Arabian Journal for Science and Engineering https://doi.org/10.1007/s13369-020-05044-x RESEARCH ARTICLE-SYSTEMS ENGINEERING Leveraging ANN and LDA Classifiers for Characterizing Different Hand Movements Using EMG Signals Bushra Saeed 1 · Muhammad Zia-ur-Rehman 2 · Syed Omer Gilani 1 · Faisal Amin 3 · Asim Waris 1 · Mohsin Jamil 4 · Muhammad Shafique 2 Received: 26 May 2020 / Accepted: 16 October 2020 © King Fahd University of Petroleum & Minerals 2020 Abstract The analysis of electromyographic (EMG) signals has expedited the use of a wearable prosthetic arm. To this end, pattern recognition-based myoelectric control schemes have shown the promising results; however, the choice of classifier and optimal features is always challenging. This paper presents the comparative analysis of classifiers for multiple EMG datasets including (1) the publicly accessible NinaPro database which provides data recorded for 52 hand movements collected from 27 subjects out of which twelve finger movements were classified, and (2) the data collected from ten healthy and six amputee subjects for 11 different hand movements. The classification results of artificial neural networks (ANN) were compared with those of linear discriminant analysis (LDA) for both datasets separately. For dataset 1, the mean classification accuracy of LDA obtained was 85.41% while ANN showed 91.14% accuracy. Similarly, for dataset 2, the mean classification accuracy achieved with LDA was 93.54% while with ANN, it was 97.69%. Besides, p-values were determined for both datasets which revealed better classification results of ANN as compared to LDA. The overall results of this study show that ANN performed better classification and recognition of hand movements as compared to LDA. The findings of this study offer important insights regarding the selection of classifiers of EMG signals which are critical to evaluating the accurate performance of prosthetic human organs. Keywords Artificial neural networks · Electromyography · Linear discriminant analysis · Bio signals · Prosthesis 1 Introduction The technique used to record the electrical activity gen- erated in the human muscles is called electromyography (EMG). This electrical activity helps control the myoelec- tric prosthetic hand for performing various hand movements. A contracting muscle generates EMG signals which may be collected either with the help of a needle, wire, or sur- B Bushra Saeed bushrasaeed.pg@smme.edu.pk 1 Department of Robotics and Intelligent Machine Engineering, School of Mechanical and Manufacturing Engineering, National University of Sciences and Technology, Islamabad, Pakistan 2 Department Biomedical Engineering, Riphah International University, Islamabad, Pakistan 3 Riphah International University, Islamabad, Pakistan 4 Faculty of Engineering and Applied Science, Memorial University of Newfoundland Libraries, St. John’s, Canada face electrode [1]. Surface electrodes are widely used for acquiring EMG signals by placing these electrodes over the human skin. This technique is called surface electromyog- raphy (sEMG). The EMG signals which are stochastic in nature must be accurately recorded and preprocessed in order to achieve the accurate performance of prosthetic hand [2]. The remarkable achievements made in proposing artificial intelligence algorithms and introducing myoelectric inter- faces have enhanced the application of wearable prosthetic devices such as prosthetic hand which act as an artificial replacement for disabled or missing organs of a human body [3]. EMG signals have been proposed to be utilized in the operation of prosthetic hand since 1948 [4, 5]. The first myoelectric-based clinical prosthetic hand was introduced at Central Prosthetic Research Institute, Moscow, in 1957–1960 [6]. Thereafter, different control methods were proposed by researchers to make advancements in the control of pros- thetic hands. These control methods may be either sequential or simultaneous [3]. Sequential control methods are used in 123