SIGNAL REPLAY FOR SELECTION OF OPTIMAL DETECTION SETTINGS

Abstract

A method for adjusting parameters of a medical device includes collecting patient data by using the medical device, in which the medical device includes programmable parameters for affecting a function carried out by the medical device, transmitting the patient data to an external device, conducting an analysis of the transmitted patient data by using the external device, and providing an automatically computed proposal for adjusting at least one parameter, several parameters or all of the parameters based on the analysis. A corresponding system is also provided.

Claims

1. A method for adjusting parameters of a medical device, the method comprising the following steps: collecting patient data of a patient by using the medical device, the medical device including programmable parameters for affecting a function carried out by the medical device; transmitting the patient data to an external device; conducting an analysis of the transmitted patient data by using the external device; and providing an automatically computed proposal to a physician for adjusting at least one of the parameters, or several of said parameters, or all of the parameters based on the analysis.

2. The method according to claim 1, which further comprises using the medical device to carry out the function by executing an algorithm implemented in the medical device, the algorithm utilizing the programmable parameters to achieve a classification of clinical events contained in the patient data.

3. The method according to claim 1, which further comprises carrying out the step of collecting the patient data by using the medical device to measure an ECG of the patient.

4. The method according to claim 1, which further comprises storing the collected patient data in the external device.

5. The method according to claim 1, which further comprises storing the collected patient data in the external device for later review.

6. The method according to claim 1, which further comprises carrying out the step of conducting the analysis by carrying out an analysis algorithm on the external device, applying the analysis algorithm to the collected patient data for a plurality of different settings of each parameter, and for each setting of the parameters storing a classification of the clinical events obtained by the algorithm executed on the external device.

7. The method according to claim 6, which further comprises using a human expert or an expert system to independently process the classifications of the clinical events made by the analysis algorithm executed on the external device to adjudicate them as true or false.

8. The method according to claim 7, which further comprises adapting the automatically computed proposal to reduce false classifications of clinical events made by the analysis algorithm and increase true classifications when the respective parameter is adjusted according to the proposal.

9. The method according to claim 3, wherein the function corresponds to one of: detecting of P, Q, R, S, and T waves in the ECG, detecting of atrial fibrillation in the ECG, detecting of bradycardia in the ECG, detecting of an asystole in the ECG, detecting of a high ventricular rate in the ECG detecting a sudden rate drop in the ECG.

10. The method according to claim 1, which further comprises adjusting the parameters of the medical device by using the external device, or another external device or an external data center by wireless programming.

11. A system for adjusting medical parameters, the system comprising: a medical device configured to collect patient data of a patient, said medical device being configured to perform a function and including programmable parameters for affecting said function; and an external device configured to receive said collected patient data transmitted by said medical device, said external device being configured to conduct an analysis of said transmitted patient data, and said external device being configured to provide a proposal to a physician for adjusting at least one of said parameters, several of said parameters, or all of said parameters using said analysis.

12. The system according to claim 11, wherein said medical device is configured to carry out said function by executing an algorithm implemented in said medical device, said algorithm utilizing said programmable parameters to achieve a classification of clinical events contained in said patient data.

13. The system according to claim 11, wherein said collected patient data is an ECG of the patient.

14. The system according to claim 11, wherein said external device is configured to conduct said analysis by executing said analysis algorithm on said external device and applying said algorithm to said collected patient data for a plurality of different settings of each parameter, said external device being configured to store classifications of clinical events obtained by said algorithm executed on said external device for each setting of said parameters.

15. The system according to claim 14, wherein the system is configured to receive, as an input, adjudications by a human expert or an expert system as to whether said classification of said respective clinical event made by said algorithm executed on said external device is true or false.

16. The system according to claim 11, wherein the system is configured to adapt said automatically computed proposal such that false classifications of clinical events made by said algorithm are reduced when said respective parameter is adjusted according to said proposal.

17. The system according to claim 13, wherein said function corresponds to one of: detecting of P, Q, R, S, and T waves in the ECG, detecting of atrial fibrillation in the ECG, detecting of bradycardia in the ECG, detecting of an asystole in the ECG, detecting of a high ventricular rate in the ECG, detecting of a sudden rate drop in the ECG.

Description

BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWING

[0065] FIG. 1 is a schematic illustration of an embodiment of a system/method according to the present invention;

[0066] FIG. 2 is a chart showing an example of an optimization matrix used to pick ideal sensing parameter settings according to an embodiment of the method of the present invention;

[0067] FIG. 3 is a chart showing an example of arrhythmia assessment across two parameters according to an embodiment of the method of the present invention;

[0068] FIG. 4 is a chart showing multiple snapshot adjudication and selection for optimization according to an embodiment of the method according to the present invention;

[0069] FIG. 5 is a chart showing an exemplary matrix for classification results of a single signal episode (snapshot); and

[0070] FIG. 6 is a chart showing an exemplary matrix for classification results of multiple signal episodes (snapshots).

