[0059] The evaluation unit 313 is further adapted to receive a specific operating parameter monitoring signal, which was recorded at a specific time by one of the blood treatment machines 201, ..., 20n during the blood treatment, from one of the blood treatment machines, here the blood treatment machine 202, and to compare it with the assignment or reference data in order to derive an operating state, a treatment course, or a deviation from an ideal or complication-free operating state or treatment course at the specific time, or a suitable user reaction to it.

[0060] The central unit 301 is adapted to an operating condition, the course of treatment, or the deviation from an ideal or complication-free operating condition or course of treatment at the specified time or the appropriate user reaction to return to the corresponding blood treatment machine 201, ... , 20n, here to the blood treatment machine 202.

[0061] The blood treatment machine 202 is adapted to respond appropriately to the operating condition, the course of treatment, or the deviation from an ideal or complication-free operating condition or course of treatment at the specified time.

[0062] The blood treatment machine 202 has an output unit 221 for issuing an alarm in response to the derived deviation from an ideal or complication-free operating state or treatment process of the blood treatment machine or the typical user reaction.

[0063] Furthermore, the blood treatment machine 202 has a control unit 223 for controlling the blood treatment machine in response to a derived deviation from an ideal or complication-free operating state or treatment process of the blood treatment machine, or the typical user reaction. In one embodiment, the control unit is adapted to initiate an interruption of the blood treatment in response to the deviation.

[0064] Figure 3 is a flowchart of a disclosed computer-implemented method 4000 for determining an operating state of a blood treatment machine, a treatment course of a blood treatment or the deviation from an ideal or complication-free operating state or treatment course in a blood treatment or a suitable user response thereto.

[0065] The individual steps of the computer-implemented procedure can be determined by the components of the dialysis system 200 (compare Fig. 1 ) or from the components of the blood treatment system 300 (compare Fig. 2 ) are executed.

[0066] The computer-implemented procedure 4000 comprises a data acquisition or learning phase 4001 and a monitoring phase 4002.

[0067] During blood treatment with a blood treatment machine 201, ... 20n, in particular dialysis treatment with a dialysis machine 100, at least one operating parameter of the blood treatment machine 401a is continuously monitored and its time-dependent course of a measurement signal of the monitored operating parameter is stored as an operating parameter monitoring signal 401b. The time-dependent measurement signal of the monitored operating parameter can be a time-dependent signal from a sensor of the blood treatment machine, such as a conductivity signal of the fresh dialysate, a conductivity signal of the used dialysate, a temperature signal of the arterial or venous blood temperature, a pressure signal from the arterial or venous temperature sensor, a power consumption signal of the blood pump, or another time-dependent measurement signal of an operating parameter.

[0068] Regarding the course of treatment, or a deviation from an ideal or complication-free course of treatment or operating condition during blood treatment, a user input of 402 is made.

[0069] User input 402 can represent a user response to a detected situation during ongoing blood treatment, such as reducing the blood flow rate due to an increase in arterial pressure observed by the user. Other examples include reducing the ultrafiltration rate or initiating electrolyte administration due to a detected hypertensive crisis.

[0070] User input 402 can include a manually triggered alarm or a manually triggered treatment stop.

[0071] Alternatively or additionally, an automatically generated alarm or a manually triggered alarm can be accompanied by a prompt to the user to specify the cause of the alarm. This allows the user to be offered a selection from various possible causes for the alarm, for example, in the form of a drop-down menu. In this case, user input 402 includes the specification of the cause of the alarm. User input can be provided by a practitioner or by the patient being treated. In one embodiment, the user input is supplemented by a timestamp indicating the point in time of the user input in relation to the course of treatment, such as the time elapsed since the start of the blood treatment.

