Intelligent Masking: Deep Q-Learning for Context Encoding in Medical Image Analysis

被引:0
|
作者
Bahrami, Mojtaba [1 ,2 ]
Ghorbani, Mahsa [1 ,2 ]
Yeganeh, Yousef [1 ]
Navab, Nassir [1 ,3 ]
机构
[1] Tech Univ Munich, Comp Aided Med Procedures, Munich, Germany
[2] Sharif Univ Technol, Tehran, Iran
[3] Johns Hopkins Univ, Whiting Sch Engn, Baltimore, MD USA
来源
MACHINE LEARNING IN MEDICAL IMAGING, MLMI 2022 | 2022年 / 13583卷
关键词
Self-supervised learning; Context encoding; Reinforcement learning; Medical image analysis;
D O I
10.1007/978-3-031-21014-3_40
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
The need for a large amount of labeled data in the supervised setting has led recent studies to utilize self-supervised learning to pre-train deep neural networks using unlabeled data. Many self-supervised training strategies have been investigated especially for medical datasets to leverage the information available in the much fewer unlabeled data. One of the fundamental strategies in image-based self-supervision is context prediction. In this approach, a model is trained to reconstruct the contents of an arbitrary missing region of an image based on its surroundings. However, the existing methods adopt a random and blind masking approach by focusing uniformly on all regions of the images. This approach results in a lot of unnecessary network updates that cause the model to forget the rich extracted features. In this work, we develop a novel self-supervised approach that occludes targeted regions to improve the pre-training procedure. To this end, we propose a reinforcement learning-based agent which learns to intelligently mask input images through deep Q-learning. We show that training the agent against the prediction model can significantly improve the semantic features extracted for downstream classification tasks. We perform our experiments on two public datasets for diagnosing breast cancer in the ultrasound images and detecting lower-grade glioma with MR images. In our experiments, we show that our novel masking strategy advances the learned features according to the performance on the classification task in terms of accuracy, macro F1, and AUROC.
引用
收藏
页码:387 / 396
页数:10
相关论文
共 50 条
  • [41] Deep Q-Learning with Multiband Sensing for Dynamic Spectrum Access
    Nguyen, Ha Q.
    Nguyen, Binh T.
    Dong, Trung Q.
    Ngo, Dat T.
    Nguyen, Tuan A.
    2018 IEEE INTERNATIONAL SYMPOSIUM ON DYNAMIC SPECTRUM ACCESS NETWORKS (DYSPAN), 2018,
  • [42] Meta-Deep Q-Learning for Eco-Routing
    Ma, Xin
    Xie, Yuanchang
    Chigan, Chunxiao
    2019 IEEE 2ND CONNECTED AND AUTOMATED VEHICLES SYMPOSIUM (CAVS), 2019,
  • [43] Trend following deep Q-Learning strategy for stock trading
    Chakole, Jagdish
    Kurhekar, Manish
    EXPERT SYSTEMS, 2020, 37 (04)
  • [44] A Novel Deep Q-learning Method for Dynamic Spectrum Access
    Tomovic, S.
    Radusinovic, I
    2020 28TH TELECOMMUNICATIONS FORUM (TELFOR), 2020, : 9 - 12
  • [45] On the Effective Transfer Learning Strategy for Medical Image Analysis in Deep Learning
    Wen, Yang
    Chen, Leiting
    Zhou, Chuan
    Deng, Yu
    Zeng, Huiru
    Xi, Shuo
    Guo, Rui
    2020 IEEE INTERNATIONAL CONFERENCE ON BIOINFORMATICS AND BIOMEDICINE, 2020, : 827 - 834
  • [46] SMAC-tuned Deep Q-learning for Ramp Metering
    ElSamadisy, Omar
    Abdulhai, Yazeed
    Xue, Haoyuan
    Smirnov, Ilia
    Khalil, Elias B.
    Abdulhai, Baher
    2023 IEEE INTERNATIONAL CONFERENCE ON SMART MOBILITY, SM, 2023, : 65 - 72
  • [47] Deep Q-Learning using Redundant Outputs in Visual Doom
    Park, Hyunsoo
    Kim, Kyung-Joong
    2016 IEEE CONFERENCE ON COMPUTATIONAL INTELLIGENCE AND GAMES (CIG), 2016,
  • [48] An adaptive deep Q-learning strategy for handwritten digit recognition
    Qiao, Junfei
    Wang, Gongming
    Li, Wenjing
    Chen, Min
    NEURAL NETWORKS, 2018, 107 : 61 - 71
  • [49] Deep Q-Learning for Illumination and Rotation Invariant Face Detection
    Ruiz-Garcia, Ariel
    Palade, Vasile
    Almakky, Ibrahim
    Elshaw, Mark
    2019 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS (IJCNN), 2019,
  • [50] Deep Q-Learning Based Reinforcement Learning Approach for Network Intrusion Detection
    Alavizadeh, Hooman
    Alavizadeh, Hootan
    Jang-Jaccard, Julian
    COMPUTERS, 2022, 11 (03)