Evaluation of misclassification matrix method in validation of an assistive device for manual wheelchair propulsion
Classification accuracy is essential in the bio signal’s performance-based assistive devices. In this study, surface electromyography (SEMG) signals acquisition was extracted from 3 healthy right-handed participants. SEMG signal was processed, and Motor Unit Action Potential (MUAP) was determined. A...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Seventh Sense Research Group
2022
|
| Subjects: | |
| Online Access: | http://umpir.ump.edu.my/id/eprint/40616/ http://umpir.ump.edu.my/id/eprint/40616/1/Evaluation%20of%20misclassification%20matrix%20method%20in%20validation.pdf |
| Summary: | Classification accuracy is essential in the bio signal’s performance-based assistive devices. In this study, surface electromyography (SEMG) signals acquisition was extracted from 3 healthy right-handed participants. SEMG signal was processed, and Motor Unit Action Potential (MUAP) was determined. Accuracy, precision, sensitivity and specificity were calculated in real-time based on individual MUAP, critically compared with pattern and non-pattern recognition control methods by Misclassification Matrix inserted into Arduino MEGA 2560 Microcontroller. The results indicated that the performance of each control method is different for every participant and a comparison tool is a must to select the best out of it. It shows that the misclassification matrix filtered the best control method for participant 1 as Probability Density Function, no for participant 2 and Maximum Point Different (MPD) for participant 3 based on determined conditions. |
|---|