Improved anterior cruciate ligament tear diagnosis using gated recurrent unit networks and Hybrid Tasmanian Devil Optimization

Chan S., Zhang M., Zhi Y., Razmjooy S., El-Sherbeeny A. M., Lin L.

Biomedical Signal Processing and Control, vol.95, 2024 (SCI-Expanded) identifier

  • Publication Type: Article / Article
  • Volume: 95
  • Publication Date: 2024
  • Doi Number: 10.1016/j.bspc.2024.106309
  • Journal Name: Biomedical Signal Processing and Control
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Compendex, EMBASE, INSPEC
  • Keywords: Anterior Cruciate Ligament (ACL), Gated Recurrent Unit (GRU) networks, Hybrid Tasmanian Devil Optimization (HTDO), Knee joints, MRI images
  • Ankara Yıldırım Beyazıt University Affiliated: No

Abstract

A tear named anterior cruciate ligament (ACL) is a frequently occurring sports injury that can lead to long-term damage to the knee and impair athletic performance. MRI (Imaging of Magnetic Resonance) is a non-invasive detection technique that is utilized to observe the internal composition of the knee joint to identify ACL tears. However, accurate identification of ACL tears from MRI images presents a difficult challenge that requires skilled expertise and extensive manual examination. In this study, a new strategy for automatically detecting ACL tears in MRI images has been proposed. This method makes use of Gated Recurrent Unit (GRU) networks and a Hybrid Tasmanian Devil Optimization (HTDO). The GRU network, derived from the RNN (Recurrent Neural Network), can analyze sequential data and detect temporal relationships. Moreover, the HTDO algorithm is a combined optimization algorithm that integrates the strengths of two optimization techniques: Tasmanian Devil Optimization (TDO) and Particle Swarm Optimization (PSO). The proposed approach has been tested on an MRI dataset of knee joints with and without ACL injuries. The findings demonstrated that the proposed approach outperformed other techniques in identifying ACL tears from MRI images. The suggested approach could be useful in sports medicine, as early diagnosis of ACL injuries could help prevent long-term knee problems and improve athletic performance.