Optimizing prediction accuracy for early recurrent lumbar disc herniation with a directional mutation-guided SVM model
Mengxian Jia,Jiaxin Lai,Kan Li,Jiyang Chen,Kelun Huang,Chaohui Ding,Ziwei Fan,Zongjie Yuan,Honglin Teng
DOI: https://doi.org/10.1016/j.compbiomed.2024.108297
IF: 7.7
2024-03-15
Computers in Biology and Medicine
Abstract:Percutaneous endoscopic lumbar discectomy (PELD) is one of the main means of minimally invasive spinal surgery, and is an effective means of treating lumbar disc herniation, but its early recurrence is still difficult to predict. With the development of machine learning technology, the auxiliary model based on the prediction of early recurrent lumbar disc herniation (rLDH) and the identification of causative risk factors have become urgent problems in current research. However, the screening ability of current models for key factors affecting the prediction of rLDH, as well as their predictive ability, needs to be improved. Therefore, this paper proposes a wrapper feature selection classification prediction model based on improved bat algorithm (BDGBA) and support vector machine (SVM). Among them, BDGBA increases the population diversity and improves the population quality by introducing directional mutation strategy and guidance-based strategy, which in turn enables the model to obtain for better feature subsets. In addition, SVM is used as a classifier for the wrapper feature selection method, and its classification prediction results are used as a fitness function for the feature subset. In the proposed classification prediction method, BDGBA is used as an optimizer for feature subset filtering and as an objective function for feature subset evaluation based on the classification results passed to the support vector machine, which improves the interpretability and prediction accuracy of the model. In order to verify the performance of the proposed method, this paper proves the performance of the model through global optimization experiments and prediction experiments on real data sets. The accuracy of the proposed rLDH prediction model is 93.49% and sensitivity is 88.33%. The experimental results show that Level of herniated disk, Modic change, Disk height, Disk length, and Disk width are the key factors for predicting rLDH, and the proposed method is an effective auxiliary diagnosis method.
engineering, biomedical,computer science, interdisciplinary applications,mathematical & computational biology,biology