Scholarlink

<p>Herniation of the intervertebral disc, also known as a slipped or ruptured disc, comprises a tear in the resilient annulus fibrosus, allowing the soft inner nucleus pulposus to bulge out and cause symptoms that may include pain, weakness, or numbness in the lower back and legs. This is the pathology that usually arises either due to age, traumatic injury, or biomechanical stress and, therefore usually creates debilitating symptoms highly affecting a person’s daily function. Machine learning (ML) has become one of the most powerful computational tools in the modern medical world for the prediction and diagnosis of different medical conditions, including disc herniation. The present study utilizes two machine learning models, namely Random Forest Classification (RFC) and Naive Bayes Classification (NBC), along with the Golf Optimization Algorithm (GOA). They are trained with much care using an extensive dataset comprising patient data, including their case history and diagnostic imaging results. The empirical study results highlight some interesting facts about the model performance. Especially, the hybrid models combining RFC with GOA and NBC with GOA perform better in predictive performance concerning disc herniation. These results indicate the superior performance of the RFC model and its hybrid version, RFGO, over those of the NBC model and its hybrid counterparts. Concretely, the RFC model achieved as high as 0.915 accuracy in the test phase, compared to 0.851 of the NBC model and 0.904 of the NBGO model. The finding underlines the discriminative effectiveness of the RFC model and its hybrid counterpart, RFGO, in correctly predicting disc herniation, confirming their usefulness within the framework of predictive modeling in the study.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Novel Disc Herniation Prediction Utilizing the Power of Random Forest Base Predictive Models

  • 期刊论文

  • 发表日期：

    2025年2月3日

查看全文
  • Peng Li,
  • Yipin Wang

摘要

Herniation of the intervertebral disc, also known as a slipped or ruptured disc, comprises a tear in the resilient annulus fibrosus, allowing the soft inner nucleus pulposus to bulge out and cause symptoms that may include pain, weakness, or numbness in the lower back and legs. This is the pathology that usually arises either due to age, traumatic injury, or biomechanical stress and, therefore usually creates debilitating symptoms highly affecting a person’s daily function. Machine learning (ML) has become one of the most powerful computational tools in the modern medical world for the prediction and diagnosis of different medical conditions, including disc herniation. The present study utilizes two machine learning models, namely Random Forest Classification (RFC) and Naive Bayes Classification (NBC), along with the Golf Optimization Algorithm (GOA). They are trained with much care using an extensive dataset comprising patient data, including their case history and diagnostic imaging results. The empirical study results highlight some interesting facts about the model performance. Especially, the hybrid models combining RFC with GOA and NBC with GOA perform better in predictive performance concerning disc herniation. These results indicate the superior performance of the RFC model and its hybrid version, RFGO, over those of the NBC model and its hybrid counterparts. Concretely, the RFC model achieved as high as 0.915 accuracy in the test phase, compared to 0.851 of the NBC model and 0.904 of the NBGO model. The finding underlines the discriminative effectiveness of the RFC model and its hybrid counterpart, RFGO, in correctly predicting disc herniation, confirming their usefulness within the framework of predictive modeling in the study.