[0072] Upon completion of the treatment, the recorded sensor signal, along with one or more user inputs made during the blood treatment, is transmitted from the blood treatment machine to a central monitoring unit 403. Before transmission to the central processing unit, the sensor signal may be subjected to signal processing 401c and / or feature extraction 401d. The signal processing may include trend analysis of the sensor signal, such as long-term trend analysis. In one embodiment, the signal processing 401c includes windowing the signal segment in the time range of the user inputs, or applying a linear or non-linear filter. Examples of a linear operation include a Fourier transform of the signal segment or a wavelet transform.In an alternative embodiment, signal processing 401c and / or feature extraction 401d can be performed on the side of the central monitoring unit. In a further alternative embodiment, the transmission 403 of the sensor signal to the central monitoring unit takes place continuously during the ongoing blood treatment.

[0073] In the central evaluation unit, 411 assignment and reference data are generated, which establish an assignment or correlation between a specific user input and the corresponding sensor signal.

[0074] Advantageously, the sequence of steps 401a to 411 is carried out as a data acquisition or learning phase 4001 to generate the mapping and reference data for a large number of treatments, preferably using a large number of blood treatment machines 201, ..., 20n. In the step 411 generating the mapping and reference data, the mapping between user input and sensor signal can be performed in the form of a learning phase of a learning algorithm, preferably a neural network. In simplified terms, the mapping or reference data establishes a mapping between a specific user input and a typical signal profile of a specific operating parameter (signal pattern) corresponding to that user input. For example, it turns out that in the case of a dislocation of the venous needle, the signal profile of the arterial blood pressure and the signal profile of the venous blood pressure exhibit a characteristic pattern.In this case, the mapping and reference data establish a correlation between this characteristic signal pattern and the deviation from the complication-free treatment course observed by the user in the form of a user input. The correlation can be made by assigning a specific cause to the treatment course, or by initiating an appropriate user response, such as, in the case of venous needle dislocation, the initiation of an immediate treatment stoppage.

[0075] In one embodiment, the mapping or reference data for a particular user input includes a plurality of mapped signal patterns, together with an indication of a frequency with which a particular signal pattern occurs, or a probability distribution over various mapped signal patterns.

[0076] In an alternative embodiment, the assignment and reference data for a given signal pattern comprise a plurality of user responses, preferably combined with a probability distribution over the user responses, wherein the user response for which the highest probability was specified is preferably characterized as the most adequate user response.

[0077] In an alternative embodiment, user input is not, or not exclusively, online during blood treatment. Instead, the recorded sensor data is examined offline by an expert to identify deviations from an ideal or complication-free operating state or treatment course, and / or to indicate a cause for this deviation or an appropriate response. In this embodiment, this indication of the cause of the deviation or the response constitutes the user input, which is then assigned to and / or correlated with the sensor signal to generate assignment or reference data, as described above. In one embodiment, the weight given to a user input by an expert in generating the assignment and reference data can vary depending on the expert's level of experience.

[0078] The mapping and reference data can be used in a subsequent monitoring phase 4002 to monitor an ongoing blood treatment. For the blood treatment being monitored, sensor data of an operating parameter are continuously recorded 421 and transmitted 422 to the central monitoring unit 40, 301, where they are received. In the central monitoring unit, a comparison 412 of the transmitted sensor data, or a portion thereof, is made with the mapping and reference data. If a similarity or correlation of the pattern of the transmitted sensor signal, or a portion thereof, is detected with a typical signal pattern contained in the mapping or reference data, the corresponding user input is derived from this.In one embodiment, a correlation is established between a transmitted sensor signal and a typical signal pattern such that the typical signal pattern exhibits the highest correlation compared to other signal patterns stored in the central monitoring unit.

[0079] In this way, a deviation from an ideal or complication-free treatment course can be detected, and / or its cause or an appropriate user response can be determined. The detected user input or a derived deviation from an ideal or complication-free operating state or treatment course of the blood treatment machine is transmitted to the blood treatment machine 423. The blood treatment machine can respond appropriately to the transmitted user input or the derived deviation from the ideal or complication-free operating state or treatment course 424. Thus, the blood treatment machine can issue an alarm, stop treatment, or adjust treatment parameters, according to the detected user input.

