Investigating cognitive fatigue recovery through mechanical massage and binaural beats: An AI-driven fNIRS study

Cognitive fatigue is a state of reduced mental performance resulting from prolonged periods of cognitive activity. It is characterized by a sense of tiredness that reduces decision-making abilities. To date, there remains a significant gap in classifying cognitive fatigue under the influence of mechanical massage via massage chair and binaural beats brain massage aided by functional Near-Infrared Spectroscopy. Our aim is to explore the impact of mechanical and binaural brain massage on cognitive fatigue recovery whilst carrying out an extensive comparative analysis of the efficacy of the existing Deep Learning (DL) models alongside conventional Machine Learning (ML) models. The experimental paradigm is consisted of two treatments: Treatment A (Control (General Rest) Group) and B (Experimental Group). Real-time data acquisition of 10 test subjects before and after both treatments is being done. Following a meticulous features extraction protocol, a comprehensive set of 8 DL and 8 ML models is utilized, and their performance is evaluated through a comparative analysis. The categorical results unequivocally demonstrate that Temporal Convolutional Network achieves superior performance by outperforming other DL models, boasting a remarkable accuracy of 97% and 96.52% for Treatment A and B, respectively. Likewise, Support Vector Machine with Radial Basis Function overtakes other ML models by yielding 91.00% and 87.50% accuracy for Treatment A and B, respectively. Upon evaluation of models' performance in Brain-Computer Interface application, it's been concluded that mechanical massage along with binaural beats significantly helps to relieve mental fatigue, enhance working memory, and mental vigilance.