Development and evaluation of a 3D ensemble framework for automatic diagnosis of early osteonecrosis of the femoral head based on MRI: a multicenter diagnostic study.

IF 1.6 4区医学 Q2 SURGERY

Frontiers in Surgery Pub Date : 2025-02-14 eCollection Date: 2025-01-01 DOI:10.3389/fsurg.2025.1555749

Miao Yang, Fuchou Hsiang, Chengfan Li, XiaoYi Chen, Changqing Zhang, Guangchen Sun, Qiliang Lou, Wenhui Zhu, Hongtao Zhao, Feng Liu, Xuehai Ding, Jun Xu

{"title":"Development and evaluation of a 3D ensemble framework for automatic diagnosis of early osteonecrosis of the femoral head based on MRI: a multicenter diagnostic study.","authors":"Miao Yang, Fuchou Hsiang, Chengfan Li, XiaoYi Chen, Changqing Zhang, Guangchen Sun, Qiliang Lou, Wenhui Zhu, Hongtao Zhao, Feng Liu, Xuehai Ding, Jun Xu","doi":"10.3389/fsurg.2025.1555749","DOIUrl":null,"url":null,"abstract":"Background: Efficient and reliable diagnosis of early osteonecrosis of the femoral head (ONFH) based on MRI is crucial for the formulation of clinical treatment plans. This study aimed to apply artificial intelligence (AI) to achieve automatic diagnosis and visualization of early ONFH, thereby improving the success rate of hip-preserving treatments.Method: This retrospective study constructed a multicenter dataset using MRI data of 381 femoral heads from 209 patients with ONFH collected from four institutions (including 239 early ONFH cases and 142 non-ONFH cases). The dataset was divided into training, validation, and internal and external test datasets. This study developed a 3D ensemble framework to automatically diagnose early osteonecrosis of the femoral head based on MRI and utilized 3D Grad-CAM to visualize its decision-making process. Finally, the diagnostic performance of the framework was experimentally evaluated on the MRI dataset and compared with the diagnostic results of three orthopedic surgeons.Results: On the internal test dataset, the 3D-ONFHNet framework achieved overall diagnostic performance with an accuracy of 93.83%, sensitivity of 89.44%, specificity of 95.56%, F1-score of 87.67%, and AUC of 95.41%. On the two external test datasets, the framework achieved overall diagnostic accuracies of 87.76% and 87.60%, respectively. Compared to three orthopedic surgeons, the diagnostic performance of the 3D-ONFHNet framework was comparable to that of senior orthopedic surgeons and superior to that of junior orthopedic surgeons.Conclusions: The framework proposed in this study can generate staging results for early ONFH and provide visualizations of internal signal changes within the femoral head. It assists orthopedic surgeons in screening for early ONFH on MRI in a clinical setting, facilitating preoperative planning and subsequent treatment strategies. This framework not only enhances diagnostic efficiency but also offers valuable diagnostic references for physicians.","PeriodicalId":12564,"journal":{"name":"Frontiers in Surgery","volume":"12 ","pages":"1555749"},"PeriodicalIF":1.6000,"publicationDate":"2025-02-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11868283/pdf/","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Frontiers in Surgery","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.3389/fsurg.2025.1555749","RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2025/1/1 0:00:00","PubModel":"eCollection","JCR":"Q2","JCRName":"SURGERY","Score":null,"Total":0}

引用次数: 0

Abstract

Background: Efficient and reliable diagnosis of early osteonecrosis of the femoral head (ONFH) based on MRI is crucial for the formulation of clinical treatment plans. This study aimed to apply artificial intelligence (AI) to achieve automatic diagnosis and visualization of early ONFH, thereby improving the success rate of hip-preserving treatments.

Method: This retrospective study constructed a multicenter dataset using MRI data of 381 femoral heads from 209 patients with ONFH collected from four institutions (including 239 early ONFH cases and 142 non-ONFH cases). The dataset was divided into training, validation, and internal and external test datasets. This study developed a 3D ensemble framework to automatically diagnose early osteonecrosis of the femoral head based on MRI and utilized 3D Grad-CAM to visualize its decision-making process. Finally, the diagnostic performance of the framework was experimentally evaluated on the MRI dataset and compared with the diagnostic results of three orthopedic surgeons.

Results: On the internal test dataset, the 3D-ONFHNet framework achieved overall diagnostic performance with an accuracy of 93.83%, sensitivity of 89.44%, specificity of 95.56%, F1-score of 87.67%, and AUC of 95.41%. On the two external test datasets, the framework achieved overall diagnostic accuracies of 87.76% and 87.60%, respectively. Compared to three orthopedic surgeons, the diagnostic performance of the 3D-ONFHNet framework was comparable to that of senior orthopedic surgeons and superior to that of junior orthopedic surgeons.

Conclusions: The framework proposed in this study can generate staging results for early ONFH and provide visualizations of internal signal changes within the femoral head. It assists orthopedic surgeons in screening for early ONFH on MRI in a clinical setting, facilitating preoperative planning and subsequent treatment strategies. This framework not only enhances diagnostic efficiency but also offers valuable diagnostic references for physicians.

查看原文本刊更多论文

求助全文

约1分钟内获得全文求助全文

来源期刊

Frontiers in Surgery Medicine-Surgery

CiteScore

1.90

自引率

11.10%

发文量

1872

审稿时长

12 weeks

期刊介绍： Evidence of surgical interventions go back to prehistoric times. Since then, the field of surgery has developed into a complex array of specialties and procedures, particularly with the advent of microsurgery, lasers and minimally invasive techniques. The advanced skills now required from surgeons has led to ever increasing specialization, though these still share important fundamental principles. Frontiers in Surgery is the umbrella journal representing the publication interests of all surgical specialties. It is divided into several “Specialty Sections” listed below. All these sections have their own Specialty Chief Editor, Editorial Board and homepage, but all articles carry the citation Frontiers in Surgery. Frontiers in Surgery calls upon medical professionals and scientists from all surgical specialties to publish their experimental and clinical studies in this journal. By assembling all surgical specialties, which nonetheless retain their independence, under the common umbrella of Frontiers in Surgery, a powerful publication venue is created. Since there is often overlap and common ground between the different surgical specialties, assembly of all surgical disciplines into a single journal will foster a collaborative dialogue amongst the surgical community. This means that publications, which are also of interest to other surgical specialties, will reach a wider audience and have greater impact. The aim of this multidisciplinary journal is to create a discussion and knowledge platform of advances and research findings in surgical practice today to continuously improve clinical management of patients and foster innovation in this field.