The DOI matches its Crossref record; 74 of the 74 references it cites were found in a registry or an index; none of the 23 sources its claims cite is checked for support yet.
Self-supervised learning for medical image analysis using image context restoration
Open accessAbstract
Machine learning, particularly deep learning has boosted medical image analysis over the past years. Training a good model based on deep learning requires large amount of labelled data. However, it is often difficult to obtain a sufficient number of labelled images for training. In many scenarios the dataset in question consists of more unlabelled images than labelled ones. Therefore, boosting the performance of machine learning models by using unlabelled as well as labelled data is an important but challenging problem. Self-supervised learning presents one possible solution to this problem. However, existing self-supervised learning strategies applicable to medical images cannot result in significant performance improvement. Therefore, they often lead to only marginal improvements. In this paper, we propose a novel self-supervised learning strategy based on context restoration in order to better exploit unlabelled images. The context restoration strategy has three major features: 1) it learns semantic image features; 2) these image features are useful for different types of subsequent image analysis tasks; and 3) its implementation is simple. We validate the context restoration strategy in three common problems in medical imaging: classification, localization, and segmentation. For classification, we apply and test it to scan plane detection in fetal 2D ultrasound images; to localise abdominal organs in CT images; and to segment brain tumours in multi-modal MR images. In all three cases, self-supervised learning based on context restoration learns useful semantic features and lead to improved machine learning models for the above tasks.
Overview
AI snapshot
Written from the paper’s full text.
TL;DR
The paper proposes context restoration as a self-supervised pretraining task for medical images: randomly swap image patches and train a CNN to restore the original context. Across fetal ultrasound classification, abdominal CT organ localization, and brain MR tumor segmentation, context restoration pretraining outperformed autoencoder, relative-position, context-prediction, and no self-supervision.
Key findings
- Context restoration pretraining improved CNNs in all three evaluated tasks: fetal standard scan plane classification, abdominal multi-organ localization, and brain tumor segmentation. It outperformed other self-supervised strategies and no self-supervision in the reported comparisons.
- For fetal ultrasound classification, when using only a quarter of the training data, context restoration pretraining improved the SonoNet performance the most among the compared pretraining strategies.
- For abdominal CT localization, context restoration pretraining produced more localization improvements than other pretraining methods; in some cases it was comparable to or better than no pretraining on more annotated training data.
- For BraTS 2017 brain tumor segmentation, U-Nets initialized by context restoration pretraining achieved the best performance in total; using 50% of the training set gave similar performance to using the whole training set, with Dice in enhanced tumor core, sensitivity in non-enhanced and enhanced tumor cores, and Hausdorff distances even slightly better.
Methods
- Self-supervised task: given an unlabelled image, randomly select two isolated patches and swap them; repeat T times to disorder context, then train a CNN to restore the original image using an L2 loss.
- Evaluated on three datasets/tasks: fetal 2D ultrasound classification with 2694 examinations and SonoNet-64; abdominal CT multi-organ localization with 150 subjects and a SonoNet-like CNN; BraTS 2017 brain tumor segmentation with 285 subjects, 142 for training/validation and 143 for testing, using a 2D U-Net on 64×64 patches.
- Compared pretraining strategies: autoencoder, relative position prediction, local context prediction, and context restoration; subsequent tasks used whole, half, or quarter of the training set. Training details include 58K iterations for ultrasound classification, 55K for CT localization, and 50K for segmentation.
Limitations
- The paper states that the L2 loss used for context restoration results in blurry restored images.
- The paper reports that data augmentation did not consistently improve performance and sometimes decreased segmentation accuracy; it also notes that self-supervised pretraining was not compared against supervised pretraining because supervised pretraining is difficult in medical imaging due to annotation requirements.
- Evident from the text: evaluation is limited to three medical imaging tasks and datasets, and the paper notes that in brain tumor segmentation the performance differences among self-supervision strategies were not always significant.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
26 Own interpretation · 25 Own result · 21 No basis given · 20 Not checked yet
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Balancing the numbers of instances in each class significantly improves the performance in all three metrics.”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Fetal standard scan plane classification
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Initialising by pretrained features, particularly those from context restoration tasks, improves the CNN performance.”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Abdominal multi-organ localization
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“A shortcoming is that the L2 loss results in image blur.”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Context restoration results
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In all three cases, self-supervised learning based on context restoration learns useful semantic features and lead to improved machine learning models for the above tasks.”
§ Self-supervised learning for medical image analysis using image context restoration > Abstract
plaindirectn = 3- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“On quarter of training data, the data augmentation techniques can be advantageous over some self-supervised learning methods such as the CP method.”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Fetal standard scan plane classification
plainconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“when learning on half of training images, the F1-scores keep stable in most cases except where the SonoNet is pretrained based on context restoration.”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Fetal standard scan plane classification
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Our context restoration pretraining improves the SonoNet performance the most.”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Fetal standard scan plane classification
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In all cases, the image context restoration achieve qualitatively good results.”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Context restoration results
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Data augmentation did not offer performance improvement in many cases.”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Abdominal multi-organ localization
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“using 50% of the training set, the proposed self-supervision strategy offers similar performance to using the whole training set.”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Brain tumour segmentation
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- It may state two separate claims as one. 18
- The quote only partly states this claim. 1
- Probably not a substantial claim. 21
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“when the SonoNet is trained on quarter of the training data, the precision decreases significantly while there is only slight decrease in terms of the recall.”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Fetal standard scan plane classification
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In all three tasks, the pretraining based on our context restoration strategy is superior to other self-supervised learning strategies, as well as no self-supervised training.”
