基于深度序列学习的计算机断层扫描影像颈椎骨折检测
图像与视频处理
2021-02-08 v4 计算机视觉与模式识别
机器学习
摘要
颈椎骨折是一种医疗急症,可能导致永久性瘫痪甚至死亡。对疑似骨折患者通过计算机断层扫描(CT)进行准确诊断对诊疗至关重要。本文中,我们提出一种带有双向长短期记忆(BLSTM)层的深度卷积神经网络(DCNN),用于CT轴位图像中颈椎骨折的自动检测。我们使用一个包含3,666例CT扫描(729例阳性与2,937例阴性)的标注数据集来训练和验证模型。验证结果显示,在平衡(104例阳性与104例阴性)与不平衡(104例阳性与419例阴性)测试数据集上,分类准确率分别为70.92%和79.18%。
引用
@article{arxiv.2010.13336,
title = {Deep Sequential Learning for Cervical Spine Fracture Detection on Computed Tomography Imaging},
author = {Hojjat Salehinejad and Edward Ho and Hui-Ming Lin and Priscila Crivellaro and Oleksandra Samorodova and Monica Tafur Arciniegas and Zamir Merali and Suradech Suthiphosuwan and Aditya Bharatha and Kristen Yeom and Muhammad Mamdani and Jefferson Wilson and Errol Colak},
journal= {arXiv preprint arXiv:2010.13336},
year = {2021}
}
备注
This paper is accepted for presentation at the IEEE International Symposium on Biomedical Imaging (ISBI) 2021