Semi-supervised classification of radiology images with NoTeacher: A teacher that is not mean

Unnikrishnan, Balagopal; Nguyen, Cuong Manh; Balaram, Shafa; Li, Chao; Foo, Chuan Sheng; Krishnaswamy, Pavitra

doi:10.1016/j.media.2021.102148

Cited by 9 publications

(7 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For the results on Chest X-Ray14 in Table 1, our method, NoTeacher [36], UPS [28], and S 2 MTS 2 [25] use the DenseNet-121 backbone, while SRC-MT [26] and GraphXNet [2] use DenseNet-169 [12]. SRC-MT [26] is a consistency-based SSL; NoTeacher [36] extends MT by replacing the EMA process with two networks combined with a probabilistic graph model; S 2 MTS 2 [25] combines self-supervised pre-training with MT fine-tuning; and GraphXNet [2] constructs a graph from dataset samples and assigns pseudo labels to unlabelled samples through label propagation; and UPS [28] applies probability and uncertainty thresholds to enable the pseudo labelling of unlabelled samples. All methods use the official test set [38].…”

Section: Thorax Disease Classification Resultsmentioning

confidence: 99%

“…The main benchmarks for SSL in MIA study the multilabel classification of chest X-ray (CXR) images [13,38] and multi-class classification of skin lesions [8,35]. For CXR SSL classification, pseudo-labelling methods have been explored [2], but SOTA results are achieved with consistency learning approaches [9,23,25,26,36]. For skin lesion SSL classification, the current SOTA is also based on consistency learning [26], with pseudo-labelling approaches [3] not being competitive.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

ACPL: Anti-curriculum Pseudo-labelling for Semi-supervised Medical Image Classification

Liu¹,

Tian²,

Chen³

et al. 2021

Preprint

View full text Add to dashboard Cite

Effective semi-supervised learning (SSL) in medical image analysis (MIA) must address two challenges: 1) work effectively on both multi-class (e.g., lesion classification) and multi-label (e.g., multiple-disease diagnosis) problems, and 2) handle imbalanced learning (because of the high variance in disease prevalence). One strategy to explore in SSL MIA is based on the pseudo labelling strategy, but it has a few shortcomings. Pseudo-labelling has in general lower accuracy than consistency learning, it is not specifically design for both multi-class and multi-label problems, and it can be challenged by imbalanced learning. In this paper, unlike traditional methods that select confident pseudo label by threshold, we propose a new SSL algorithm, called anti-curriculum pseudo-labelling (ACPL), which introduces novel techniques to select informative unlabelled samples, improving training balance and allowing the model to work for both multi-label and multi-class problems, and to estimate pseudo labels by an accurate ensemble of classifiers (improving pseudo label accuracy). We run extensive experiments to evaluate ACPL on two public medical image classification benchmarks: Chest X-Ray14 for thorax disease multi-label classification and ISIC2018 for skin lesion multi-class classification. Our method outperforms previous SOTA SSL methods on both datasets. (b) Imbalanced distribution on multi-label Chest X-ray14 [38] (left) and multi-class ISIC2018 [35] (right)

show abstract

Section: Thorax Disease Classification Resultsmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

ACPL: Anti-curriculum Pseudo-labelling for Semi-supervised Medical Image Classification

Liu¹,

Tian²,

Chen³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…In our study, only ACL tears are considered as positive samples and other abnormal or normal patients are considered as negative samples. We train and test our presented model using the publicly available training set (1,130 scans) and test set (120 scans) with no patient overlaps, and the split refers to previous work ( Unnikrishnan et al, 2021 ; Azcona et al, 2020 ). Meanwhile, of the 1,130 scans available for training, we randomly sample 20% as the validation set.…”

Section: Methodsmentioning

confidence: 99%

Automated diagnosis of anterior cruciate ligament via a weighted multi-view network

Li,

Zhai,

Yang

et al. 2023

Front. Bioeng. Biotechnol.

View full text Add to dashboard Cite

Objective: To build a three-dimensional (3D) deep learning-based computer-aided diagnosis (CAD) system and investigate its applicability for automatic detection of anterior cruciate ligament (ACL) of the knee joint in magnetic resonance imaging (MRI).Methods: In this study, we develop a 3D weighted multi-view convolutional neural network by fusing different views of MRI to detect ACL. The network is evaluated on two MRI datasets, the in-house MRI-ACL dataset and the publicly available MRNet-v1.0 dataset. In the MRI-ACL dataset, the retrospective study collects 100 cases, and four views per patient are included. There are 50 ACL patients and 50 normal patients, respectively. The MRNet-v1.0 dataset contains 1,250 cases with three views, of which 208 are ACL patients, and the rest are normal or other abnormal patients.Results: The area under the receiver operating characteristic curve (AUC) of the ACL diagnosis system is 97.00% and 92.86% at the optimal threshold for the MRI-ACL dataset and the MRNet-v1.0 dataset, respectively, indicating a high overall diagnostic accuracy. In comparison, the best AUC of the single-view diagnosis methods are 96.00% (MRI-ACL dataset) and 91.78% (MRNet-v1.0 dataset), and our method improves by about 1.00% and 1.08%. Furthermore, our method also improves by about 1.00% (MRI-ACL dataset) and 0.28% (MRNet-v1.0 dataset) compared with the multi-view network (i.e., MRNet).Conclusion: The presented 3D weighted multi-view network achieves superior AUC in diagnosing ACL, not only in the in-house MRI-ACL dataset but also in the publicly available MRNet-v1.0 dataset, which demonstrates its clinical applicability for the automatic detection of ACL.

show abstract

“…NoTeacher (NoT) [52]: In the Mean Teacher, the consistency target, which is the teacher, relies on the EMA of the student. In other words, the teacher's weight is an ensemble of student weights.…”

Section: Consistency-based Methodsmentioning

confidence: 99%

Survey on Recent Trends in Medical Image Classification Using Semi-Supervised Learning

Solatidehkordi

Zualkernan

2022

Applied Sciences

View full text Add to dashboard Cite

Training machine learning and deep learning models for medical image classification is a challenging task due to a lack of large, high-quality labeled datasets. As the labeling of medical images requires considerable time and effort from medical experts, models need to be specifically designed to train on low amounts of labeled data. Therefore, an application of semi-supervised learning (SSL) methods provides one potential solution. SSL methods use a combination of a small number of labeled datasets with a much larger number of unlabeled datasets to achieve successful predictions by leveraging the information gained through unsupervised learning to improve the supervised model. This paper provides a comprehensive survey of the latest SSL methods proposed for medical image classification tasks.

show abstract

Semi-supervised classification of radiology images with NoTeacher: A teacher that is not mean

Cited by 9 publications

References 14 publications

ACPL: Anti-curriculum Pseudo-labelling for Semi-supervised Medical Image Classification

ACPL: Anti-curriculum Pseudo-labelling for Semi-supervised Medical Image Classification

Automated diagnosis of anterior cruciate ligament via a weighted multi-view network

Survey on Recent Trends in Medical Image Classification Using Semi-Supervised Learning

Contact Info

Product

Resources

About