Uneven Terrain Recognition Using Neuromorphic Haptic Feedback

Area of Science:

• Biomedical Engineering
• Neuroscience
• Rehabilitation Technology

Background:

• Technological advancements in prosthetics have improved quality of life for amputees.
• Current lower limb prostheses lack sensory feedback regarding foot-ground interaction, especially terrain irregularities.
• This deficit increases fall risk, particularly on uneven surfaces.

Purpose of the Study:

• To develop and assess a biomimetic vibrotactile feedback system for conveying gait and terrain information to users.
• To investigate the human ability to decode vibrotactile feedback for terrain discrimination.
• To explore the temporal dynamics of human perception in processing this sensory information.

Main Methods:

• A biomimetic vibrotactile feedback system was integrated with a sensor-equipped insole.
• The system was tested on intact subjects who experienced even and uneven terrains.
• A K-Nearest Neighbors (KNN) classifier was employed to analyze the decoding of terrain features from the feedback.

Main Results:

• Subjects could discriminate between even and uneven terrains with 87.5% accuracy, relying solely on vibrotactile feedback.
• Analysis indicated that human subjects processed this feedback with a temporal dynamic of approximately 45 milliseconds.
• The KNN classifier demonstrated the potential for recognizing terrain types based on the feedback patterns.

Conclusions:

• The vibrotactile feedback system shows promise in providing crucial foot-sole information to lower-limb amputees.
• This technology can help users perceive floor conditions, adapt their gait, and enhance confidence in using prosthetic limbs.
• Further development could significantly reduce fall risks and improve mobility for individuals with lower limb amputations.