Cross-Center Validation of Deep Learning Model for
Musculoskeletal Fracture Detection in Radiographic Imaging: A
Feasibility Study

j 2024

Cross-Center Validation of Deep Learning Model for Musculoskeletal Fracture Detection in Radiographic Imaging: A Feasibility Study

HRUBÝ, Robert; Daniel KVAK; Jakub DANDÁR; Anora ATAKHANOVA; Matěj MISAŘ et al.

Základní údaje

Originální název

Cross-Center Validation of Deep Learning Model for Musculoskeletal Fracture Detection in Radiographic Imaging: A Feasibility Study

Autoři

HRUBÝ, Robert; Daniel KVAK; Jakub DANDÁR; Anora ATAKHANOVA; Matěj MISAŘ a Daniel DUFEK

Vydání

MedRxiv - The Preprint Server for Health Sciences, Cold Spring Harbor Laboratory, Yale, BMJ, and Cold Spring Harbor Lab. 2024

Další údaje

Typ výsledku

Publikace v odborném periodiku – kromě recenzovaných typů article, review a letter

Utajení

není předmětem státního či obchodního tajemství

Odkazy

URL

Označené pro přenos do RIV

DOI

https://doi.org/10.1101/2024.01.17.24301244

Klíčová slova anglicky

Bone Fracture, Computer-Aided Detection, Deep Learning, Musculoskeletal X-ray, YOLO

Změněno: 18. 1. 2024 14:43, Mgr. Daniel Kvak

Anotace

V originále

Fractures, often resulting from trauma, overuse, or osteoporosis, pose diagnostic challenges due to their variable clinical manifestations. To address this, we propose a deep learning-based decision support system to enhance the efficacy of fracture detection in radiographic imaging. For the purpose of our study, we utilized 720 annotated musculoskeletal (MSK) X-rays from the MURA dataset, augmented by bounding box-level annotation, for training the YOLO (You Only Look Once) model. The model's performance was subsequently tested on two datasets, sampled FracAtlas dataset (Dataset 1, 840 images, n_NORMAL=696, n_FRACTURE=144) and our own internal dataset (Dataset 2, 124 images, n_NORMAL=50, n_FRACTURE=74), encompassing a diverse range of MSK radiographs. The results showed a Sensitivity (Se) of 0.910 (95% CI: 0.852-0.946) and Specificity (Sp) of 0.557 (95% CI: 0.520-0.594) on Dataset 1, and a Se of 0.622 (95% CI: 0.508-0.724) and Sp of 0.740 (95% CI: 0.604-0.841) on Dataset 2. This study underscores the promising role of AI in medical imaging, providing a solid foundation for future research and advancements in the field of radiographic diagnostics.

Citovat

HRUBÝ, Robert; Daniel KVAK; Jakub DANDÁR; Anora ATAKHANOVA; Matěj MISAŘ a Daniel DUFEK. Cross-Center Validation of Deep Learning Model for Musculoskeletal Fracture Detection in Radiographic Imaging: A Feasibility Study. MedRxiv - The Preprint Server for Health Sciences. Cold Spring Harbor Laboratory: Yale, BMJ, and Cold Spring Harbor Lab., 2024, s. 1-9. Dostupné z: https://doi.org/10.1101/2024.01.17.24301244.

@article{2362561,
   author = {Hrubý, Robert and Kvak, Daniel and Dandár, Jakub and Atakhanova, Anora and Misař, Matěj and Dufek, Daniel},
   article_location = {Cold Spring Harbor Laboratory},
   doi = {https://doi.org/10.1101/2024.01.17.24301244},
   keywords = {Bone Fracture, Computer-Aided Detection, Deep Learning, Musculoskeletal X-ray, YOLO},
   journal = {MedRxiv - The Preprint Server for Health Sciences},
   title = {Cross-Center Validation of Deep Learning Model for Musculoskeletal Fracture Detection in Radiographic Imaging: A Feasibility Study},
   url = {https://medrxiv.org/cgi/content/short/2024.01.17.24301244v1},
   year = {2024}
}

TY  - JFULL
ID  - 2362561
AU  - Hrubý, Robert - Kvak, Daniel - Dandár, Jakub - Atakhanova, Anora - Misař, Matěj - Dufek, Daniel
PY  - 2024
TI  - Cross-Center Validation of Deep Learning Model for Musculoskeletal Fracture Detection in Radiographic Imaging: A Feasibility Study
JF  - MedRxiv - The Preprint Server for Health Sciences
SP  - 1-9
EP  - 1-9
PB  - Yale, BMJ, and Cold Spring Harbor Lab.
KW  - Bone Fracture, Computer-Aided Detection, Deep Learning, Musculoskeletal X-ray, YOLO
UR  - https://medrxiv.org/cgi/content/short/2024.01.17.24301244v1
N2  - Fractures, often resulting from trauma, overuse, or osteoporosis, pose diagnostic challenges due to their variable clinical manifestations. To address this, we propose a deep learning-based decision support system to enhance the efficacy of fracture detection in radiographic imaging. For the purpose of our study, we utilized 720 annotated musculoskeletal (MSK) X-rays from the MURA dataset, augmented by bounding box-level annotation, for training the YOLO (You Only Look Once) model. The model's performance was subsequently tested on two datasets, sampled FracAtlas dataset (Dataset 1, 840 images, n_NORMAL=696, n_FRACTURE=144) and our own internal dataset (Dataset 2, 124 images, n_NORMAL=50, n_FRACTURE=74), encompassing a diverse range of MSK radiographs. The results showed a Sensitivity (Se) of 0.910 (95% CI: 0.852-0.946) and Specificity (Sp) of 0.557 (95% CI: 0.520-0.594) on Dataset 1, and a Se of 0.622 (95% CI: 0.508-0.724) and Sp of 0.740 (95% CI: 0.604-0.841) on Dataset 2. This study underscores the promising role of AI in medical imaging, providing a solid foundation for future research and advancements in the field of radiographic diagnostics.
ER  -

HRUBÝ, Robert; Daniel KVAK; Jakub DANDÁR; Anora ATAKHANOVA; Matěj MISAŘ a Daniel DUFEK. Cross-Center Validation of Deep Learning Model for Musculoskeletal Fracture Detection in Radiographic Imaging: A Feasibility Study. \textit{MedRxiv - The Preprint Server for Health Sciences}. Cold Spring Harbor Laboratory: Yale, BMJ, and Cold Spring Harbor Lab., 2024, s.~1-9. Dostupné z: https://doi.org/10.1101/2024.01.17.24301244.

Přehled o publikaci