HKA-Net: clinically-adapted deep learning for automated measurement of hip-knee-ankle angle on lower limb radiography for knee osteoarthritis assessment

Abstract Background Accurate measurement of the hip-knee-ankle (HKA) angle is essential for informed clinical decision-making in the management of knee osteoarthritis (OA). Knee OA is commonly associated with varus deformity, where the alignment of the knee shifts medially, leading to increased stre...

Full description

Saved in:
Bibliographic Details
Main Authors: Young-Tak Kim, Beom-Su Han, Jung Bin Kim, Jason K. Sa, Je Hyeong Hong, Yunsik Son, Jae-Ho Han, Synho Do, Ji Seon Chae, Jung-Kwon Bae
Format: Article
Language:English
Published: BMC 2024-11-01
Series:Journal of Orthopaedic Surgery and Research
Subjects:
Online Access:https://doi.org/10.1186/s13018-024-05265-y
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850127973915557888
author Young-Tak Kim
Beom-Su Han
Jung Bin Kim
Jason K. Sa
Je Hyeong Hong
Yunsik Son
Jae-Ho Han
Synho Do
Ji Seon Chae
Jung-Kwon Bae
author_facet Young-Tak Kim
Beom-Su Han
Jung Bin Kim
Jason K. Sa
Je Hyeong Hong
Yunsik Son
Jae-Ho Han
Synho Do
Ji Seon Chae
Jung-Kwon Bae
author_sort Young-Tak Kim
collection DOAJ
description Abstract Background Accurate measurement of the hip-knee-ankle (HKA) angle is essential for informed clinical decision-making in the management of knee osteoarthritis (OA). Knee OA is commonly associated with varus deformity, where the alignment of the knee shifts medially, leading to increased stress and deterioration of the medial compartment. The HKA angle, which quantifies this alignment, is a critical indicator of the severity of varus deformity and helps guide treatment strategies, including corrective surgeries. Current manual methods are labor-intensive, time-consuming, and prone to inter-observer variability. Developing an automated model for HKA angle measurement is challenging due to the elaborate process of generating handcrafted anatomical landmarks, which is more labor-intensive than the actual measurement. This study aims to develop a ResNet-based deep learning model that predicts the HKA angle without requiring explicit anatomical landmark annotations and to assess its accuracy and efficiency compared to conventional manual methods. Methods We developed a deep learning model based on the variants of the ResNet architecture to process lower limb radiographs and predict HKA angles without explicit landmark annotations. The classification performance for the four stages of varus deformity (stage I: 0°–10°, stage II: 10°–20°, stage III: > 20°, others: genu valgum or normal alignment) was also evaluated. The model was trained and validated using a retrospective cohort of 300 knee OA patients (Kellgren-Lawrence grade 3 or higher), with horizontal flip augmentation applied to double the dataset to 600 samples, followed by fivefold cross-validation. An extended temporal validation was conducted on a separate cohort of 50 knee OA patients. The model's accuracy was assessed by calculating the mean absolute error between predicted and actual HKA angles. Additionally, the classification of varus deformity stages was conducted to evaluate the model's ability to provide clinically relevant categorizations. Time efficiency was compared between the automated model and manual measurements performed by an experienced orthopedic surgeon. Results The ResNet-50 model achieved a bias of − 0.025° with a standard deviation of 1.422° in the retrospective cohort and a bias of − 0.008° with a standard deviation of 1.677° in the temporal validation cohort. Using the ResNet-152 model, it accurately classified the four stages of varus deformity with weighted F1-score of 0.878 and 0.859 in the retrospective and temporal validation cohorts, respectively. The automated model was 126.7 times faster than manual measurements, reducing the total time from 49.8 min to 23.6 sec for the temporal validation cohort. Conclusions The proposed ResNet-based model provides an efficient and accurate method for measuring HKA angles and classifying varus deformity stages without the need for extensive landmark annotations. Its high accuracy and significant improvement in time efficiency make it a valuable tool for clinical practice, potentially enhancing decision-making and workflow efficiency in the management of knee OA.