§ Self-supervised learning for medical image analysis using image context restoration > Introduction
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“When the SonoNet is trained on half of the training data, the precision and recall both decrease, which lead to the decrease of the F1-score.”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Fetal standard scan plane classification
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Data augmentation does not lead to improvement when the SonoNet is trained on half data; while it lead to obvious improvement when the SonoNet is trained on quarter of data.”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Fetal standard scan plane classification
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“The SonoNet pretrained using context restoration can only offers marginal improvement.”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Fetal standard scan plane classification
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
The quote only partly states this claim.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“When learning using only a quarter of training images, the SonoNet with feature initialisation from the auto-encoder pretraining still cannot improve the baseline; while SonoNets using other pretraining strategies perform better than the baseline.”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Fetal standard scan plane classification
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“in terms of the liver, the CNN performance decreases sharply with half training data; while it remains stable with quarter training data.”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Abdominal multi-organ localization
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Although the CNN pretrained using the RP method on quarter training data outperforms that using context restoration pretraining, it performs much worse in left kidney.”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Abdominal multi-organ localization
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“The Dice score in enhanced tumour core, the sensitivity in non-enhanced and enhanced tumour cores, and the Hausdorff distances in all aspects are even slightly better.”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Brain tumour segmentation
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“pretraining based on relative positions and context prediction tasks improves the segmentations but they are not as good as the pretraining based on the context restoration task.”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Brain tumour segmentation
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“For many common problems in medical imaging such as classification, localization, and segmentation, learning image context is key.”
§ Self-supervised learning for medical image analysis using image context restoration > Self-supervision based on context restoration > Context restoration
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper may give no reason why it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In case of subsequent segmentation tasks, the whole CNN architecture can be shared by the pretraining network and the subsequent segmentation network. As a result, almost all the weights of the subsequent segmentation CNN can be initialised using those learned in the self-supervised pretraining. This results in better segmentation results.”
§ Self-supervised learning for medical image analysis using image context restoration > Self-supervision based on context restoration > Network architectures
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
In the paper
“However, the removal of context changes the image intensity distribution. Thus the resulting images belong to another domain and the learned features may not be useful for images in the original domain.”
§ Self-supervised learning for medical image analysis using image context restoration > Related work
What the sources say
- Not checked yetcites Pathak et al., 2016
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“The pretrained models from the natural image domain are not useful in the medical imaging domain since the intensity distribution of natural images is different from that of medical images.”
§ Self-supervised learning for medical image analysis using image context restoration > Introduction
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper may give no reason why it holds.
In the paper
“Since each data instance is regarded as one class, the exemplar learning method is difficult to apply to large datasets.”
§ Self-supervised learning for medical image analysis using image context restoration > Related work
What the sources say
- Not checked yetcites Dosovitskiy et al., 2016 source
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“More precisely, the CNN weights learned in the CP method can be used to initialise subsequent CNNs for classification and segmentation; while CNN weights learned in the RP method can initialise subsequent classification CNNs and only the analysis part of the subsequent segmentation CNNs.”
§ Self-supervised learning for medical image analysis using image context restoration > Related work
What the sources say
- Not checked yetcites Pathak et al., 2016
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
In the paper
“In case of subsequent classification tasks, simple CNN layers with a few deconvolution layers are sufficient (see Fig. 3) (Doersch and Zisserman, 2017). In this condition, the analysis part makes most contributions to the context restoration. Therefore, the feature maps resulting from the analysis part are more useful.”
§ Self-supervised learning for medical image analysis using image context restoration > Self-supervision based on context restoration > Network architectures
What the sources say
- Not checked yetcites Doersch and Zisserman, 2017
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“This may because the augmented images do not cover the variance of organ shapes and positions and introduce additional variance to pixel intensities.”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Abdominal multi-organ localization
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“The disadvantage of multi-task self-supervised learning is the training requires significant computational resources, i.e. 64 GPUs for approximately 16.8K GPU hours.”
§ Self-supervised learning for medical image analysis using image context restoration > Related work
What the sources say
- Not checked yetcites Doersch and Zisserman, 2017
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“In many medical image analysis problems, localization anatomical structures is a prerequisite.”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Abdominal multi-organ localization
What the sources say
- Not checked yetcites Heimann et al., 2009 source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim.
In the paper
“As suggested by Pathak et al. (2016), the L2 loss is sufficient for feature learning although the outputs from context restoration outputs may be blurry.”
§ Self-supervised learning for medical image analysis using image context restoration > Self-supervision based on context restoration > Network architectures
What the sources say
- Not checked yetcites Pathak et al., 2016
hedgeddirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In this scenario, the baseline (i.e. random initialisation) is not far away from the ceiling (i.e. SonoNet on the whole training set). Therefore, it is difficult to obtain improvements.”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Fetal standard scan plane classification
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
In the paper
“de Vos et al. (2017) proposed a novel approach which can localize anatomical structures in 3D medical images.”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Abdominal multi-organ localization
What the sources say
- Not checked yetcites de Vos et al., 2017 source
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim.