DETAILED DESCRIPTION OF THE INVENTION

[0071] Referring now to the figures of the drawings in detail and first, particularly, to FIG. 1 thereof, there is seen a schematic illustration of an embodiment of a system 1/method according to the present invention. According to FIG. 1, the system 1 includes a medical device 2 configured to collect patient data 4 of a patient P. The medical device 2 is configured to perform a function and includes programmable parameters for affecting that function. Further, the system 1 includes an external device 3. The medical device 2 is configured to transmit the collected patient data 4 to the external device 3, the external device 3 is further configured to conduct an analysis of the transmitted patient data 4, and the external device 3 is configured to provide a proposal 5 to a physician P for adjusting at least one of the parameters, or several of the parameters, or all of the parameters using the analysis.

[0072] Particularly, the external device 3 can be a programmer that is operable to program the programmable parameters of the medical device 2. The medical device 2 can be a monitoring device that is e.g. configured to monitor a signal 4 of the patient P. Alternatively the medical device 2 can be a cardiac pacemaker or a cardioverter defibrillator.

[0073] Particularly, according to an embodiment, for QRS detection, a real-time electrocardiogram (ECG) is captured by the external device 3 as the patient data 4. The external device 3 can be formed by a physician's programmer. The data 4 can be either directly received from the implant 2 or can be stored in the implant before. Particularly, a QRS detection algorithm that would normally be applied by the device 2 is applied to the captured ECG 4 by the external device 3, using e.g. all variations of the available parameter set. Particularly, a performance metric such as a Receiver Operating Curve (sensitivity versus 1-specificity) allows the clinician P to see which parameter set(s) would be most appropriate for this patient P. The clinician P can select a candidate from this ROC plot directly, or may examine the ECG 4 with detection annotations before selecting a candidate.

[0074] Since the algorithm for conducting data analysis is now running on the external device 3 (e.g. programmer), which does not have the processor and battery limitations of the implant 2, a much wider parameter space can be explored.

[0075] Particularly, a human expert (e.g. clinician/physician) P or an expert system independently processes the classifications of the events made by the algorithm executed on the external device 3 to adjudicate them as true or false. The automatically computed proposal is then adapted such that false classifications of clinical events made by the algorithm are reduced or true classifications are increased when the respective parameter is adjusted according to the proposal.

[0076] Apart from QRS detection described above, the function/algorithm may also be configured to detect one of: atrial fibrillation, bradycardia, asystole, sudden rate drop, or a high ventricular rate.

[0077] Particularly, according to an embodiment of the invention, when the function of the implantable device 2 relates to detecting bradycardia, a false snapshot can be re-evaluated with different bradycardia settings. These settings can be applied to other non-bradycardia snapshots of the measured ECG 4, to ensure that they are not falsely detected. A more sophisticated approach would be to evaluate both bradycardia and QRS detection for optimal settings of QRS detection in the context of bradycardia. For instance, various QRS detection settings can be applied to find the settings where a false bradycardia snapshot is no longer detected as false. The user P can adjudicate the snapshot or adjudicate specific events in order to provide a gold standard for grading the parameters.

[0078] Particularly, the present invention allows for improving performance by leveraging the parameter space of existing algorithms. The human or machine experts P are presented with performance data, so that they can use their domain experience to choose the best parameters for their patient P without going through the drudgery and time of manually selecting each parameter. This minimizes the need for customer support of patients P. By providing a graphical matrix of parameter settings (cf. e.g. FIG. 2 and FIG. 3), along with performance at those settings, the user P can at a glance select settings of the programmable parameters that perform to their expectations. In addition, by selecting multiple snapshots, the algorithm can provide a pooled performance metric that has a more generalized behavior applicable to the patient's signals (cardiac rhythm or ECG morphology (FIG. 4), where the medical device measures cardiac signals). Snapshots can be adjudicated based on review and then selected for optimization to achieve the performance desired for a given subset of snapshots. This achieves a personalized medicine outcome with appropriate settings for either sensing settings or arrhythmia settings based on individual patients P.

[0079] FIG. 5 shows an example for classification of a single signal episode (snapshot) of an ECG according to an embodiment of the invention. The adjustable parameters are sensitivity, which is technically defined by the formula (true positives)/(true positives+false negatives) and RR variability, which measures the variation of time intervals between successive R-peaks in an ECG for a given time window. The matrix has three columns which stand for sensitivity in three levels low, medium and high. and three rows which stand for RR variability also in three levels low, medium and high. The cells of the matrix contain for each possible combination of the two parameter and levels exemplary classification results as detected or not detected for the detection of a clinical event in the snapshot. In this case, a clinical event was detected for 6 of the 9 combinations, while for 3 of the 9 combinations, no clinical event was detected.

[0080] FIG. 6 shows an example for classification according to FIG. 5 for multiple snapshots. The first number of each matrix cell refers to the sensitivity in percent %, the second number to the precision (Positive Prediction Value PPV, defined by the formula (true positives)/(true positives+false positives)) in percent. The user may choose the desired parameter settings according to the indicated sensitivity vs. precision.

[0081] It will be apparent to those skilled in the art that numerous modifications and variations of the described examples and embodiments are possible in light of the above teaching. The disclosed examples and embodiments are presented for purposes of illustration only. Other alternate embodiments may include some or all of the features disclosed herein. Therefore, it is the intent to cover all such modifications and alternate embodiments as may come within the true scope of this invention.