AUTHOR=Prisilla Ardha Ardea , Guo Yue Leon , Jan Yih-Kuen , Lin Chih-Yang , Lin Fu-Yu , Liau Ben-Yi , Tsai Jen-Yung , Ardhianto Peter , Pusparani Yori , Lung Chi-Wen 

TITLE=An approach to the diagnosis of lumbar disc herniation using deep learning models

JOURNAL=Frontiers in Bioengineering and Biotechnology

VOLUME=Volume 11 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/bioengineering-and-biotechnology/articles/10.3389/fbioe.2023.1247112

DOI=10.3389/fbioe.2023.1247112

ISSN=2296-4185

ABSTRACT=1. The YOLOv5x in the dataset with augmentation successfully performed the highest mAP compared to all YOLO models tested, which indicated the model with the highest performance model to detect lumbar disc herniation (LDH). 2. YOLOv5x without augmentation (non-AUG) dataset performed well in detecting LDH in L2-L3, L3-L4, L4-L5, and L5-S1 regions with values above 90%, which showed effectiveness in using a non-AUG dataset for training. 3. YOLOv5x showed the shortest training duration and the lightest weight compared to YOLOv6 and YOLOv7, indicating the most efficient LDH detection model.