Abstract

In the tragic situation when a person loses his or her hand, they are usually faced with only one option if they wish to regain a good level of mobility; learn to control an artificial hand. It has been suggested that our brain stores a "body map" of the different parts in our body. Thus, if a person loses a hand, their "body map" remains intact and produces phantom sensations that permit the person to feel like they still have their hand. Some discomfort is felt during these sensations; nevertheless, there is a positive side to them as they enable patients to control prosthetic replacements. Sensations experienced can be measured using a method known as Electromyography (EMG) and can be acquired and processed to control an artificial hand. This research involved the acquisition, analysis and classification of EMG signals through construction of a recording device and the development of classification models based on heuristic approaches and Artificial Intelligence classifiers based on Neural Networks to control artificial hands.