[0080] Figure 4This illustrates the mapping between user input and a signal pattern using the example of a venous and an arterial pressure pattern at the time of venous needle dislocation during blood treatment with a blood treatment machine 100, as used in connection with the Figure 4 is described.

[0081] The time-dependent signal of the venous pressure sensor 26 is labeled 501, and the time-dependent signal of the arterial pressure sensor 27 is labeled 502. Between time T1 and time T2, the venous needle becomes dislodged. At time T0, a user input is received, such as initiating a treatment stop and / or triggering an alarm, advantageously accompanied by a description of the cause of the alarm.

[0082] Based on the user input at time T0, a segment of the venous pressure signal and the arterial pressure signal are generated (windowing) between times T1 and T2. These signal segments of the venous and arterial pressure signals, together with the user input, constitute an assignment or reference datum as defined in this disclosure.

Claims

1. Blood treatment system (300) having at least one blood treatment machine (100, 201, 202, 20..., 20n) for conducting a blood treatment, having at least one sensor (211, 10, 21, 24, 25, 20, 27) for determining a time-dependent progression of an operating parameter during the blood treatment, and having a computer system for identifying an operational state of one of the blood treatment machines (100, 201, 202, 20..., 20n) or of a treatment course of a blood treatment or a deviation from an ideal or complication-free operational state or treatment course, wherein the computer system comprises: a first interface (20, 312) adapted to receive an operational parameter monitoring signal, wherein the operational parameter monitoring signal represents a signal from the sensor for monitoring an operational parameter of the blood treatment machine (100, 201, 202, 20..., 20n) comprising the sensor, the signal being derived from a time-dependent signal and derived signal construction comprising a temporal portion or a time window being cut out of the time-dependent signal of the operational parameter, a second interface (20, 312) adapted to receive a user reaction signal, wherein a user reaction signal represents a signal from a user input in view of the operational state or the treatment course or the deviation from an ideal or complication-free treatment course or operational state of a blood treatment conducted with the blood treatment machine of the blood treatment machines (100, 201, 202, 20..., 20n) comprising the sensor, the user input being found in the portion or time window or in an adjacent portion or time window to the portion or time window and the user reaction signal being an adjustment of a blood flow rate or an ultrafiltration rate, an initiation of electrolyte administration, a determination of a blood pressure crisis, or a determination of a needle dislocation, or a triggering of an alarm or a treatment stop; an evaluation unit (40, 313) adapted to create assignment or reference data that establish a correlation of a specific user reaction signal with an assigned operational parameter monitoring signal, a storage unit (311) configured to store the assignment and reference data, wherein the evaluation unit (40, 313) is furthermore adapted to compare a specific operational parameter monitoring signal with the assignment or reference data in order to derive therefrom a typical user reaction thereto, wherein one of the blood treatment machines (100, 201, 202, 20..., 20n) comprises a control unit (223) for initiating an interruption of the blood treatment in response to a deviation of a typical user reaction at the blood treatment machine comprising the control unit from the derived typical user reaction, and / or wherein one of the blood treatment machines (100, 201, 202, 20..., 20n) is configured to output via a user interface a proposition for the user in relation to implementing the derived typical user reaction, and / or wherein one of the blood treatment machines (100, 201, 202, 20..., 20n) is configured to automatically implement an adjustment of an operational parameter corresponding to the derived typical user reaction, and / or wherein the blood treatment system (300) comprises an output unit (212) for outputting an alarm in response to a deviation of a typical user reaction at one of the blood treatment machines (100, 201, 202, 20..., 20n) from the derived typical user reaction.

2. Blood treatment system (300) according to Claim 1, wherein the evaluation unit (40, 313) is adapted to apply a learning algorithm, in particular a neural network, during the creation of the assignment or reference data.

