A Novel Deep Hybrid Model for Automatic Femoral Stem Classification in Hip Arthroplasty From Radiographs: MSFT-Net With CBAM and Transformer Modules

Accurate identification of femoral stem implants in hip arthroplasty is essential for effective revision surgery, minimizing operative complexity, patient morbidity, intraoperative blood loss, and postoperative recovery time. In cases where prior implant data are unavailable, manual identification i...

Full description

Saved in:
Bibliographic Details
Main Authors: Emre Gogus, Atinc Yilmaz, Meric Enercan
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/11030603/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850218995653804032
author Emre Gogus
Atinc Yilmaz
Meric Enercan
author_facet Emre Gogus
Atinc Yilmaz
Meric Enercan
author_sort Emre Gogus
collection DOAJ
description Accurate identification of femoral stem implants in hip arthroplasty is essential for effective revision surgery, minimizing operative complexity, patient morbidity, intraoperative blood loss, and postoperative recovery time. In cases where prior implant data are unavailable, manual identification is often required, posing significant challenges due to its time-consuming and error-prone nature. To solve this problem, a novel hybrid deep learning architecture that includes a convolutional block attention module and a swin transformer with multi-scale feature fusion from pre-trained architectures DenseNet201, VGG19, and InceptionV3 under the transfer learning paradigm was proposed in this study. The proposed multi-scale feature transformer network was trained and validated on a dataset comprising 1266 anteroposterior (A.P.) hip radiographs of 10 different femoral stem implant types. The proposed hybrid deep learning architecture achieved a training accuracy of 96.7% and validation accuracy of 94.86%, significantly outperforming other baseline models. Compared with state-of-the-art methods, the proposed model achieved an absolute accuracy improvement of 9.5% over VGG19 and 7.4% over DenseNet201 and 8.8% over InceptionV3, demonstrating a significant advancement over existing models in femoral stem classification. The average inference time per image was under 1 second. The experimental results demonstrated that the proposed architecture enhances classification performance while reducing overfitting through attention and transformer-based feature refinement. This automated approach facilitates real-time preoperative implant recognition, thereby streamlining surgical planning, potentially reducing operative costs and duration, and improving clinical outcomes.
format Article
id doaj-art-26169f0187774b66aa17c29fdaabae9d
institution OA Journals
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-26169f0187774b66aa17c29fdaabae9d2025-08-20T02:07:31ZengIEEEIEEE Access2169-35362025-01-011310256410257710.1109/ACCESS.2025.357891911030603A Novel Deep Hybrid Model for Automatic Femoral Stem Classification in Hip Arthroplasty From Radiographs: MSFT-Net With CBAM and Transformer ModulesEmre Gogus0https://orcid.org/0009-0008-5200-5301Atinc Yilmaz1https://orcid.org/0000-0003-0038-7519Meric Enercan2https://orcid.org/0000-0003-3263-9290The Koç School, İstanbul, TürkiyeDepartment of Computer Engineering, İstanbul Beykent University, İstanbul, TürkiyeDepartment of Orthopedic Surgery, İstanbul Florence Nightingale Hospital, Demiroglu Science University, İstanbul, TürkiyeAccurate identification of femoral stem implants in hip arthroplasty is essential for effective revision surgery, minimizing operative complexity, patient morbidity, intraoperative blood loss, and postoperative recovery time. In cases where prior implant data are unavailable, manual identification is often required, posing significant challenges due to its time-consuming and error-prone nature. To solve this problem, a novel hybrid deep learning architecture that includes a convolutional block attention module and a swin transformer with multi-scale feature fusion from pre-trained architectures DenseNet201, VGG19, and InceptionV3 under the transfer learning paradigm was proposed in this study. The proposed multi-scale feature transformer network was trained and validated on a dataset comprising 1266 anteroposterior (A.P.) hip radiographs of 10 different femoral stem implant types. The proposed hybrid deep learning architecture achieved a training accuracy of 96.7% and validation accuracy of 94.86%, significantly outperforming other baseline models. Compared with state-of-the-art methods, the proposed model achieved an absolute accuracy improvement of 9.5% over VGG19 and 7.4% over DenseNet201 and 8.8% over InceptionV3, demonstrating a significant advancement over existing models in femoral stem classification. The average inference time per image was under 1 second. The experimental results demonstrated that the proposed architecture enhances classification performance while reducing overfitting through attention and transformer-based feature refinement. This automated approach facilitates real-time preoperative implant recognition, thereby streamlining surgical planning, potentially reducing operative costs and duration, and improving clinical outcomes.https://ieeexplore.ieee.org/document/11030603/CBAM attention mechanismdeep learningfemoral stem classificationhip implanthip arthroplastymulti-scale feature fusion
spellingShingle Emre Gogus
Atinc Yilmaz
Meric Enercan
A Novel Deep Hybrid Model for Automatic Femoral Stem Classification in Hip Arthroplasty From Radiographs: MSFT-Net With CBAM and Transformer Modules
IEEE Access
CBAM attention mechanism
deep learning
femoral stem classification
hip implant
hip arthroplasty
multi-scale feature fusion
title A Novel Deep Hybrid Model for Automatic Femoral Stem Classification in Hip Arthroplasty From Radiographs: MSFT-Net With CBAM and Transformer Modules
title_full A Novel Deep Hybrid Model for Automatic Femoral Stem Classification in Hip Arthroplasty From Radiographs: MSFT-Net With CBAM and Transformer Modules
title_fullStr A Novel Deep Hybrid Model for Automatic Femoral Stem Classification in Hip Arthroplasty From Radiographs: MSFT-Net With CBAM and Transformer Modules
title_full_unstemmed A Novel Deep Hybrid Model for Automatic Femoral Stem Classification in Hip Arthroplasty From Radiographs: MSFT-Net With CBAM and Transformer Modules
title_short A Novel Deep Hybrid Model for Automatic Femoral Stem Classification in Hip Arthroplasty From Radiographs: MSFT-Net With CBAM and Transformer Modules
title_sort novel deep hybrid model for automatic femoral stem classification in hip arthroplasty from radiographs msft net with cbam and transformer modules
topic CBAM attention mechanism
deep learning
femoral stem classification
hip implant
hip arthroplasty
multi-scale feature fusion
url https://ieeexplore.ieee.org/document/11030603/
work_keys_str_mv AT emregogus anoveldeephybridmodelforautomaticfemoralstemclassificationinhiparthroplastyfromradiographsmsftnetwithcbamandtransformermodules
AT atincyilmaz anoveldeephybridmodelforautomaticfemoralstemclassificationinhiparthroplastyfromradiographsmsftnetwithcbamandtransformermodules
AT mericenercan anoveldeephybridmodelforautomaticfemoralstemclassificationinhiparthroplastyfromradiographsmsftnetwithcbamandtransformermodules
AT emregogus noveldeephybridmodelforautomaticfemoralstemclassificationinhiparthroplastyfromradiographsmsftnetwithcbamandtransformermodules
AT atincyilmaz noveldeephybridmodelforautomaticfemoralstemclassificationinhiparthroplastyfromradiographsmsftnetwithcbamandtransformermodules
AT mericenercan noveldeephybridmodelforautomaticfemoralstemclassificationinhiparthroplastyfromradiographsmsftnetwithcbamandtransformermodules