format Article
id doaj-art-5bdad98d35e241aa84e307a044e7c09f
institution OA Journals
issn 1749-799X
language English
publishDate 2024-11-01
publisher BMC
record_format Article
series Journal of Orthopaedic Surgery and Research
spelling doaj-art-5bdad98d35e241aa84e307a044e7c09f2025-08-20T02:33:31ZengBMCJournal of Orthopaedic Surgery and Research1749-799X2024-11-0119111410.1186/s13018-024-05265-yHKA-Net: clinically-adapted deep learning for automated measurement of hip-knee-ankle angle on lower limb radiography for knee osteoarthritis assessmentYoung-Tak Kim0Beom-Su Han1Jung Bin Kim2Jason K. Sa3Je Hyeong Hong4Yunsik Son5Jae-Ho Han6Synho Do7Ji Seon Chae8Jung-Kwon Bae9Department of Radiology, Massachusetts General Hospital, Harvard Medical SchoolDepartment of Orthopedic Surgery, Seoul Medical CenterDepartment of Neurology, Korea University Anam Hospital, Korea University College of MedicineDepartment of Biomedical Sciences, Korea University College of MedicineDepartment of Electronic Engineering, Hanyang UniversityDepartment of Computer Science and Engineering, Dongguk UniversityDepartment of Brain and Cognitive Engineering, Korea UniversityDepartment of Radiology, Massachusetts General Hospital, Harvard Medical SchoolDepartment of Anesthesiology and Pain Medicine, College of Medicine, Ewha Womans UniversityDepartment of Orthopedic Surgery, Seoul Medical CenterAbstract Background Accurate measurement of the hip-knee-ankle (HKA) angle is essential for informed clinical decision-making in the management of knee osteoarthritis (OA). Knee OA is commonly associated with varus deformity, where the alignment of the knee shifts medially, leading to increased stress and deterioration of the medial compartment. The HKA angle, which quantifies this alignment, is a critical indicator of the severity of varus deformity and helps guide treatment strategies, including corrective surgeries. Current manual methods are labor-intensive, time-consuming, and prone to inter-observer variability. Developing an automated model for HKA angle measurement is challenging due to the elaborate process of generating handcrafted anatomical landmarks, which is more labor-intensive than the actual measurement. This study aims to develop a ResNet-based deep learning model that predicts the HKA angle without requiring explicit anatomical landmark annotations and to assess its accuracy and efficiency compared to conventional manual methods. Methods We developed a deep learning model based on the variants of the ResNet architecture to process lower limb radiographs and predict HKA angles without explicit landmark annotations. The classification performance for the four stages of varus deformity (stage I: 0°–10°, stage II: 10°–20°, stage III: > 20°, others: genu valgum or normal alignment) was also evaluated. The model was trained and validated using a retrospective cohort of 300 knee OA patients (Kellgren-Lawrence grade 3 or higher), with horizontal flip augmentation applied to double the dataset to 600 samples, followed by fivefold cross-validation. An extended temporal validation was conducted on a separate cohort of 50 knee OA patients. The model's accuracy was assessed by calculating the mean absolute error between predicted and actual HKA angles. Additionally, the classification of varus deformity stages was conducted to evaluate the model's ability to provide clinically relevant categorizations. Time efficiency was compared between the automated model and manual measurements performed by an experienced orthopedic surgeon. Results The ResNet-50 model achieved a bias of − 0.025° with a standard deviation of 1.422° in the retrospective cohort and a bias of − 0.008° with a standard deviation of 1.677° in the temporal validation cohort. Using the ResNet-152 model, it accurately classified the four stages of varus deformity with weighted