2017/11/23 by Pulkit Kumar, Kumar, Pulkit, Monika Grewal +3 · 1 citation
Medicine · #COVID-19 diagnosis using AI #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Lung Cancer Diagnosis and Treatment #Radiomics and Machine Learning in Medical Imaging
paper · pdf · doi:10.48550/arxiv.1711.08760
openalex publication_date 2017/11/23 · openalex created_date 2022/10/02 · openalex updated_date 2026/07/28
Chest X-ray is one of the most accessible medical imaging technique for\ndiagnosis of multiple diseases. With the availability of ChestX-ray14, which is\na massive dataset of chest X-ray images and provides annotations for 14\nthoracic diseases; it is possible to train Deep Convolutional Neural Networks\n(DCNN) to build Computer Aided Diagnosis (CAD) systems. In this work, we\nexperiment a set of deep learning models and present a cascaded deep neural\nnetwork that can diagnose all 14 pathologies better than the baseline and is\ncompetitive with other published methods. Our work provides the quantitative\nresults to answer following research questions for the dataset: 1) What loss\nfunctions to use for training DCNN from scratch on ChestX-ray14 dataset that\ndemonstrates high class imbalance and label co occurrence? 2) How to use\ncascading to model label dependency and to improve accuracy of the deep\nlearning model?\n