The study implements an algorithm for classifying lower limb movement patterns. This task has significant practical applications in various areas of medical technology, including lower limb prosthetics, rehabilitation and diagnostic equipment in physiology, as well as the development of exoskeletons and assistive devices.
Currently, there are numerous domestic studies in a similar field focusing on upper limbs, which have found widespread use. However, domestic developments for lower limbs are limited to methods based on the "agonist-antagonist" muscle principle, making existing systems incomplete.
KEY FINDINGS OF THE STUDY As a result of the research, a Python-based algorithm for classifying lower limb movement patterns was developed, consisting of:
-
data loading module;
-
preprocessing module (digital filtering);
-
segmentation module;
-
feature extraction module;
-
Z-score-based data normalization;
-
classification module (NN – 92% accuracy).
The movement recognition algorithm presented in this thesis has the potential to become the first medical software of its kind in the Russian Federation for controlling bionic lower limb prostheses.