Advancing Affordable Bionic Limbs: Critical Research Gaps in Deep Learning-Based Transradial Prosthetic Arm Development

Abstract

Background: Upper-limb amputation significantly affects an individual's ability to perform daily activities, creating an increasing demand for intelligent, lightweight, and affordable prosthetic systems. Recent advances in electromyography (EMG), artificial intelligence, deep learning, computer vision, soft robotics, and additive manufacturing have substantially improved prosthetic arm technology. However, challenges related to reliable intention recognition, grasp force regulation, sensory feedback, affordability, and personalization continue to limit widespread clinical adoption. Objective: This review critically analyzes recent developments in transradial prosthetic arm technology, identifies current research gaps, and proposes future directions for developing affordable and intelligent EMG-driven prosthetic systems. Methods: A comprehensive review of recent literature was conducted covering prosthetic classifications, anthropometric requirements, commercial prosthetic devices, material selection, EMG-based control strategies, deep learning techniques, computer vision, force myography, sensory feedback, soft robotics, and rehabilitation technologies. Existing approaches were comparatively analyzed to identify technological limitations and emerging research opportunities. Results: The review indicates that although EMG-based control and deep learning have significantly improved gesture recognition and prosthetic functionality, existing systems still suffer from limitations in adaptive control, sensory feedback, lightweight design, affordability, and real-time performance. Emerging technologies such as multimodal sensing, artificial intelligence, and additive manufacturing demonstrate considerable potential for improving prosthetic functionality; however, their integration into a unified prosthetic control framework remains limited. Conclusion: The review identifies critical research gaps in intelligent prosthetic arm development and highlights the need for integrated frameworks combining multimodal sensing, deep learning, advanced control strategies, and low-cost manufacturing. The presented analysis provides a foundation for future research toward developing affordable, adaptive, and intelligent transradial prosthetic arms capable of enhancing rehabilitation outcomes and improving the quality of life of amputees