{"title":"A deep learning based fracture detection in arm bone X-ray images","authors":"Hoai Phuong Nguyen, T. Hoang, Huy Hoang Nguyen","doi":"10.1109/MAPR53640.2021.9585292","DOIUrl":null,"url":null,"abstract":"A large number of arm fracture-related injuries are reported in hospitals and clinics around the world. In this paper, we propose a novel deep learning based fracture detection in arm bone X-ray images. First, we preprocess the Xray image by using an algorithm that is a combination of the YOLACT++ for image segmentation and Contrast Limited Adaptive Histogram Equalization for image contrast enhancement. Then, YOLOv4 is trained on a small dataset with four data augmentation techniques to identify and locate the position of bone fracture on X-ray images. The topmost result obtained is 81.91% by using our proposed method. Experimental results also confirm that our method outperforms the Faster-RCNN based solution while implementing on the small dataset.","PeriodicalId":233540,"journal":{"name":"2021 International Conference on Multimedia Analysis and Pattern Recognition (MAPR)","volume":"95 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"8","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2021 International Conference on Multimedia Analysis and Pattern Recognition (MAPR)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/MAPR53640.2021.9585292","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 8
Abstract
A large number of arm fracture-related injuries are reported in hospitals and clinics around the world. In this paper, we propose a novel deep learning based fracture detection in arm bone X-ray images. First, we preprocess the Xray image by using an algorithm that is a combination of the YOLACT++ for image segmentation and Contrast Limited Adaptive Histogram Equalization for image contrast enhancement. Then, YOLOv4 is trained on a small dataset with four data augmentation techniques to identify and locate the position of bone fracture on X-ray images. The topmost result obtained is 81.91% by using our proposed method. Experimental results also confirm that our method outperforms the Faster-RCNN based solution while implementing on the small dataset.