Stimulus Artifact Removal in Neuronal Recordings
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for removing stimulus artifacts from recordings of neuronal action potentials in cochlear implants are inefficient, often requiring increased measurement time and relying on assumptions that do not always hold, such as phase-invariance of action potentials or linear scaling of artifacts with amplitude.
Innovation Solution
A method that processes waveform signals by deriving an optimized electrical stimulation signal based on a cost function comparison between stimulus artifact and neuronal action potential waveforms, using a gradient descent procedure and source separation algorithms to separate the stimulus artifact from the neuronal action potential, thereby improving signal-to-noise ratio without increasing measurement time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If alternating stimulation approach is used to average out stimulus artifact, then stimulus artifact removal is improved, but measurement time increases
Solution Approach 1:
The patent applies preliminary action by measuring the stimulus artifact template before the actual NAP measurement. The artifact template is obtained through preliminary measurements at different stimulation levels, then stored for subsequent subtraction from the NAP signal. This preliminary measurement approach allows the actual NAP measurement to proceed without requiring alternating stimulation sequences, thereby reducing measurement time while maintaining artifact removal effectiveness.
2Measurement precision
If masker probe method is used to measure artifact template, then stimulus artifact removal is improved, but measurement time increases
Solution Approach 1:
The patent measures the stimulus artifact template in advance at different stimulation levels and stores these templates for later use. This preliminary measurement of the artifact characteristics allows the actual NAP measurements to proceed efficiently without requiring additional masker probe pulses during the measurement sequence, thus avoiding the time penalty associated with real-time artifact template acquisition.
3Measurement precision
If triphasic pulse is applied to reduce stimulus artifact, then stimulus artifact removal is improved, but recording delay increases
Solution Approach 1:
The patent extracts the stimulus artifact component from the recorded signal by measuring it separately at different stimulation levels and storing it as a template. This extracted artifact template is then subtracted from the NAP signal, allowing the use of simple biphasic stimulation pulses without requiring triphasic pulses. This approach eliminates the recording delay inherent in triphasic pulse methods while achieving effective artifact reduction.
4Measurement precision
If sub-threshold level recording is used to measure artifact template, then stimulus artifact removal is improved, but measurement accuracy deteriorates due to non-linear scaling
Solution Approach 1:
The patent measures the stimulus artifact template at different stimulation levels (amplitudes) and uses these multiple measurements to characterize the artifact's behavior. By collecting artifact data across a range of parameters rather than relying on linear scaling from a single sub-threshold measurement, the method accurately captures the non-linear relationship between stimulation amplitude and artifact magnitude, enabling precise artifact removal at supra-threshold levels.
Data Source
Figure 1
Figure 2
AI summary
A method is described for processing a waveform signal containing a stimulus artifact and one or more neuronal action potentials. An electrical stimulation signal is derived based on satisfying a cost function comparison with a plurality of known neuronal action potential waveforms.