In the paper
“The results showed that multitask self-supervised learning improves subsequent tasks more than single-task self-supervised learning.”
§ Self-supervised learning for medical image analysis using image context restoration > Related work
What the sources say
- Not checked yetcites Doersch and Zisserman, 2017
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
None of the sources cited for this claim have been checked yet.
In the paper
“However, existing self-supervised learning strategies applicable to medical images cannot result in significant performance improvement. Therefore, they often lead to only marginal improvements.”
§ Self-supervised learning for medical image analysis using image context restoration > Abstract
What the sources say
- Not checked yetcites Doersch et al., 2015
- Not checked yetcites Pathak et al., 2016
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“the CNNs for context restoration can be structured in flexible architectures depending on subsequent tasks.”
§ Self-supervised learning for medical image analysis using image context restoration > Discussion and conclusion
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
In the paper
“They are routinely assessed using MR imaging (Bakas et al., 2017b).”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Brain tumour segmentation
What the sources say
- Not checked yetcites Bakas et al., 2017b
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
None of the sources cited for this claim have been checked yet.
In the paper
“In computer vision, many CNNs are pretrained before the main task.”
§ Self-supervised learning for medical image analysis using image context restoration > Discussion and conclusion
What the sources say
- Not checked yetcites Ren et al., 2015
- Not checked yetcites Simonyan and Zis-serman, 2014
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“the implementation of the context restoration task is simple and straightforward, meaning that it can be widely used.”
§ Self-supervised learning for medical image analysis using image context restoration > Discussion and conclusion
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: fails — It may be too vague to check against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. It may be too vague to check against evidence. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“2) The augmentation does not introduce any more useful variance in terms of tumour appearance.”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Brain tumour segmentation
plainconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“3) The augmentation introduces additional intensity variations which are not meaningful.”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Brain tumour segmentation
plainconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
None of the sources cited for this claim have been checked yet.
In the paper
“the Faster R-CNN (Ren et al., 2015) is based on the pretraining of the VGG-Net (Simonyan and Zis-serman, 2014).”
§ Self-supervised learning for medical image analysis using image context restoration > Discussion and conclusion
What the sources say
- Not checked yetcites Ren et al., 2015
- Not checked yetcites Simonyan and Zis-serman, 2014
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“which may because of three reasons: 1) The segmentation is based on patches so a large number of image patches can be sampled.”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Brain tumour segmentation
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“We conclude from the existing self-supervised feature learning literature that the ideal pretraining task should have similar goal to the subsequent task.”
§ Self-supervised learning for medical image analysis using image context restoration > Discussion and conclusion
plaincorrelational- Asserted: passes — The authors put it forward as true.
- Checkable: fails — It may be too vague to check against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. It may be too vague to check against evidence. The paper may give no reason why it holds.
In the paper
“This type of pretraining leads to good detection results in the Faster R-CNN.”
§ Self-supervised learning for medical image analysis using image context restoration > Discussion and conclusion
What the sources say
- Not checked yetcites Ren et al., 2015
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“It is noteworthy that if less training data leads to significant decrease of results, self-supervised learning is likely to improve the results significantly.”
§ Self-supervised learning for medical image analysis using image context restoration > Experiments and results > Abdominal multi-organ localization
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“if the reduction of training data causes significant performance decrease, the context restoration pretraining can offer significant performance improvement over the baselines.”
§ Self-supervised learning for medical image analysis using image context restoration > Discussion and conclusion
plainconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Compared with the existing strategies such as relative positions and context prediction, solving the context restoration task requires pattern recognition and prediction, which ensures the context restoration task offers more efficient image semantics.”
§ Self-supervised learning for medical image analysis using image context restoration > Discussion and conclusion
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: fails — It may be too vague to check against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. It may be too vague to check against evidence. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
The paper relies on this without support.
Training CNNs only using a small number of labelled images cannot always achieve satisfactory results and does not exploit the potentially large number of unlabelled images that may be available.
“Training CNNs only using the small number of labelled images cannot always achieve satisfactory results and does not exploit the potentially large number of unlabelled images that may be available.”
The paper relies on this without support.
It is usually easier to access a large number of unlabelled medical images rather than a large number of annotated images.
“This means it is usually easier to access a large number of unlabelled medical images rather than a large number of annotated images.”
The paper relies on this without support.
In medical imaging, most images are in greyscale so that no colour information is available.
“In medical imaging, most images are in greyscale so that no colour information is available.”
The paper relies on this without support.
The key challenge for self-supervised learning is identifying a suitable self-supervision task, i.e. generating input and output instance pairs from data.
“The key challenge for self-supervised learning is identifying a suitable self supervision task, i.e. generating input and out-put instance pairs from data.”
The paper relies on this without support.
Manual cropping requires expert knowledge and is costly.
“However, manual cropping requires expert knowledge and costly.”
The paper relies on this without support.
Not all patients have two kidneys.
“This is because not all patients have two kidneys.”
The paper relies on this without support.
Gliomas are the major brain tumours occurring in adults.
“Gliomas are the major brain tumours occurring in adults.”
The paper relies on this without support.
Accurate segmentation of gliomas on MR image is a key step for quantification.
