Convolutional neural network model-based prediction of human muscle activity by analyzing urine in body fluid using Raman spectroscopy

In recent years, with the popularization of the concept of exercise, the determination of fatigue state during exercise in order to achieve the purpose of scientific exercise has become an important research topic. The concentration of urea in urine fluctuates with the change in exercise intensity, so it is widely used as a biochemical indicator for judging sports fatigue. In this paper, a method combining Raman spectroscopy and convolutional neural network is proposed for quantitative analysis of urea in urine. Averaged spectra are combined with the baseline correction of Raman spectra, an approach that significantly improves the quality of the data and further enhances the prediction accuracy of the model. Finally, in the actual quantitative analysis of urine urea, it demonstrated not only high efficiency and simplicity but also very high stability compared with the traditional optical colorimetric method. Thus, it provides a basis for the rapid and accurate assessment of muscle fatigue.