Evaluation of Natural Hand Movements and Grasp Force in Prosthetic Hands: A Systematic Review

Abstract

Traditional myoelectric prostheses remain limited to only one or two functional grip patterns and are prohibitively expensive, rendering them inaccessible to most amputees in developing countries. The biggest challenge for prosthetic engineers remains how to make an artificial hand move naturally and squeeze with exactly the right force for whatever object is in front of it. This systematic review searched five major databases (Medline, PubMed, Embase, Scopus, Cochrane) from 2010 to 2025 and included 47 studies that met predefined criteria. We examined how modern prosthetic hands read user intent from residual-muscle surface electromyography, translate it into smooth proportional grip strength—often aided by embedded pressure sensors—and how far we have really come toward natural control. Deep learning approaches, particularly convolutional neural networks, generally outperform traditional machine learning methods in motion classification and force estimation among able-bodied subjects, although performance remains substantially lower in amputees and traditional algorithms can achieve higher peak accuracy in specific settings. In actual amputees, however, performance drops sharply because residual signals are weak, noisy, and highly individual. Adding recurrent layers (Long Short-Term Memory), cheap inertial sensors (inertial measurement units), or vision dramatically narrows that gap and improves robustness. Overall, this systematic review shows that deep learning plus multimodal sensing is finally pushing prosthetic hands toward intuitive daily-life use, although major challenges in signal quality, training burden, and cost still stand in the way of truly seamless embodiment.



Received: 12 March 2026 | Revised: 19 May 2026 | Accepted: 15 June 2026



Conflicts of Interest

The authors declare that they have no conflicts of interest to this work.



Data Availability Statement

Data sharing is not applicable to this article as no new data were created or analyzed in this study.



Author Contribution Statement

Krishnakumar Sankar: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data curation, Writing – original draft, Writing – review & editing, Visualization, Supervision, Project administration. Ayra Afreen: Methodology, Software, Investigation, Visualization. Akshaay Sathish Kumar: Methodology, Software, Investigation, Visualization.