“Accurate segmentation of gliomas on MR image is a key step for quantification.”
The paper relies on this without support.
In medical image analysis, the image context is the common feature for classification, localization/detection, and segmentation tasks.
“Particularly, in medical image analysis, the image context is the common feature for classification, localization/detection, and segmentation tasks.”
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Self-supervised learning for medical image analysis using image context restoration > Self-supervision based on context restoration > Context restoration
“a new dataset (1)χ˜=f(χ˜) is generated.”
§ Self-supervised learning for medical image analysis using image context restoration > Self-supervision based on context restoration > Context restoration
“xi=g(x˜i)=f−1(x˜i)”
Citations
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474
Very Deep Convolutional Networks for Large-Scale Image Recognition
Simonyan, K., & Zisserman, A. (2014) · arXiv · doi:10.48550/arxiv.1409.1556 · Cited by 114,148
ImageNet classification with deep convolutional neural networks
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017) · Communications of the ACM · doi:10.1145/3065386 · Cited by 112,155
U-Net: Convolutional Networks for Biomedical Image Segmentation
Ronneberger, O., Fischer, P., & Brox, T. (2015) · Lecture notes in computer science · doi:10.1007/978-3-319-24574-4_28 · Cited by 99,404
Adam: A Method for Stochastic Optimization
Kingma, D. P., & Ba, J. (2014) · arXiv · doi:10.48550/arxiv.1412.6980 · Cited by 98,946
Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
Ren, S., He, K., Girshick, R., & Sun, J. (2017) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2016.2577031 · Cited by 73,472
Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
Ren, S., He, K., Girshick, R., & Sun, J. (2015) · arXiv · doi:10.48550/arxiv.1506.01497 · Cited by 70,812
Densely Connected Convolutional Networks
Huang, G., Liu, Z., van der Maaten, L., & Weinberger, K. Q. (2016) · arXiv · doi:10.48550/arxiv.1608.06993 · Cited by 48,960
Densely Connected Convolutional Networks
Huang, G., Liu, Z., Van Der Maaten, L., & Weinberger, K. Q. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.243 · Cited by 48,566
Going deeper with convolutions
Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., Erhan, D., Vanhoucke, V., & Rabinovich, A. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298594 · Cited by 48,562
Cited by
Sorted by citations
Recent advances and clinical applications of deep learning in medical image analysis
Chen, X., Wang, X., Zhang, K., Fung, K.-M. A., Thai, T., Moore, K. N., Mannel, R. S., Liu, H., Zheng, B., & Qiu, Y. (2022) · Medical Image Analysis · doi:10.1016/j.media.2022.102444 · Cited by 1,180
CutPaste: Self-Supervised Learning for Anomaly Detection and Localization
Li, C.-L., Sohn, K., Yoon, J., & Pfister, T. (2021) · 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr46437.2021.00954 · Cited by 1,107
COVIDX-Net: A Framework of Deep Learning Classifiers to Diagnose COVID-19 in X-Ray ImagesPreprint
Hemdan, E. E., Shouman, M. A., & Karar, M. E. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2003.11055 · Cited by 1,218
Using machine learning approaches for multi-omics data analysis: A review
Reel, P. S., Reel, S., Pearson, E., Trucco, E., & Jefferson, E. (2021) · Biotechnology Advances · doi:10.1016/j.biotechadv.2021.107739 · Cited by 903
Self-Supervised Pre-Training of Swin Transformers for 3D Medical Image Analysis
Tang, Y., Yang, D., Li, W., Roth, H. R., Landman, B., Xu, D., Nath, V., & Hatamizadeh, A. (2022) · 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr52688.2022.02007 · Cited by 849
Big Self-Supervised Models Advance Medical Image Classification
Azizi, S., Mustafa, B., Ryan, F., Beaver, Z., Freyberg, J., Deaton, J., Loh, A., Karthikesalingam, A., Kornblith, S., Chen, T., et al. (2021) · 2021 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.00346 · Cited by 685
Pérez‐García, F., Sparks, R. E., & Sebastien Ourselin, A. (2021) · Computer Methods and Programs in Biomedicine · doi:10.1016/j.cmpb.2021.106236 · Cited by 672
Artificial intelligence and machine learning for medical imaging: A technology review
Montero, A. M. B., Javaid, U., Valdés, G., Nguyen, D. T. D., Desbordes, P., Macq, B. M. M., Willems, S., Vandewinckele, L., Holmström, M., Löfman, F., et al. (2021) · Physica Medica · doi:10.1016/j.ejmp.2021.04.016 · Cited by 510
Self-Supervised Representation Learning: Introduction, advances, and challenges
