eCAP-Based Autoencoder Detection of Implantable Lead Movement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The movement of leads relative to anatomical elements, such as nerves in the spine, can negatively impact patient satisfaction and the effectiveness of neuromodulation therapy by causing diminished treatment efficacy and increased patient pain.
Innovation Solution
A system utilizing an auto-encoder neural network to analyze evoked Compound Action Potentials (eCAPs) to detect lead movement by training on a first set of data and classifying lead movement based on a growth curve loss waveform threshold, enabling automatic notification of lead movement to physicians or automated programming routines.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If lead positioning is maintained manually without automated monitoring, then device complexity is reduced, but lead movement detection capability is lost resulting in diminished treatment efficacy
Solution Approach 1:
The system uses the lead's own eCAP signals for self-diagnosis of lead movement. The auto-encoder neural network is trained on eCAP data from the specific lead configuration and automatically detects deviations indicating lead movement, allowing the system to self-monitor without external intervention or complex additional sensors.
Solution Approach 2:
The patent replaces manual mechanical monitoring of lead position with an automated computational system. Instead of physical checks or complex mechanical sensors, the system uses machine learning algorithms that analyze electrical signals (eCAPs) to detect lead movement, substituting mechanical monitoring with intelligent signal processing.
2Measurement precision
If automated lead movement detection is implemented, then early detection capability is improved, but device complexity increases due to neural network requirements
Solution Approach 1:
The auto-encoder neural network creates a computational model (copy) of the expected eCAP signal pattern when the lead is properly positioned. By comparing actual eCAP signals against this learned model, the system can detect deviations indicating lead movement with high accuracy while keeping the implementation relatively simple through data-driven modeling rather than complex rule-based systems.
3Ease of operation
If lead movement is detected early, then patient satisfaction is improved through timely intervention, but loss of time for therapy adjustment occurs
Solution Approach 1:
The system performs preliminary detection of lead movement continuously in the background, so that when movement occurs, it is immediately identified and can be addressed before significantly impacting patient outcomes. This allows for proactive rather than reactive lead management, reducing the overall time lost to therapy adjustments by catching issues early.
Data Source
AI summary
A method according to at least one embodiment of the present disclosure includes receiving a first set of data including information about evoked Compound Action Potentials (eCAPs), the first set of data generated by an electrical lead; and training, using the first set of data, an auto-encoder neural network. The auto-encoder neural network may be used in a device with an implantable electrical lead. The auto-encoder neural network may receive information collected by the device, analyze growth curve waveforms, and determine, based on the growth curve waveform analysis, whether or not the implantable electrical lead has moved.