Objective classification of masseter muscle and deep inferior tendon morphology: A comparative performance analysis of machine learning algorithms.

IF 5.4 4区医学 Q2 DENTISTRY, ORAL SURGERY & MEDICINE

Cranio-The Journal of Craniomandibular & Sleep Practice Pub Date : 2026-10-05 DOI:10.1080/08869634.2026.2741836

Merva Güneyli, Merve Edik, Fatma Latifoğlu, Aykağan Çukurluoğlu, Şerife Arzu Çopur

{"title":"Objective classification of masseter muscle and deep inferior tendon morphology: A comparative performance analysis of machine learning algorithms.","authors":"Merva Güneyli, Merve Edik, Fatma Latifoğlu, Aykağan Çukurluoğlu, Şerife Arzu Çopur","doi":"10.1080/08869634.2026.2741836","DOIUrl":null,"url":null,"abstract":"Objectives: This study objectively classified masseter internal structure and deep inferior tendon (DIT) morphology using ultrasonography and machine learning on 229 images from 101 (DIT) and 97 (MUSCLE) patients.Methods: Following hybrid feature extraction (radiomics, LBP, HOG), patient-grouped, fold-internal SMOTE balancing, and LASSO-based feature selection, four algorithms were cross-validated under a leakage-free scheme.Results: All four classifiers achieved accuracy modestly but consistently above the one-third chance level for this three-class problem (47.9-54.3%), with no algorithm showing a consistent advantage. LASSO selection-frequency analysis identified a feature-specific asymmetry: a single GLSZM feature was selected in 46 of 50 folds for MUSCLE classification but never for DIT classification, alongside HOG-dominated features in both tasks.Conclusions: This AI-based framework provides an objective approach for classifying masseter internal structure and DIT morphology, potentially reducing observer-dependent variability. These findings may underpin future studies investigating whether automated morphological classification can support clinical assessment and procedure planning.","PeriodicalId":56318,"journal":{"name":"Cranio-The Journal of Craniomandibular & Sleep Practice","volume":" ","pages":"1-16"},"PeriodicalIF":5.4000,"publicationDate":"2026-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Cranio-The Journal of Craniomandibular & Sleep Practice","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1080/08869634.2026.2741836","RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"DENTISTRY, ORAL SURGERY & MEDICINE","Score":null,"Total":0}

引用次数: 0

Abstract

Objectives: This study objectively classified masseter internal structure and deep inferior tendon (DIT) morphology using ultrasonography and machine learning on 229 images from 101 (DIT) and 97 (MUSCLE) patients.

Methods: Following hybrid feature extraction (radiomics, LBP, HOG), patient-grouped, fold-internal SMOTE balancing, and LASSO-based feature selection, four algorithms were cross-validated under a leakage-free scheme.

Results: All four classifiers achieved accuracy modestly but consistently above the one-third chance level for this three-class problem (47.9-54.3%), with no algorithm showing a consistent advantage. LASSO selection-frequency analysis identified a feature-specific asymmetry: a single GLSZM feature was selected in 46 of 50 folds for MUSCLE classification but never for DIT classification, alongside HOG-dominated features in both tasks.

Conclusions: This AI-based framework provides an objective approach for classifying masseter internal structure and DIT morphology, potentially reducing observer-dependent variability. These findings may underpin future studies investigating whether automated morphological classification can support clinical assessment and procedure planning.

查看原文本刊更多论文

咬肌和深下肌腱形态的客观分类：机器学习算法的比较性能分析。

目的：对101例（DIT）和97例（MUSCLE）患者229张图像进行超声和机器学习，对咬肌内部结构和深下肌腱（DIT）形态进行客观分类。方法：采用混合特征提取（radiomics， LBP， HOG），患者分组，fold-internal SMOTE平衡和基于lasso的特征选择，在无泄漏方案下交叉验证四种算法。结果：对于这三个类别的问题（47.9-54.3%），所有四种分类器都达到了适度的准确性，但始终高于三分之一的机会水平，没有算法显示出一致的优势。LASSO选择频率分析确定了特征特异性的不对称性：在50次分类中，有46次选择了单个GLSZM特征用于MUSCLE分类，但从未选择DIT分类，同时在两个任务中都选择了hog主导特征。结论：这个基于人工智能的框架提供了客观的方法来分类咬咬器内部结构和DIT形态，潜在地减少了观察者依赖的变异。这些发现可能为未来的研究提供基础，调查自动形态学分类是否可以支持临床评估和手术计划。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Cranio-The Journal of Craniomandibular & Sleep Practice DENTISTRY, ORAL SURGERY & MEDICINE-

CiteScore

4.50

自引率

12.50%

发文量

审稿时长

>12 weeks

期刊介绍： CRANIO: The Journal of Craniomandibular & Sleep Practice is the oldest and largest journal in the world devoted to temporomandibular disorders, and now also includes articles on all aspects of sleep medicine. The Journal is multidisciplinary in its scope, with editorial board members from all areas of medicine and dentistry, including general dentists, oral surgeons, orthopaedists, radiologists, chiropractors, professors and behavioural scientists, physical therapists, acupuncturists, osteopathic and ear, nose and throat physicians. CRANIO publishes commendable works from outstanding researchers and clinicians in their respective fields. The multidisciplinary format allows individuals practicing with a TMD emphasis to stay abreast of related disciplines, as each issue presents multiple topics from overlapping areas of interest. CRANIO''s current readership (thousands) is comprised primarily of dentists; however, many physicians, physical therapists, chiropractors, osteopathic physicians and other related specialists subscribe and contribute to the Journal.