Ericsson, L., Gouk, H., Loy, C. C., & Hospedales, T. M. (2022) · IEEE Signal Processing Magazine · doi:10.1109/msp.2021.3134634 · Cited by 433
Chaitanya, K., Erdil, E., Karani, N., & Konukoglu, E. (2020) · Neural Information Processing Systems · doi:10.3929/ethz-b-000443425 · Cited by 720
Deep Learning for Medical Image-Based Cancer Diagnosis
Jiang, X., Hu, Z., Wang, S., & Zhang, Y. (2023) · Cancers · doi:10.3390/cancers15143608 · Cited by 409
Zhou, Z., Sodha, V., Pang, J., Gotway, M. B., & Liang, J. (2020) · Medical Image Analysis · doi:10.1016/j.media.2020.101840 · Cited by 375
The Role of Artificial Intelligence in Early Cancer Diagnosis
Hunter, B., Hindocha, S., & Lee, R. W. (2022) · Cancers · doi:10.3390/cancers14061524 · Cited by 384
Self-supervised learning: A succinct review
Rani, V., Nabi, S. T., Kumar, M., Mittal, A., & Kumar, K. (2023) · Archives of Computational Methods in Engineering · doi:10.1007/s11831-023-09884-2 · Cited by 295
Self-supervision with Superpixels: Training Few-Shot Medical Image Segmentation Without Annotation
Ouyang, C., Biffi, C., Chen, C., Kart, T., Qiu, H., & Rueckert, D. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-58526-6_45 · Cited by 293
Self-citationSelf-supervised learning methods and applications in medical imaging analysis: a survey
Shurrab, S., & Duwairi, R. (2022) · PeerJ Computer Science · doi:10.7717/peerj-cs.1045 · Cited by 299
Robust and data-efficient generalization of self-supervised machine learning for diagnostic imaging
Azizi, S., Culp, L., Freyberg, J., Mustafa, B., Baur, S., Kornblith, S., Chen, T., Tomasev, N., Mitrović, J., Strachan, P., et al. (2023) · Nature Biomedical Engineering · doi:10.1038/s41551-023-01049-7 · Cited by 250
Deep Learning for PET Image Reconstruction
Reader, A. J., Corda-D’Incan, G., Mehranian, A., da Costa-Luis, C. O., Ellis, S., & Schnabel, J. A. (2020) · IEEE Transactions on Radiation and Plasma Medical Sciences · doi:10.1109/trpms.2020.3014786 · Cited by 255
Self-Supervised Learning for Few-Shot Medical Image Segmentation
Ouyang, C., Biffi, C., Chen, C., Kart, T., Qiu, H., & Rueckert, D. (2022) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2022.3150682 · Cited by 232
Self-citationChaitanya, K., Erdil, E., Karani, N., & Konukoğlu, E. (2023) · Medical Image Analysis · doi:10.1016/j.media.2023.102792 · Cited by 234
A review on deep-learning algorithms for fetal ultrasound-image analysis
Fiorentino, M. C., Villani, F. P., Cosmo, M. D., Frontoni, E., & Moccia, S. (2022) · Medical Image Analysis · doi:10.1016/j.media.2022.102629 · Cited by 224
Ten deep learning techniques to address small data problems with remote sensing
Safonova, A. N., Ghazaryan, G., Stiller, S., Main-Knorn, M., Nendel, C., & Ryo, M. (2023) · International Journal of Applied Earth Observation and Geoinformation · doi:10.1016/j.jag.2023.103569 · Cited by 242
Applications of artificial intelligence and deep learning in molecular imaging and radiotherapy
Arabi, H., & Zaidi, H. (2020) · European Journal of Hybrid Imaging · doi:10.1186/s41824-020-00086-8 · Cited by 201
Applications of artificial intelligence in cardiovascular imaging
Maxime Sermesant, B., Delingette, H., Cochet, H., Jaı̈s, P., & Ayache, N. (2021) · Nature Reviews Cardiology · doi:10.1038/s41569-021-00527-2 · Cited by 181
Hansen, S., Gautam, S., Jenssen, R., & Kampffmeyer, M. (2022) · Medical Image Analysis · doi:10.1016/j.media.2022.102385 · Cited by 176
Rubik’s Cube+: A self-supervised feature learning framework for 3D medical image analysis
Zhu, J., Li, Y., Hu, Y., Ma, K., Zhou, S. K., & Zheng, Y. (2020) · Medical Image Analysis · doi:10.1016/j.media.2020.101746 · Cited by 162
Domain Adaptive Relational Reasoning for 3D Multi-organ Segmentation
Fu, S., Lu, Y., Wang, Y., Zhou, Y., Shen, W., Fishman, E. K., & Yuille, A. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-59710-8_64 · Cited by 156
A medical multimodal large language model for future pandemics
Liu, F., Zhu, T., Wu, X., Yang, B., You, C., Wang, C., Lu, L., Liu, Z., Zheng, Y., Sun, X., et al. (2023) · npj Digital Medicine · doi:10.1038/s41746-023-00952-2 · Cited by 155
Albelwi, S. (2022) · Entropy · doi:10.3390/e24040551 · Cited by 150
Self-Supervised Feature Learning via Exploiting Multi-Modal Data for Retinal Disease Diagnosis
Li, X., Jia, M. J., Islam, M. T., Yu, L., & Xing, L. (2020) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2020.3008871 · Cited by 150
Haghighi, F., Taher, M. R. H., Zhou, Z., Gotway, M. B., & Liang, J. (2021) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2021.3060634 · Cited by 151
