Context-Aware Level-Wise Feature Fusion Network with Anomaly Focus for Precise Classification of Incomplete Atypical Femoral Fractures in X-Ray Images

Incomplete Atypical Femoral Fracture (IAFF) is a precursor to Atypical Femoral Fracture (AFF). If untreated, it progresses to a complete fracture, increasing mortality risk. However, due to their small and ambiguous features, IAFFs are often misdiagnosed even by specialists. In this paper, we propos...

Full description

Saved in:
Bibliographic Details
Main Authors: Joonho Chang, Junwon Lee, Doyoung Kwon, Jin-Han Lee, Minho Lee, Sungmoon Jeong, Joon-Woo Kim, Heechul Jung, Chang-Wug Oh
Format: Article
Language:English
Published: MDPI AG 2024-11-01
Series:Mathematics
Subjects:
Online Access:https://www.mdpi.com/2227-7390/12/22/3613
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850068667985821696
author Joonho Chang
Junwon Lee
Doyoung Kwon
Jin-Han Lee
Minho Lee
Sungmoon Jeong
Joon-Woo Kim
Heechul Jung
Chang-Wug Oh
author_facet Joonho Chang
Junwon Lee
Doyoung Kwon
Jin-Han Lee
Minho Lee
Sungmoon Jeong
Joon-Woo Kim
Heechul Jung
Chang-Wug Oh
author_sort Joonho Chang
collection DOAJ
description Incomplete Atypical Femoral Fracture (IAFF) is a precursor to Atypical Femoral Fracture (AFF). If untreated, it progresses to a complete fracture, increasing mortality risk. However, due to their small and ambiguous features, IAFFs are often misdiagnosed even by specialists. In this paper, we propose a novel approach for accurately classifying IAFFs in X-ray images across various radiographic views. We design a Dual Context-aware Complementary Extractor (DCCE) to capture both the overall femur characteristics and IAFF details with the surrounding context, minimizing information loss. We also develop a Level-wise Perspective-preserving Fusion Network (LPFN) that preserves the perspective of features while integrating them at different levels to enhance model representation and sensitivity by learning complex correlations and features that are difficult to obtain independently. Additionally, we incorporate the Spatial Anomaly Focus Enhancer (SAFE) to emphasize anomalous regions, preventing the model bias toward normal regions, and reducing False Negatives and missed IAFFs. Experimental results show significant improvements across all evaluation metrics, demonstrating high reliability in terms of accuracy (0.931), F1-score (0.9456), and AUROC (0.9692), proving the model’s potential for application in real medical settings.
format Article
id doaj-art-feae55d3ef444bf0bd90a5af7019ca49
institution DOAJ
issn 2227-7390
language English
publishDate 2024-11-01
publisher MDPI AG
record_format Article
series Mathematics
spelling doaj-art-feae55d3ef444bf0bd90a5af7019ca492025-08-20T02:48:00ZengMDPI AGMathematics2227-73902024-11-011222361310.3390/math12223613Context-Aware Level-Wise Feature Fusion Network with Anomaly Focus for Precise Classification of Incomplete Atypical Femoral Fractures in X-Ray ImagesJoonho Chang0Junwon Lee1Doyoung Kwon2Jin-Han Lee3Minho Lee4Sungmoon Jeong5Joon-Woo Kim6Heechul Jung7Chang-Wug Oh8Department of Artificial Intelligence, Kyungpook National University, Daegu 41566, Republic of KoreaDepartment of Artificial Intelligence, Kyungpook National University, Daegu 41566, Republic of KoreaDepartment of Artificial Intelligence, Kyungpook National University, Daegu 41566, Republic of KoreaDepartment of Orthopedic Surgery, School of Medicine, Kyungpook National University Hospital, Daegu 41566, Republic of KoreaDepartment of Artificial Intelligence, Kyungpook National University, Daegu 41566, Republic of KoreaDepartment of Medical Informatics, School of Medicine, Kyungpook National University, Daegu 41566, Republic of KoreaDepartment of Orthopedic Surgery, School of Medicine, Kyungpook National University Hospital, Daegu 41566, Republic of KoreaDepartment of Artificial Intelligence, Kyungpook National University, Daegu 41566, Republic of KoreaDepartment of Orthopedic Surgery, School of Medicine, Kyungpook National University Hospital, Daegu 41566, Republic of KoreaIncomplete Atypical Femoral Fracture (IAFF) is a precursor to Atypical Femoral Fracture (AFF). If untreated, it progresses to a