Exploring the impact of hyperparameter and data augmentation in YOLO V10 for accurate bone fracture detection from X-ray images DOI Creative Commons
Parvathaneni Naga Srinivasu, Gorli L. Aruna Kumari,

Sujatha Canavoy Narahari

et al.

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: March 21, 2025

Accurately identifying bone fractures from the X-ray image is essential to prompt timely and appropriate medical treatment. This research explores impact of hyperparameters data augmentation techniques on performance You Only Look Once (YOLO) V10 architecture for fracture detection. While YOLO architectures have been widely employed in object detection tasks, recognizing fractures, which can appear as subtle complicated patterns images, requires rigorous model tuning. Image was done using unsharp masking approach contrast-limited adaptive histogram equalization before training model. The augmented images assist feature identification contribute overall current study has performed extensive experiments analyze influence like number epochs learning rate, along with analysis input data. experimental outcome proven that particular hyperparameter combinations, when paired targeted strategies, improve accuracy precision It observed proposed yielded an 0.964 evaluation over statistical classification across raw 0.98 0.95, respectively. In comparison other deep models, empirical clearly demonstrates its superior conventional approaches

Language: Английский

Exploring the impact of hyperparameter and data augmentation in YOLO V10 for accurate bone fracture detection from X-ray images DOI Creative Commons
Parvathaneni Naga Srinivasu, Gorli L. Aruna Kumari,

Sujatha Canavoy Narahari

et al.

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: March 21, 2025

Accurately identifying bone fractures from the X-ray image is essential to prompt timely and appropriate medical treatment. This research explores impact of hyperparameters data augmentation techniques on performance You Only Look Once (YOLO) V10 architecture for fracture detection. While YOLO architectures have been widely employed in object detection tasks, recognizing fractures, which can appear as subtle complicated patterns images, requires rigorous model tuning. Image was done using unsharp masking approach contrast-limited adaptive histogram equalization before training model. The augmented images assist feature identification contribute overall current study has performed extensive experiments analyze influence like number epochs learning rate, along with analysis input data. experimental outcome proven that particular hyperparameter combinations, when paired targeted strategies, improve accuracy precision It observed proposed yielded an 0.964 evaluation over statistical classification across raw 0.98 0.95, respectively. In comparison other deep models, empirical clearly demonstrates its superior conventional approaches

Language: Английский

Citations

0