Paper Info
Title
Semi-Supervised Active Learning For Covid-19 Lung Ultrasound Multi-Symptom Classification
Abstract
Ultrasound (US) is a non-invasive yet effective medical diagnostic imaging technique for the COVID-19 global pandemic. However, due to complex feature behaviors and expensive annotations of US images, it is difficult to apply Artificial Intelligence (AI) assisting approaches for the lung's multi-symptom (multi-label) classification. To overcome these difficulties, we propose a novel semi-supervised Two-Stream Active Learning (TSAL) method to model complicated features and reduce labeling costs in an iterative manner. The core component of TSAL is the multi-label learning mechanism, in which label correlation information is used to design a multi-label margin (MLM) strategy and a confidence validation for automatically selecting informative samples and confident labels. In this framework, a multi-symptom multi-label (MSML) classification network is proposed to learn discriminative features of lung symptoms, and a human-machine interaction (HMI) is exploited to confirm the final annotations that are used to fine-tune MSML. Moreover, a novel lung US dataset named COVID19-LUSMS is built, currently containing 71 clinical patients with 6,836 images sampled from 678 videos. Experimental evaluations show that TSAL can achieve superior performance to the baseline and the state-of-the-art using only 20% data. Qualitatively, visualization of the attention map confirms a good consistency between the model prediction and the clinical knowledge.
Year
DOI
Venue
2020
10.1109/ICTAI50040.2020.00191
2020 IEEE 32ND INTERNATIONAL CONFERENCE ON TOOLS WITH ARTIFICIAL INTELLIGENCE (ICTAI)
Keywords
DocType
ISSN
COVID-19, Ultrasound Imaging, Multi-Label Classification, Active Learning, Semi-Supervised Learning
Conference
1082-3409
Citations 
PageRank 
References 
1
0.35
0
Authors
6
Name
Order
Citations
PageRank
Liu Lei113.06
Wentao Lei210.35
Xiang Wan312.04
Li Liu411.03
Yongfang Luo510.69
Cheng Feng610.35