3. Blood treatment system (300) according to Claim 1, wherein the blood treatment machine (100, 201, 202, 20..., 20n) is a dialysis machine (100) and wherein the sensor (211, 10, 21, 24, 25, 20, 27) is selected from a blood temperature sensor (10, 21), a haematocrit sensor, a blood pressure sensor (20, 27), a conductivity sensor (24, 25) for determining a dialysate conductivity or a sensor for determining the power consumption of an actuator (23), preferably a blood pump (23).

4. Blood treatment system (300) according to any of Claims 1 to 3, wherein an evaluation unit (20) of the blood treatment machine (100, 201, 202, 20..., 20n) is adapted to provide the user input signal with a timestamp specifying the time of user input in relation to the treatment course, and wherein the evaluation unit of the computer system or the evaluation unit of the blood treatment machine is adapted to cut out a temporal portion or a time window from the time-dependent signal, and wherein the time of the user input is in the temporal portion or time window or adjacent thereto.

5. Blood treatment system (300) according to any of Claims 1 to 4, having an input unit (30) for inputting the user input in view of an operational state or a treatment course or in view of a deviation from an ideal or complication-free operational state of the blood treatment machine or a deviation from an ideal or complication-free treatment course of the blood treatment.

6. Computer- implemented method (4000) for identifying an operational state of a blood treatment machine or a treatment course of a blood treatment or a deviation from an ideal or complication-free operational state or treatment course, comprising: receiving an operational parameter monitoring signal (401b), wherein the operational parameter monitoring signal represents a signal of a sensor for monitoring an operational parameter of the blood treatment machine the signal being derived from a time-dependent signal and derived signal construction comprising a temporal portion or a time window being cut out of the time-dependent signal of the operational parameter, receiving a user reaction signal (402), wherein the user reaction signal represents a signal from a user input in view of the operational state or the treatment course or the deviation from an ideal or complication-free treatment course or operational state of a blood treatment conducted with the blood treatment machine, a user input time being found in the portion or time window or adjacent thereto and the user reaction signal being an adjustment of a blood flow rate or an ultrafiltration rate, an initiation of electrolyte administration, a determination of a blood pressure crisis, or a determination of a needle dislocation, or a triggering of an alarm or a treatment stop; storing a plurality of user reaction signals and operational parameter monitoring signals in relation to one another, creating assignment or reference data (411) that establish an assignment or correlation of a specific user reaction signal with an assigned operational parameter monitoring signal in the portion or the time window, storing the assignment and reference data, receiving a specific operational parameter monitoring signal (422), comparing the specific operational parameter monitoring signal with the assignment or reference data (412) in order to derive therefrom a typical user reaction thereto, and by way of a control unit of the blood treatment machine initiating an interruption of the blood treatment in response to a deviation of a typical user reaction at the blood treatment machine comprising the control unit from the derived typical user reaction, and / or outputting via a user interface a proposition for the user in relation to implementing the derived typical user reaction, and / or automatically implementing an adjustment of an operational parameter corresponding to the derived typical user reaction, and / or outputting on an output unit an alarm in response to a deviation of a typical user reaction at one of the blood treatment machines (100, 201, 202, 20..., 20n) from the derived typical user reaction.

7. Computer-implemented method (4000) according to Claim 6, wherein the step of creating the assignment or reference data comprises the following: applying a learning algorithm, in particular a neural network.

8. Computer-implemented method (4000) according to either of Claims 6 and 7, containing the following steps: providing the user input signal with a timestamp specifying the time of the user input in relation to the treatment course, forming a time window or a temporal portion from the time-dependent signal, wherein the time of the user input is found in the time window or the temporal portion or adjacent thereto.

9. Computer program product comprising instructions which when loaded into at least one storage module of a computer system (300) and worked through by at least one processor of the computer system execute the computer-implemented method (4000) according to any of Claims 6 to 8.

10. Computer program comprising instructions for carrying out the computer-implemented method (4000) according to any of Claims 6 to 8 when the computer program is executed on a computer.