Automatic Meniscus Segmentation Using YOLO-Based Deep Learning Models with Ensemble Methods in Knee MRI Images

The meniscus is a C-shaped connective tissue with a cartilage-like structure in the knee joint. This study proposes an innovative method based on You Only Look Once (YOLO) series models and ensemble methods for meniscus segmentation from knee magnetic resonance imaging (MRI) images to improve segmen...

Full description

Saved in:
Bibliographic Details
Main Authors: Mehmet Ali Şimşek, Ahmet Sertbaş, Hadi Sasani, Yaşar Mahsut Dinçel
Format: Article
Language:English
Published: MDPI AG 2025-03-01
Series:Applied Sciences
Subjects:
Online Access:https://www.mdpi.com/2076-3417/15/5/2752
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850225334287335424
author Mehmet Ali Şimşek
Ahmet Sertbaş
Hadi Sasani
Yaşar Mahsut Dinçel
author_facet Mehmet Ali Şimşek
Ahmet Sertbaş
Hadi Sasani
Yaşar Mahsut Dinçel
author_sort Mehmet Ali Şimşek
collection DOAJ
description The meniscus is a C-shaped connective tissue with a cartilage-like structure in the knee joint. This study proposes an innovative method based on You Only Look Once (YOLO) series models and ensemble methods for meniscus segmentation from knee magnetic resonance imaging (MRI) images to improve segmentation performance and evaluate generalization capability. In this study, five different segmentation models were trained, and masks were created from the YOLO series. These masks are combined with pixel-based voting, weighted multiple voting, and dynamic weighted multiple voting optimized by grid search. Tests were conducted on internal and external sets and various metrics. The dynamic weighted multiple voting method optimized with grid search performed the best on both the test set (DSC: 0.8976 ± 0.0071, PPV: 0.8561 ± 0.0121, Sensitivity: 0.9467 ± 0.0077) and the external set (DSC: 0.9004 ± 0.0064, PPV: 0.8876 ± 0.0134, Sensitivity: 0.9200 ± 0.0119). The proposed ensemble methods offer high accuracy, reliability, and generalization capability for meniscus segmentation.
format Article
id doaj-art-e3b24ed808fe4569b81c92d49b792f12
institution OA Journals
issn 2076-3417
language English
publishDate 2025-03-01
publisher MDPI AG
record_format Article
series Applied Sciences
spelling doaj-art-e3b24ed808fe4569b81c92d49b792f122025-08-20T02:05:23ZengMDPI AGApplied Sciences2076-34172025-03-01155275210.3390/app15052752Automatic Meniscus Segmentation Using YOLO-Based Deep Learning Models with Ensemble Methods in Knee MRI ImagesMehmet Ali Şimşek0Ahmet Sertbaş1Hadi Sasani2Yaşar Mahsut Dinçel3Department of Computer Technologies, Vocational School of Technical Sciences, Tekirdag Namik Kemal University, Tekirdag 59030, TurkeyDepartment of Computer Engineering, Faculty of Engineering, University of Istanbul-Cerrahpasa, Istanbul 34320, TurkeyDepartment of Radiology, Faculty of Medicine, Tekirdag Namik Kemal University, Tekirdag 59030, TurkeyDepartment of Orthopedics and Traumatology, Faculty of Medicine, Tekirdag Namik Kemal University, Tekirdag 59030, TurkeyThe meniscus is a C-shaped connective tissue with a cartilage-like structure in the knee joint. This study proposes an innovative method based on You Only Look Once (YOLO) series models and ensemble methods for meniscus segmentation from knee magnetic resonance imaging (MRI) images to improve segmentation performance and evaluate generalization capability. In this study, five different segmentation models were trained, and masks were created from the YOLO series. These masks are combined with pixel-based voting, weighted multiple voting, and dynamic weighted multiple voting optimized by grid search. Tests were conducted on internal and external sets and various metrics. The dynamic weighted multiple voting method optimized with grid search performed the best on both the test set (DSC: 0.8976 ± 0.0071, PPV: 0.8561 ± 0.0121, Sensitivity: 0.9467 ± 0.0077) and the external set (DSC: 0.9004 ± 0.0064, PPV: 0.8876 ± 0.0134, Sensitivity: 0.9200 ± 0.0119). The proposed ensemble methods offer high accuracy, reliability, and generalization capability for meniscus segmentation.https://www.mdpi.com/2076-3417/15/5/2752meniscus segmentationmagnetic resonance imagingYOLO seriesensemble methodsvoting methods
spellingShingle Mehmet Ali Şimşek
Ahmet Sertbaş
Hadi Sasani
Yaşar Mahsut Dinçel
Automatic Meniscus Segmentation Using YOLO-Based Deep Learning Models with Ensemble Methods in Knee MRI Images
Applied Sciences
meniscus segmentation
magnetic resonance imaging
YOLO series
ensemble methods
voting methods
title Automatic Meniscus Segmentation Using YOLO-Based Deep Learning Models with Ensemble Methods in Knee MRI Images
title_full Automatic Meniscus Segmentation Using YOLO-Based Deep Learning Models with Ensemble Methods in Knee MRI Images
title_fullStr Automatic Meniscus Segmentation Using YOLO-Based Deep Learning Models with Ensemble Methods in Knee MRI Images
title_full_unstemmed Automatic Meniscus Segmentation Using YOLO-Based Deep Learning Models with Ensemble Methods in Knee MRI Images
title_short Automatic Meniscus Segmentation Using YOLO-Based Deep Learning Models with Ensemble Methods in Knee MRI Images
title_sort automatic meniscus segmentation using yolo based deep learning models with ensemble methods in knee mri images
topic meniscus segmentation
magnetic resonance imaging
YOLO series
ensemble methods
voting methods
url https://www.mdpi.com/2076-3417/15/5/2752
work_keys_str_mv AT mehmetalisimsek automaticmeniscussegmentationusingyolobaseddeeplearningmodelswithensemblemethodsinkneemriimages
AT ahmetsertbas automaticmeniscussegmentationusingyolobaseddeeplearningmodelswithensemblemethodsinkneemriimages
AT hadisasani automaticmeniscussegmentationusingyolobaseddeeplearningmodelswithensemblemethodsinkneemriimages
AT yasarmahsutdincel automaticmeniscussegmentationusingyolobaseddeeplearningmodelswithensemblemethodsinkneemriimages