{"title":"Deep Learning-Driven MRI analysis for accurate diagnosis and grading of lumbar spinal stenosis","authors":"Hasan Genç , Ebubekir Seyyarer , Faruk Ayata","doi":"10.1016/j.measurement.2025.117294","DOIUrl":null,"url":null,"abstract":"<div><div>In recent years, deep neural networks (DNN) have emerged as an important solution due to the increasing complexity of healthcare data. Machine learning (ML) algorithms provide effective and powerful analytical methods that can uncover hidden patterns and important information from large healthcare data sets that cannot be detected in a reasonable time frame using traditional methods. Deep learning (DL) techniques have shown promise in areas such as pattern recognition and diagnosis in healthcare systems. This study aims to contribute to easier interpretation of medical data by applying different DL algorithms to MRI images of the lumbar spine collected between 2020and 2023 in a private clinic. In this context, Convolutional Neural Network (CNN) variations, EfficientNET models and methods such as k-fold cross-validation for more acceptable results, early stopping to save time and Genetic Algorithm (GA) to optimize hyperparameters are preferred. As a result of the study, success rates between 61% and 83.25% are achieved with CNN and between 86.25% and 91.56% with EfficientNET. Overall, this study aims to support medical professionals by mitigating some of the challenges in diagnosis and classification caused by image complexity when interpreting medical data.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"251 ","pages":"Article 117294"},"PeriodicalIF":5.2000,"publicationDate":"2025-03-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Measurement","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0263224125006530","RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, MULTIDISCIPLINARY","Score":null,"Total":0}
引用次数: 0
Abstract
In recent years, deep neural networks (DNN) have emerged as an important solution due to the increasing complexity of healthcare data. Machine learning (ML) algorithms provide effective and powerful analytical methods that can uncover hidden patterns and important information from large healthcare data sets that cannot be detected in a reasonable time frame using traditional methods. Deep learning (DL) techniques have shown promise in areas such as pattern recognition and diagnosis in healthcare systems. This study aims to contribute to easier interpretation of medical data by applying different DL algorithms to MRI images of the lumbar spine collected between 2020and 2023 in a private clinic. In this context, Convolutional Neural Network (CNN) variations, EfficientNET models and methods such as k-fold cross-validation for more acceptable results, early stopping to save time and Genetic Algorithm (GA) to optimize hyperparameters are preferred. As a result of the study, success rates between 61% and 83.25% are achieved with CNN and between 86.25% and 91.56% with EfficientNET. Overall, this study aims to support medical professionals by mitigating some of the challenges in diagnosis and classification caused by image complexity when interpreting medical data.
期刊介绍:
Contributions are invited on novel achievements in all fields of measurement and instrumentation science and technology. Authors are encouraged to submit novel material, whose ultimate goal is an advancement in the state of the art of: measurement and metrology fundamentals, sensors, measurement instruments, measurement and estimation techniques, measurement data processing and fusion algorithms, evaluation procedures and methodologies for plants and industrial processes, performance analysis of systems, processes and algorithms, mathematical models for measurement-oriented purposes, distributed measurement systems in a connected world.