Zhou, H.-Y., Chen, X., Zhang, Y., Luo, R., Wang, L., & Yu, Y. (2022) · Nature Machine Intelligence · doi:10.1038/s42256-021-00425-9 · Cited by 126
How Do Machines Learn? Artificial Intelligence as a New Era in Medicine
Koteluk, O., Wartecki, A., Mazurek, S., Kołodziejczak, I., & Mackiewicz, A. A. (2021) · Journal of Personalized Medicine · doi:10.3390/jpm11010032 · Cited by 129
3D Self-Supervised Methods for Medical ImagingPreprint
Taleb, A., Loetzsch, W., Danz, N., Severin, J., Gaertner, T., Bergner, B., & Lippert, C. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2006.03829 · Cited by 145
Self-supervised driven consistency training for annotation efficient histopathology image analysis
Srinidhi, C. L., Kim, S. W., Chen, F.-D., & Martel, A. L. (2022) · Medical Image Analysis · doi:10.1016/j.media.2021.102256 · Cited by 116
Machine Learning for Auto-Segmentation in Radiotherapy Planning
Harrison, K., Pullen, H., Welsh, C., Oktay, O., Alvarez-Valle, J., & Jena, R. (2022) · Clinical Oncology · doi:10.1016/j.clon.2021.12.003 · Cited by 123
An Overview of Organs-on-Chips Based on Deep Learning
Li, J., Chen, J., Bai, H., Wang, H., Hao, S., Ding, Y., Peng, B., Zhang, J., Li, L., & Huang, W. (2022) · Research · doi:10.34133/2022/9869518 · Cited by 115
Xu, Z., Wang, Y., Lu, D., Yu, L., Yan, J., Luo, J., Ma, K., Zheng, Y., & Tong, R. K. (2022) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2022.3162043 · Cited by 110
Self-Supervised Multi-Modal Hybrid Fusion Network for Brain Tumor Segmentation
Fang, F., Yao, Y., Zhou, T., Xie, G.-S., & Lu, J. (2021) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2021.3109301 · Cited by 107
Rotation-Oriented Collaborative Self-Supervised Learning for Retinal Disease Diagnosis
Li, X., Hu, X., Qi, X., Yu, L., Zhao, W., Heng, P.-A., & Xing, L. (2021) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2021.3075244 · Cited by 105
Zhou, H.-Y., Lu, C., Yang, S., Han, X., & Yu, Y. (2021) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.00348 · Cited by 105
Haghighi, F., Taher, M., Gotway, M., & Liang, J.-M. (2022) · Computer Vision and Pattern Recognition · doi:10.1109/cvpr52688.2022.02016 · Cited by 97
Automated detection of COVID-19 cases from chest X-ray images using deep neural network and XGBoost
Nasiri, H., & Hasani, S. (2022) · Radiography · doi:10.1016/j.radi.2022.03.011 · Cited by 102
Delving into Masked Autoencoders for Multi-Label Thorax Disease Classification
Xiao, J., Bai, Y., Yuille, A., & Zhou, Z. (2023) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv56688.2023.00358 · Cited by 110
Multi-task contrastive learning for automatic CT and X-ray diagnosis of COVID-19
Li, J., Zhao, G., Tao, Y., Zhai, P., Chen, H., He, H., & Cai, T. (2021) · Pattern Recognition · doi:10.1016/j.patcog.2021.107848 · Cited by 104
SAM: Self-Supervised Learning of Pixel-Wise Anatomical Embeddings in Radiological Images
Yan, K., Cai, J., Jin, D., Miao, S., Guo, D., Harrison, A. P., Tang, Y., Xiao, J., Lu, J., & Lu, L. (2022) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2022.3169003 · Cited by 95
Rethinking pre-training on medical imaging
Wen, Y., Chen, L., Deng, Y., & Zhou, C. (2021) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2021.103145 · Cited by 92
Self-citationDive into the details of self-supervised learning for medical image analysis
Zhang, C., Zheng, H., & Gu, Y. (2023) · Medical Image Anal. · doi:10.1016/j.media.2023.102879 · Cited by 101
Transformers in medical image segmentation: a narrative review
Khan, R. F., Lee, B.-D., & Lee, M. S. (2023) · Quantitative Imaging in Medicine and Surgery · doi:10.21037/qims-23-542 · Cited by 104
The importance of resource awareness in artificial intelligence for healthcare
Jia, Z., Chen, J., Xu, X., Kheir, J. N., Hu, J., Xiao, H., Peng, S., Hu, X. S., Chen, D., & Shi, Y. (2023) · Nature Machine Intelligence · doi:10.1038/s42256-023-00670-0 · Cited by 93
Montero, A. M. B., Bibal, A., Huet‐Dastarac, M., Draguet, C., Valdés, G., Nguyen, D. T. D., Willems, S., Vandewinckele, L., Holmström, M., Löfman, F., et al. (2022) · Physics in Medicine and Biology · doi:10.1088/1361-6560/ac678a · Cited by 91
Huang, H. B., Wu, R., Li, Y., & Peng, C. (2022) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2022.3171851 · Cited by 84
Abbas, A., Abdelsamea, M. M., & Gaber, M. M. (2021) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2021.3082015 · Cited by 84
Qu, L., Liu, S., Liu, X., Wang, M., & Song, Z. (2022) · Physics in Medicine and Biology · doi:10.1088/1361-6560/ac910a · Cited by 89
