The assessment of dyskinesia in Parkinson's disease (PD) based on Artificial Intelligence technology is a significant and challenging task. At present, doctors usually use MDS-UPDRS scale to assess the severity of patients. This method is time-consuming and laborious, and there are subjective differences. The evaluation method based on sensor equipment is also widely used, but this method is expensive and needs professional guidance, which is not suitable for remote evaluation and patient self-examination. In addition, it is difficult to collect patient data in medical research, so it is of great significance to find an objective and automatic assessment method for Parkinson's dyskinesia based on small samples.
In this study, we design an automatic evaluation method combining manual features and convolutional neural network (CNN), which is suitable for small sample classification. Based on the finger tapping video of Parkinson's patients, we use the pose estimation model to obtain the action skeleton information and calculate the feature data. We then use the 5-folds cross validation training model to achieve optimum trade-of between bias and variance, and finally make multi-class prediction through fully connected network (FCN).
Our proposed method achieves the current optimal accuracy of 79.7% in this research. We have compared with the latest methods of related research, and our method is superior to them in terms of accuracy, number of parameters and FLOPs.
The method in this paper does not require patients to wear sensor devices, and has obvious advantages in remote clinical evaluation. At the same time, the method of using motion feature data to train CNN model obtains the optimal accuracy, effectively solves the problem of difficult data acquisition in medicine, and provides a new idea for small sample classification.