Figure 4 - available via license: Creative Commons Attribution 4.0 International
Content may be subject to copyright.
Source publication
Background
Brain-computer interfaces (BCIs) can restore communication in movement- and/or speech-impaired individuals by enabling neural control of computer typing applications. Single command “click” decoders provide a basic yet highly functional capability.
Methods
We sought to test the performance and long-term stability of click-decoding using...
Contexts in source publication
Context 1
... all repeated samples and normalizing the resulting mean values between 0 and 1. We repeated this process using HG features from all channels except one (channel 112) and again by using features from a subset of 12 electrodes over cortical hand-knob (anatomically determined as channels 92, 93, 94, 100, 101, 102, 108, 109, 110, 116, 117, 118; Fig. 4e, Supplementary Figs. 2,3). Neither of these two model architectures were deployed for real-time BCI ...
Context 2
... assess which channels produced the most important HG features for classication of attempted grasp, we generated a saliency map across all channels used to train our original model (Fig. 4a). As expected, channels covering cortical face region were generally not salient for grasp classication. The channel producing the most salient HG features was located in the upper-limb area of somatosensory cortex (channel 112, Supplementary Fig. 16), with a saliency value 55% and 88% higher than the next two most salient channels ...
Context 3
... the corresponding oine classication accuracy of our original model architecture for comparison to a model architecture without channel 112 and an architecture using channels only over cortical hand-knob (see Methods: Channel contributions and oine classication comparisons); the mean accuracy from repeated 10-fold cross-validation (CV) was 92.9% (Fig. ...
Context 4
... ensure that real-time classication accuracy was not entirely driven by channel 112, we evaluated a model trained on HG features from all other channels oine. As expected, this model relied strongly on channels covering the cortical hand-knob region (Fig. 4c), and notably was not as dependent on a single channel; the saliency of the most important channel was only 23% and 60% larger than the next two most salient channels, respectively (Supplementary Fig. 16b). The oine mean classication accuracy from repeated 10-fold CV was 91.7% (Fig. 4d), which was not signicantly lower compared to the ...
Context 5
... strongly on channels covering the cortical hand-knob region (Fig. 4c), and notably was not as dependent on a single channel; the saliency of the most important channel was only 23% and 60% larger than the next two most salient channels, respectively (Supplementary Fig. 16b). The oine mean classication accuracy from repeated 10-fold CV was 91.7% (Fig. 4d), which was not signicantly lower compared to the mean accuracy using all channels (P = 0.139, Wilcoxon Rank-Sum test with 3-way Bonferroni-Holm correction, Fig. ...
Context 6
... was only 23% and 60% larger than the next two most salient channels, respectively (Supplementary Fig. 16b). The oine mean classication accuracy from repeated 10-fold CV was 91.7% (Fig. 4d), which was not signicantly lower compared to the mean accuracy using all channels (P = 0.139, Wilcoxon Rank-Sum test with 3-way Bonferroni-Holm correction, Fig. ...
Context 7
... channels covering the cortical hand-knob region made relatively larger contributions to decoding results, we investigated the classication accuracy of a model trained on HG features from a subset of electrodes covering only this region (Fig. 4e). Saliency values followed a atter distribution; the saliency of the most important channel was only 21% and 44% larger than the next two most salient channels respectively ( Supplementary Fig. 16c). Though the oine mean classication accuracy from repeated 10-fold CV remained high at 90.4% (Fig. 4f), it was statistically lower compared ...
Context 8
... a subset of electrodes covering only this region (Fig. 4e). Saliency values followed a atter distribution; the saliency of the most important channel was only 21% and 44% larger than the next two most salient channels respectively ( Supplementary Fig. 16c). Though the oine mean classication accuracy from repeated 10-fold CV remained high at 90.4% (Fig. 4f), it was statistically lower compared to the mean accuracy using all channels (P = 0.015, Wilcoxon Rank-Sum test with 3-way Bonferroni-Holm correction, Fig. 4g). This suggests that a model trained on HG features from only the cortical hand-knob could still produce effective click detection, but parameters used for data labeling, model ...