Contrastive Learning with Continuous Proxy Meta-data for 3D MRI Classification
Dufumier, B., Gori, P., Victor, J., Grigis, A., Wessa, M., Brambilla, P., Favre, P., Polosan, M., McDonald, C., Piguet, C. M., et al. (2021) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-87196-3_6 · Cited by 82
Self-evolving vision transformer for chest X-ray diagnosis through knowledge distillation
Park, S. J., Kim, G., Oh, Y., Seo, J. B., Lee, S. M., Kim, J. H., Moon, S., Lim, J., Park, C. M., & Ye, J. C. (2022) · Nature Communications · doi:10.1038/s41467-022-31514-x · Cited by 82
A review of artificial intelligence in prostate cancer detection on imaging
Bhattacharya, I., Khandwala, Y. S., Vesal, S., Shao, W., Yang, Q., Soerensen, S. J. C., Fan, R. E., Ghanouni, P., Kunder, C. A., Brooks, J. D., et al. (2022) · Therapeutic Advances in Urology · doi:10.1177/17562872221128791 · Cited by 97
Zhou, H.-Y., Lu, C., Chen, C., Yang, S., & Yu, Y. (2023) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2023.3234002 · Cited by 87
PRIOR: Prototype Representation Joint Learning from Medical Images and Reports
Cheng, P., Lin, L., Lyu, J., Huang, Y., Luo, W., & Tang, X. (2023) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51070.2023.01953 · Cited by 93
Best of Both Worlds: Multimodal Contrastive Learning with Tabular and Imaging Data
Hager, P., Menten, M. J., & Rueckert, D. (2023) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52729.2023.02291 · Cited by 96
Self-citationZhang, J., Zhang, S., Shen, X., Lukasiewicz, T., & Xu, Z.-H. (2023) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2023.3290356 · Cited by 72
ContIG: Self-supervised Multimodal Contrastive Learning for Medical Imaging with Genetics
Taleb, A., Kirchler, M., Monti, R., & Lippert, C. (2022) · 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr52688.2022.02024 · Cited by 83
Haghighi, F., Taher, M. R. H., Zhou, Z., Gotway, M. B., & Liang, J. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-59710-8_14 · Cited by 78
Hybrid Masked Image Modeling for 3D Medical Image Segmentation
Xing, Z., Zhu, L., Yu, L., Xing, Z., & Wan, L. (2023) · IEEE journal of biomedical and health informatics · doi:10.1109/jbhi.2024.3360239 · Cited by 49
UniMiSS: Universal Medical Self-supervised Learning via Breaking Dimensionality Barrier
Xie, Y., Zhang, J., Xia, Y., & Wu, Q. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-19803-8_33 · Cited by 73
Machine Learning in Oncology: What Should Clinicians Know?
Nagy, M. R., Radakovich, N., & Nazha, A. (2020) · JCO Clinical Cancer Informatics · doi:10.1200/cci.20.00049 · Cited by 75
Artificial intelligence in obstetric ultrasound: A scoping review.
Horgan, R., Nehme, L., & Abuhamad, A. (2023) · Prenatal Diagnosis · doi:10.1002/pd.6411 · Cited by 72
“…assessed the ability of deep learning models to undergo self ‐ supervised learning by correcting the order of a reshuffled fetal video clip, 115 by context restoration of unlabeled 2D fetal images 116 and the use of random forests classifiers for classification of unlabeled fetal ultrasound images.”
BackgroundSelf-supervised learning for medical image analysis: a comprehensive review
Rani, V., Kumar, M., Gupta, A., Sachdeva, M., Mittal, A., & Kumar, K. (2024) · Evolving Systems · doi:10.1007/s12530-024-09581-w · Cited by 83
Progressive Pretraining Network for 3D System Matrix Calibration in Magnetic Particle Imaging
Shi, G., Yin, L., An, Y., Li, G., Zhang, L., Bian, Z., Chen, Z., Zhang, H., Hui, H., & Tian, J. (2023) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2023.3297173 · Cited by 72
Challenges for machine learning in clinical translation of big data imaging studies
Dinsdale, N. K., Bluemke, E., Sundaresan, V., Jenkinson, M., Smith, S. M., & Namburete, A. I. L. (2022) · Neuron · doi:10.1016/j.neuron.2022.09.012 · Cited by 74
Xu, Z., Lu, D., Luo, J., Wang, Y., Yan, J., Ma, K., Zheng, Y., & Tong, R. K. (2022) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2022.3176915 · Cited by 69
OpenBHB: a Large-Scale Multi-Site Brain MRI Data-set for Age Prediction and Debiasing
Dufumier, B., Grigis, A., Victor, J., Ambroise, C., Frouin, V., & Duchesnay, É. (2022) · NeuroImage · doi:10.1016/j.neuroimage.2022.119637 · Cited by 78
Contrastive self-supervised learning from 100 million medical images with optional supervision
Ghesu, F. C., Georgescu, B., Mansoor, A., Yoo, Y., Neumann, D., Patel, P. K., Vishwanath, R. S., Balter, J. M., Cao, Y., Grbić, S., et al. (2022) · Journal of medical imaging · doi:10.1117/1.jmi.9.6.064503 · Cited by 111