complete fracture, increasing mortality risk. However, due to their small and ambiguous features, IAFFs are often misdiagnosed even by specialists. In this paper, we propose a novel approach for accurately classifying IAFFs in X-ray images across various radiographic views. We design a Dual Context-aware Complementary Extractor (DCCE) to capture both the overall femur characteristics and IAFF details with the surrounding context, minimizing information loss. We also develop a Level-wise Perspective-preserving Fusion Network (LPFN) that preserves the perspective of features while integrating them at different levels to enhance model representation and sensitivity by learning complex correlations and features that are difficult to obtain independently. Additionally, we incorporate the Spatial Anomaly Focus Enhancer (SAFE) to emphasize anomalous regions, preventing the model bias toward normal regions, and reducing False Negatives and missed IAFFs. Experimental results show significant improvements across all evaluation metrics, demonstrating high reliability in terms of accuracy (0.931), F1-score (0.9456), and AUROC (0.9692), proving the model’s potential for application in real medical settings.https://www.mdpi.com/2227-7390/12/22/3613Incomplete Atypical Femoral FractureAtypical Femoral FractureX-rayfeature fusionanomaly focustiny lesion
spellingShingle Joonho Chang
Junwon Lee
Doyoung Kwon
Jin-Han Lee
Minho Lee
Sungmoon Jeong
Joon-Woo Kim
Heechul Jung
Chang-Wug Oh
Context-Aware Level-Wise Feature Fusion Network with Anomaly Focus for Precise Classification of Incomplete Atypical Femoral Fractures in X-Ray Images
Mathematics
Incomplete Atypical Femoral Fracture
Atypical Femoral Fracture
X-ray
feature fusion
anomaly focus
tiny lesion
title Context-Aware Level-Wise Feature Fusion Network with Anomaly Focus for Precise Classification of Incomplete Atypical Femoral Fractures in X-Ray Images
title_full Context-Aware Level-Wise Feature Fusion Network with Anomaly Focus for Precise Classification of Incomplete Atypical Femoral Fractures in X-Ray Images
title_fullStr Context-Aware Level-Wise Feature Fusion Network with Anomaly Focus for Precise Classification of Incomplete Atypical Femoral Fractures in X-Ray Images
title_full_unstemmed Context-Aware Level-Wise Feature Fusion Network with Anomaly Focus for Precise Classification of Incomplete Atypical Femoral Fractures in X-Ray Images
title_short Context-Aware Level-Wise Feature Fusion Network with Anomaly Focus for Precise Classification of Incomplete Atypical Femoral Fractures in X-Ray Images
title_sort context aware level wise feature fusion network with anomaly focus for precise classification of incomplete atypical femoral fractures in x ray images
topic Incomplete Atypical Femoral Fracture
Atypical Femoral Fracture
X-ray
feature fusion
anomaly focus
tiny lesion
url https://www.mdpi.com/2227-7390/12/22/3613
work_keys_str_mv AT joonhochang contextawarelevelwisefeaturefusionnetworkwithanomalyfocusforpreciseclassificationofincompleteatypicalfemoralfracturesinxrayimages
AT junwonlee contextawarelevelwisefeaturefusionnetworkwithanomalyfocusforpreciseclassificationofincompleteatypicalfemoralfracturesinxrayimages
AT doyoungkwon contextawarelevelwisefeaturefusionnetworkwithanomalyfocusforpreciseclassificationofincompleteatypicalfemoralfracturesinxrayimages
AT jinhanlee contextawarelevelwisefeaturefusionnetworkwithanomalyfocusforpreciseclassificationofincompleteatypicalfemoralfracturesinxrayimages
AT minholee contextawarelevelwisefeaturefusionnetworkwithanomalyfocusforpreciseclassificationofincompleteatypicalfemoralfracturesinxrayimages
AT sungmoonjeong contextawarelevelwisefeaturefusionnetworkwithanomalyfocusforpreciseclassificationofincompleteatypicalfemoralfracturesinxrayimages
AT joonwookim contextawarelevelwisefeaturefusionnetworkwithanomalyfocusforpreciseclassificationofincompleteatypicalfemoralfracturesinxrayimages
AT heechuljung contextawarelevelwisefeaturefusionnetworkwithanomalyfocusforpreciseclassificationofincompleteatypicalfemoralfracturesinxrayimages
AT changwugoh contextawarelevelwisefeaturefusionnetworkwithanomalyfocusforpreciseclassificationofincompleteatypicalfemoralfracturesinxrayimages