Avoid is a pre-prosthetic serious game developed in Unity, controlled through the Myo Armband.
It is specially designed for people with upper limb amputation, helping them train their residual muscle signals before using a myoelectric prosthesis.

The main goal of the game is to dodge obstacles along an endless track, while promoting motor coordination, muscle training, and rehabilitation.

• The player controls a yellow sphere that moves sideways.
• Movements are performed exclusively using muscle contractions of flexion and extension, captured by the Myo Armband.
• The objective is to avoid collisions with colored blocks that appear on the track.
• The score increases as the player survives longer.

• Unity Engine (game development)
• Myo Armband (flexion and extension EMG signal recognition)
• Python (training the machine learning model and processing EMG signals)
• C# (game logic and integration)

• The Myo Armband captures EMG (electromyographic) signals from the residual limb.
• A Python-based model is trained to detect flexion and extension patterns.
• After training, the processed signals are transmitted in real-time to Unity.
• This enables pre-prosthetic training, preparing amputees for the control of myoelectric prostheses.

Avoid was created as a serious pre-prosthetic game with the following goals:

• Support people with upper limb amputation in training EMG-based control.
• Provide an engaging environment for pre-prosthetic rehabilitation.
• Improve motor coordination, muscle strength, and reaction time.
• Serve as a complementary tool in physiotherapy and rehabilitation programs.

1. Wear the Myo Armband on the residual limb or the healthy arm.
2. Run the Python script to capture and classify flexion and extension muscle contractions.
3. The trained model sends EMG data to Unity in real time.
4. Control the yellow sphere by contracting muscles for flexion or extension to dodge the blocks.
5. Survive as long as possible and achieve the highest score.