mlsanityData to 14 Oct 2026

A novel multiple instance learning framework for COVID-19 severity assessment via data augmentation and self-supervised learning

arXiv 2102.03837 ↗First posted 7 Feb 2021eess.IV cs.CVcited by 19

Not on the map. Why?

Abstract (arXiv)

How to fast and accurately assess the severity level of COVID-19 is an essential problem, when millions of people are suffering from the pandemic around the world. Currently, the chest CT is regarded as a popular and informative imaging tool for COVID-19 diagnosis. However, we observe that there are two issues -- weak annotation and insufficient data that may obstruct automatic COVID-19 severity assessment with CT images. To address these challenges, we propose a novel three-component method, i.e., 1) a deep multiple instance learning component with instance-level attention to jointly classify the bag and also weigh the instances, 2) a bag-level data augmentation component to generate virtual bags by reorganizing high confidential instances, and 3) a self-supervised pretext component to aid the learning process. We have systematically evaluated our method on the CT images of 229 COVID-19 cases, including 50 severe and 179 non-severe cases. Our method could obtain an average accuracy of 95.8%, with 93.6% sensitivity and 96.4% specificity, which outperformed previous works.

Key facts. Cited by 19 papers in the map (data through 14 Oct 2026). Builds on 25 of its 47 references that are in the map. First posted on arXiv 7 Feb 2021.

Builds on

25 of 47 references.

References · 25

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  3. 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,513
  4. 2016Densely Connected Convolutional Networkscited by 6,720
  5. 2017mixup: Beyond Empirical Risk Minimizationcited by 5,183
  6. 2017Improved Regularization of Convolutional Neural Networks with Cutoutcited by 1,628
  7. 2018Unsupervised Representation Learning by Predicting Image Rotationscited by 1,591
  8. 2016Context Encoders: Feature Learning by Inpaintingcited by 1,428
  9. 2016Unsupervised Learning of Visual Representations by Solving Jigsaw Puzzlescited by 1,382
  10. 2017Random Erasing Data Augmentationcited by 1,239
  11. 2016Colorful Image Colorizationcited by 1,114
  12. 2018Attention-based Deep Multiple Instance Learningcited by 1,095
13 more
  1. 2015Unsupervised Visual Representation Learning by Context Predictioncited by 1,066
  2. 2017The Effectiveness of Data Augmentation in Image Classification using Deep Learningcited by 370
  3. 2016Learning Representations for Automatic Colorizationcited by 251
  4. 2018Not-so-supervised: a survey of semi-supervised, multi-instance, and transfer learning in medical image analysiscited by 179
  5. 2015Patch-based Convolutional Neural Network for Whole Slide Tissue Image Classificationcited by 159
  6. 2019Boosting Few-Shot Visual Learning with Self-Supervisioncited by 140
  7. 2018Synthetic Data Augmentation using GAN for Improved Liver Lesion Classificationcited by 136
  8. 2017Smart Augmentation - Learning an Optimal Data Augmentation Strategycited by 54
  9. 2017Exploiting the potential of unlabeled endoscopic video data with self-supervised learningcited by 33
  10. 2020Severity Assessment of Coronavirus Disease 2019 (COVID-19) Using Quantitative Features from Chest CT Imagescited by 8
  11. 2020Diversity Helps: Unsupervised Few-shot Learning via Distribution Shift-based Data Augmentationcited by 7
  12. 2016Multiple Instance Learning Convolutional Neural Networks for Object Recognitioncited by 6
  13. 2016Semi-supervised Learning using Denoising Autoencoders for Brain Lesion Detection and Segmentationcited by 6

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2021: 2; 2022: 6; 2023: 7; 2024: 0; 2025: 2; 2026: 2.

19 papers.

Sort:

  1. 2022Multiple Instance Learning for Digital Pathology: A Review on the State-of-the-Art, Limitations & Future Potentialcited by 54
  2. 2022Anomaly Detection-Inspired Few-Shot Medical Image Segmentation Through Self-Supervision With Supervoxelscited by 42
  3. 2023Label-Efficient Deep Learning in Medical Image Analysis: Challenges and Future Directionscited by 16
  4. 2023A Review of Predictive and Contrastive Self-supervised Learning for Medical Imagescited by 10
  5. 2021Better Pseudo-label: Joint Domain-aware Label and Dual-classifier for Semi-supervised Domain Generalizationcited by 9
  6. 2022MixUp-MIL: Novel Data Augmentation for Multiple Instance Learning and a Study on Thyroid Cancer Diagnosiscited by 9
  7. 2022Temporal Context Matters: Enhancing Single Image Prediction with Disease Progression Representationscited by 8
  8. 2023MixBag: Bag-Level Data Augmentation for Learning from Label Proportionscited by 8
  9. 2023Data efficient deep learning for medical image analysis: A surveycited by 6
  10. 2023Attention2Minority: A salient instance inference-based multiple instance learning for classifying small lesions in whole slide imagescited by 5
  11. 2021Information Bottleneck Attribution for Visual Explanations of Diagnosis and Prognosiscited by 4
  12. 2022Lung Swapping Autoencoder: Learning a Disentangled Structure-texture Representation of Chest Radiographscited by 4
7 more
  1. 2023Learning Self-Supervised Representations for Label Efficient Cross-Domain Knowledge Transfer on Diabetic Retinopathy Fundus Imagescited by 2
  2. 2025Learning from Majority Label: A Novel Problem in Multi-class Multiple-Instance Learningcited by 2
  3. 2023MixUp-MIL: A Study on Linear & Multilinear Interpolation-Based Data Augmentation for Whole Slide Image Classificationcited by 1
  4. 2022Dive into Self-Supervised Learning for Medical Image Analysis: Data, Models and Taskscited by 0
  5. 2025Multi-instance Learning as Downstream Task of Self-Supervised Learning-based Pre-trained Modelcited by 0
  6. 2026Knowing When Not to Predict: Self Supervised Learning and Abstention for Safer DR Screeningcited by 0
  7. 2026Towards Whole-Study Screening for Congenital Heart Disease in Fetal Ultrasound Using Multiple Instance Learningcited by 0