Recently, VinBigdata’s research on Medical Imaging has been published by Neurocomputing, one of the most cited Artificial Intelligence (AI) journals based on Google’s Scholar Metrics.
This paper presents a supervised multi-label classification framework based on deep convolutional neural networks (CNNs) for predicting the presence of 14 common thoracic diseases and observations. Our scientists tackle this problem by training state-of-the-art CNNs that exploit hierarchical dependencies among abnormality labels. We also propose to use the label smoothing technique for a better handling of uncertain samples, which occupy a significant portion of almost every CXR dataset.
Our model is trained on over 200,000 CXRs of the recently released CheXpert dataset and achieves a mean area under the curve (AUC) of 0.940 in predicting 5 selected pathologies from the validation set. This is the highest AUC score yet reported to date. The proposed method is also evaluated on the independent test set of the CheXpert competition, which is composed of 500 CXR studies annotated by a panel of 5 experienced radiologists. The performance is on average better than 2.6 out of 3 other individual radiologists with a mean AUC of 0.930, which ranks first on the CheXpert leaderboard at the time of writing this paper.
The paper is an achievement of scientific team of Medical Image processing department, including PhD. Nguyen Quy Ha (University of Illinois, USA), PhD. Pham Huy Hieu (University of Toulouse, France), Tran Quang Dat, Ngo Thanh Dat, Le Tung. Currently, the group of scientists and experts in the Medical Image Processing Department are continuing to develop and complete VinDr – a comprehensive AI solution for medical imaging diagnostics. The main goal is to assist radiologists or clinicians in making fast and accurate diagnosis; therefore improving the patient care and facilitating effective clinical workflows in Vietnam.
Previously, the Medical Imaging team has affirmed their scientific capacity in many international arenas, including: No. 01 in CheXpert competition organized by Stanford University in 2019; No. 01 in Abnormal Image
Detection in Endoscopy Videos (EndoCV), 2020; top 03 in Pulmonary Embolism Detection Challenge, organized by the Radiological, Society of North America (RSNA), 2020.
Learn more about VinBigdata’s scientific paper here.