Context 9
... and 44% larger than the next two most salient channels respectively ( Supplementary Fig. 16c). Though the oine mean classication accuracy from repeated 10-fold CV remained high at 90.4% (Fig. 4f), it was statistically lower compared to the mean accuracy using all channels (P = 0.015, Wilcoxon Rank-Sum test with 3-way Bonferroni-Holm correction, Fig. 4g). This suggests that a model trained on HG features from only the cortical hand-knob could still produce effective click detection, but parameters used for data labeling, model training, and post-processing may need to be more thoroughly explored to optimize click ...
Similar publications
Brain‐computer interfaces (BCIs) can be used to control assistive devices by patients with neurological disorders like amyotrophic lateral sclerosis (ALS) that limit speech and movement. For assistive control, it is desirable for BCI systems to be accurate and reliable, preferably with minimal setup time. In this study, a participant with severe dy...
Background
Brain-computer interfaces (BCIs) can restore communication for movement- and/or speech-impaired individuals by enabling neural control of computer typing applications. Single command click detectors provide a basic yet highly functional capability.
Methods
We sought to test the performance and long-term stability of click decoding using...
Citations
Background
Brain-computer interfaces (BCIs) can restore communication for movement- and/or speech-impaired individuals by enabling neural control of computer typing applications. Single command click detectors provide a basic yet highly functional capability.
Methods
We sought to test the performance and long-term stability of click decoding using a chronically implanted high density electrocorticographic (ECoG) BCI with coverage of the sensorimotor cortex in a human clinical trial participant (ClinicalTrials.gov, NCT03567213) with amyotrophic lateral sclerosis. We trained the participant’s click detector using a small amount of training data (<44 min across 4 days) collected up to 21 days prior to BCI use, and then tested it over a period of 90 days without any retraining or updating.
Results
Using a click detector to navigate a switch scanning speller interface, the study participant can maintain a median spelling rate of 10.2 characters per min. Though a transient reduction in signal power modulation can interrupt usage of a fixed model, a new click detector can achieve comparable performance despite being trained with even less data (<15 min, within 1 day).
Conclusions
These results demonstrate that a click detector can be trained with a small ECoG dataset while retaining robust performance for extended periods, providing functional text-based communication to BCI users.
Background
To enhance the information transfer rate (ITR) of a steady-state visual evoked potential (SSVEP)-based speller, more characters with flickering symbols should be used. Increasing the number of symbols might reduce the classification accuracy. A hybrid brain-computer interface (BCI) improves the overall performance of a BCI system by taking advantage of two or more control signals. In a simultaneous hybrid BCI, various modalities work with each other simultaneously, which enhances the ITR.
Methods
In our proposed speller, simultaneous combination of electromyogram (EMG) and SSVEP was applied to increase the ITR. To achieve 36 characters, only nine stimulus symbols were used. Each symbol allowed the selection of four characters based on four states of muscle activity. The SSVEP detected which symbol the subject was focusing on and the EMG determined the target character out of the four characters dedicated to that symbol. The frequency rate for character encoding was applied in the EMG modality and latency was considered in the SSVEP modality. Online experiments were carried out on 10 healthy subjects.
Results
The average ITR of this hybrid system was 96.1 bit/min with an accuracy of 91.2%. The speller speed was 20.9 char/min. Different subjects had various latency values. We used an average latency of 0.2 s across all subjects. Evaluation of each modality showed that the SSVEP classification accuracy varied for different subjects, ranging from 80% to 100%, while the EMG classification accuracy was approximately 100% for all subjects.
Conclusions
Our proposed hybrid BCI speller showed improved system speed compared with state-of-the-art systems based on SSVEP or SSVEP-EMG, and can provide a user-friendly, practical system for speller applications.
