Methods and systems for automatically identifying detection parameters for an implantable medical device

Patent metadata

Field	Value
Publication number	US-9955921-B2
Application number	US-201715654656-A
Country	US
Kind code	B2
Filing date	Jul 19, 2017
Priority date	Nov 27, 2012
Publication date	May 1, 2018
Grant date	May 1, 2018

Abstract

An initial set of parameters for operating one or more detection tools is automatically derived and subsequently adjusted so that each detection tool is more or less sensitive to signal characteristics in a region of interest. Detection tool(s) may be applied to physiological signals sensed from a patient (such as EEG signals) and may be configured to run in an implanted medical device that is programmable with the parameters to look for rhythmic activity, spike activity, and power changes in the sensed signals, etc. A detection tool may be selected and parameter values derived in a logical sequence and/or in pairs based on a graphical representation of an activity type which may be selected by a user, for example, by clicking and dragging on the graphic via a GUI. Displayed simulations allow a user to assess what will be detected with a derived parameter set and then to adjust the sensitivity of the set or start over as desired.

First claim

What is claimed is:

1. A method comprising:
obtaining an electrographic signal of a type capable of being sensed by an active implantable device;
defining a region of interest within the electrographic signal;
processing the region of interest to determine a plurality of parameter values for each of a plurality of different types of detection tools;
for each of the plurality of different types of detection tools:
running a simulation of the type of detection tool on the electrographic signal to obtain one or more detection outputs, and
calculating a plurality of detection metrics based on the one or more detection outputs and one or more characteristics of the electrographic signal; and

processing the detection metrics of each of the plurality of different types of detection tools to select a best type of detection tool to be implemented in the active implantable device.

2. The method of claim 1 , wherein the plurality of different types of detection tools comprises a rhythmic detection tool, a spike detection tool, and a nonlinear energy increase detection tool.

3. The method of claim 1 , wherein each respective plurality of parameter values defines how a corresponding one of the plurality of different types of detection tools operates.

4. The method of claim 1 , wherein the plurality of detection metrics comprises a true positive detection metric and a false positive detection metric, and processing the detection metrics of each of the plurality of different types of detection tools comprises:
ranking the plurality of different types of detection tools from highest to lowest based on their respective true positive detection metrics;
determining if the plurality of different types of detection tools meets a first criterion;
if the first criterion is met, temporarily selecting from the plurality of different types of detection tools, the type of detection tool having the highest true positive detection metric;
determining if the temporarily selected type of detection tool meets a second criterion; and
if the second criterion is met, choosing the temporarily selected type of detection tool as the best type of detection tool.

5. The method of claim 4 , wherein:
the first criterion is that at least one of the different types of detection tools has a false positive detection metric less than a specified value, and
the second criterion is that the temporarily selected type of detection tool is the highest ranked type of detection tool.

6. The method of claim 5 , further comprising:
if the second criterion is not met, determining which one, if any, of the different types of detection tools not corresponding to the temporarily selected type of detection tool meets a third criterion;
if the third criterion is met, choosing the type of detection tool ranked immediately above the determined type of detection tool, as the best type of detection tool; and
if the third criterion is not met, choosing the determined type of detection tool as the best type of detection tool.

7. The method of claim 6 , wherein the third criterion is that each of a plurality of equalities derived from the true positive detection metric and the false positive detection metric of the plurality of different types of detection tools are satisfied.

8. The method of claim 4 , further comprising:
if the first criterion is not met, selecting from the plurality of different types of detection tools, the type of detection tool having the lowest false positive detection metric;
determining if the selected type of detection tool meets a fourth criterion; and
if the fourth criterion is not met, choosing the selected type of detection tool as the best type of detection tool.

9. The method of claim 8 , wherein the fourth criterion is that the true positive detection metric of the selected type of detection tool is less than a specified value.

10. The method of claim 8 , further comprising:
if the fourth criterion is met, selecting from the plurality of different types of detection tools, the type of detection tool having the second lowest false positive detection metric;
determining if the selected type of detection tool meets a fifth criterion;
if the fifth criterion is met, choosing the type of detection tool having the second lowest false positive detection metric as the best type of detection tool; and
if the fifth criterion is not met, choosing the type of detection tool having the lowest false positive detection metric as the best type of detection tool.

11. The method of claim 10 , wherein the fifth criterion is that each of a plurality of equalities derived from the true positive detection metric and the false positive detection metric of the types of detection tools having the lowest false positive detection metric and the second lowest false positive detection metric are satisfied.

12. The method of claim 1 , wherein processing the region of interest to determine a plurality of parameter values for each of a plurality of different types of detection tools is performed subsequent to determining that an activity pattern in the region of interest is undetermined.

13. The method of claim 12 , wherein determining that an activity pattern in the region of interest is undetermined comprises:
computing a plurality of metrics for the region of interest, wherein each of the plurality of metrics corresponds to a value derived from one or more amplitudes of the region of interest, or one or more frequencies of the region of interest; and
applying one or more of the plurality of metrics to a series of rules until an outcome corresponding to an undetermined pattern is reached.

14. The method of claim 1 , further comprising determining final values for a plurality of parameters for a detection tool corresponding to the best type of detection tool.

15. An apparatus comprising:
a memory; and
a processor coupled to the memory and configured to:
obtain an electrographic signal of a type capable of being sensed by an active implantable device;
define a region of interest within the electrographic signal;
process the region of interest to determine a plurality of parameter values for each of a plurality of different types of detection tools;
for each of the plurality of different types of detection tools:
run a simulation of the type of detection tool on the electrographic signal to obtain one or more detection outputs, and
calculate a plurality of detection metrics based on the one or more detection outputs and one or more characteristics of the electrographic signal; and

process the detection metrics of each of the plurality of different types of detection tools to select a best type of detection tool to be implemented in the active implantable device.

16. The apparatus of claim 15 , wherein the plurality of detection metrics comprises a true positive detection metric and a false positive detection metric, and the processor processes the detection metrics of each of the plurality of different types of detection tools by being further configured to:
rank the plurality of different types of detection tools from highest to lowest based on their respective true positive detection metrics;
determine if the plurality of different types of detection tools meets a first criterion;
if the first criterion is met, temporarily select from the plurality of different types of detection tools, the type of detection tool having the highest true positive detection metric;
determine if the temporarily selected type of detection tool meets a second criterion; and
if the second criterion is met, choose the temporari

Assignees

• Neuropace Inc

Inventors

• Esteller Rosana
• Sun Felice

Classifications

• A61N1/36135

  using physiological parameters · CPC title

• A61B5/748

  Selection of a region of interest, e.g. using a graphics tablet · CPC title

• A61B5/7282

  Event detection, e.g. detecting unique waveforms indicative of a medical condition (cough events A61B5/0823; seizures A61B5/4094; sleep apnoea A61B5/4818) · CPC title

• A61B5/7278

  Artificial waveform generation or derivation, e.g. synthesizing signals from measured signals · CPC title

• A61B5/6868

  Brain · CPC title