F1-score of 0.878 and 0.859 in the retrospective and temporal validation cohorts, respectively. The automated model was 126.7 times faster than manual measurements, reducing the total time from 49.8 min to 23.6 sec for the temporal validation cohort. Conclusions The proposed ResNet-based model provides an efficient and accurate method for measuring HKA angles and classifying varus deformity stages without the need for extensive landmark annotations. Its high accuracy and significant improvement in time efficiency make it a valuable tool for clinical practice, potentially enhancing decision-making and workflow efficiency in the management of knee OA.https://doi.org/10.1186/s13018-024-05265-yHip-knee-ankle angleDeep learningResidual networkVarus deformityKnee osteoarthritisLower limb scanogram
spellingShingle Young-Tak Kim
Beom-Su Han
Jung Bin Kim
Jason K. Sa
Je Hyeong Hong
Yunsik Son
Jae-Ho Han
Synho Do
Ji Seon Chae
Jung-Kwon Bae
HKA-Net: clinically-adapted deep learning for automated measurement of hip-knee-ankle angle on lower limb radiography for knee osteoarthritis assessment
Journal of Orthopaedic Surgery and Research
Hip-knee-ankle angle
Deep learning
Residual network
Varus deformity
Knee osteoarthritis
Lower limb scanogram
title HKA-Net: clinically-adapted deep learning for automated measurement of hip-knee-ankle angle on lower limb radiography for knee osteoarthritis assessment
title_full HKA-Net: clinically-adapted deep learning for automated measurement of hip-knee-ankle angle on lower limb radiography for knee osteoarthritis assessment
title_fullStr HKA-Net: clinically-adapted deep learning for automated measurement of hip-knee-ankle angle on lower limb radiography for knee osteoarthritis assessment
title_full_unstemmed HKA-Net: clinically-adapted deep learning for automated measurement of hip-knee-ankle angle on lower limb radiography for knee osteoarthritis assessment
title_short HKA-Net: clinically-adapted deep learning for automated measurement of hip-knee-ankle angle on lower limb radiography for knee osteoarthritis assessment
title_sort hka net clinically adapted deep learning for automated measurement of hip knee ankle angle on lower limb radiography for knee osteoarthritis assessment
topic Hip-knee-ankle angle
Deep learning
Residual network
Varus deformity
Knee osteoarthritis
Lower limb scanogram
url https://doi.org/10.1186/s13018-024-05265-y
work_keys_str_mv AT youngtakkim hkanetclinicallyadapteddeeplearningforautomatedmeasurementofhipkneeankleangleonlowerlimbradiographyforkneeosteoarthritisassessment
AT beomsuhan hkanetclinicallyadapteddeeplearningforautomatedmeasurementofhipkneeankleangleonlowerlimbradiographyforkneeosteoarthritisassessment
AT jungbinkim hkanetclinicallyadapteddeeplearningforautomatedmeasurementofhipkneeankleangleonlowerlimbradiographyforkneeosteoarthritisassessment
AT jasonksa hkanetclinicallyadapteddeeplearningforautomatedmeasurementofhipkneeankleangleonlowerlimbradiographyforkneeosteoarthritisassessment
AT jehyeonghong hkanetclinicallyadapteddeeplearningforautomatedmeasurementofhipkneeankleangleonlowerlimbradiographyforkneeosteoarthritisassessment
AT yunsikson hkanetclinicallyadapteddeeplearningforautomatedmeasurementofhipkneeankleangleonlowerlimbradiographyforkneeosteoarthritisassessment
AT jaehohan hkanetclinicallyadapteddeeplearningforautomatedmeasurementofhipkneeankleangleonlowerlimbradiographyforkneeosteoarthritisassessment
AT synhodo hkanetclinicallyadapteddeeplearningforautomatedmeasurementofhipkneeankleangleonlowerlimbradiographyforkneeosteoarthritisassessment
AT jiseonchae hkanetclinicallyadapteddeeplearningforautomatedmeasurementofhipkneeankleangleonlowerlimbradiographyforkneeosteoarthritisassessment
AT jungkwonbae hkanetclinicallyadapteddeeplearningforautomatedmeasurementofhipkneeankleangleonlowerlimbradiographyforkneeosteoarthritisassessment