Liver, kidney and spleen segmentation from CT scans and MRI with deep learning: A survey
Altini, N., Prencipe, B., Cascarano, G. D., Brunetti, A., Brunetti, G., Triggiani, V., Carnimeo, L., Marino, F., Guerriero, A., Villani, L., et al. (2022) · Neurocomputing · doi:10.1016/j.neucom.2021.08.157 · Cited by 66
Deep learning-based approach for detecting COVID-19 in chest X-rays
Şahin, M. E. (2022) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2022.103977 · Cited by 66
MISSU: 3D Medical Image Segmentation via Self-Distilling TransUNet
Wang, N., Lin, S., Li, X., Li, K., Shen, Y., Gao, Y., & Ma, L. (2023) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2023.3264433 · Cited by 64
Image data augmentation techniques based on deep learning: A survey
Zeng, W. (2024) · Mathematical Biosciences & Engineering · doi:10.3934/mbe.2024272 · Cited by 93
Intra- and Inter-Slice Contrastive Learning for Point Supervised OCT Fluid Segmentation
He, X., Fang, L., Tan, M., & Chen, X. (2022) · IEEE Transactions on Image Processing · doi:10.1109/tip.2022.3148814 · Cited by 65
PyMIC: A deep learning toolkit for annotation-efficient medical image segmentation
Wang, G., Luo, X., Gu, R., Yang, S., Qu, Y., Zhai, S., Zhao, Q., Li, K., & Zhang, S. (2023) · Computer Methods and Programs in Biomedicine · doi:10.1016/j.cmpb.2023.107398 · Cited by 62
Lei, B., yu, L., Xie, J., Wu, Y., Liang, E., Liu, Y., Yang, P., Wang, T., Liu, C., Du, J., et al. (2024) · Pattern Recognition · doi:10.1016/j.patcog.2024.110423 · Cited by 65
MuRCL: Multi-Instance Reinforcement Contrastive Learning for Whole Slide Image Classification
Zhu, Z., Yu, L., Wu, W., Yu, R., Zhang, D., & Wang, L. (2022) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2022.3227066 · Cited by 61
Wang, W., Shen, X., Yi, B., Zhang, H., Liu, J., & Dai, C. (2024) · Expert Systems with Applications · doi:10.1016/j.eswa.2024.123710 · Cited by 63
Huang, W., Li, C., Zhou, H.-Y., Yang, H., Liu, J., Liang, Y., Zheng, H., Zhang, S., & Wang, S. (2024) · Nature Communications · doi:10.1038/s41467-024-51749-0 · Cited by 62
Li, Z., Zhao, W., Shi, F., Qi, L., Xie, X., Wei, Y., Ding, Z., Gao, Y., Wu, S., Liu, J., et al. (2021) · Medical Image Analysis · doi:10.1016/j.media.2021.101978 · Cited by 64
Multi-Scale Transformer Network With Edge-Aware Pre-Training for Cross-Modality MR Image Synthesis
Li, Y., Zhou, T., He, K., Zhou, Y., & Shen, D. G. (2023) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2023.3288001 · Cited by 60
Stacked-autoencoder-based model for COVID-19 diagnosis on CT images
Li, D., Fu, Z., & Xu, J. (2020) · Applied Intelligence · doi:10.1007/s10489-020-02002-w · Cited by 59
Zhou, H.-Y., Wang, C., Li, H., Wang, G., Zhang, S., Li, W., & Yu, Y. (2021) · Medical Image Analysis · doi:10.1016/j.media.2021.102117 · Cited by 58
Yang, X., Chen, D., Sun, Q., Wang, Y., Xia, Y., Yang, J., Lin, C., Dang, X., Cen, Z., Liang, D., et al. (2023) · Cell Discovery · doi:10.1038/s41421-023-00543-1 · Cited by 59
Self-supervised Visual Representation Learning for Histopathological Images
Yang, P., Hong, Z., Yin, X., Zhu, C., & Jiang, R. (2021) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-87196-3_5 · Cited by 56
A Review of Predictive and Contrastive Self-supervised Learning for Medical Images
Wang, W.-C., Ahn, E., Feng, D. D., & Kim, J. (2023) · Machine Intelligence Research · doi:10.1007/s11633-022-1406-4 · Cited by 64
Consequential Advancements of Self-Supervised Learning (SSL) in Deep Learning Contexts
Abdulrazzaq, M. M., Ramaha, N. T. A., Hameed, A. A., Salman, M., Yon, D. K., Fitriyani, N. L., Syafrudin, M., & Lee, S. W. (2024) · Mathematics · doi:10.3390/math12050758 · Cited by 67
Automatic detection of stroke lesion from diffusion-weighted imaging via the improved YOLOv5
Chen, S., Duan, J., Wang, H., Wang, R., Li, J., Qi, M., Duan, Y., & Qi, S. (2022) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2022.106120 · Cited by 56
Deep multiple-instance learning for abnormal cell detection in cervical histopathology images
Pal, A., Xue, Z., Desai, K. T., Banjo, A. A. F., Adepiti, C. A., Long, L. R., SCHIFFMAN, M. H., & Antani, S. (2021) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2021.104890 · Cited by 55
Wu, R., Liang, C., Li, Y., Shi, X., Zhang, J., & Huang, H. B. (2022) · Expert Systems with Applications · doi:10.1016/j.eswa.2022.119339 · Cited by 57
Self-Supervised Graph Learning With Hyperbolic Embedding for Temporal Health Event Prediction
Lu, C. H., Reddy, C. K. A., & Ning, Y. (2021) · IEEE Transactions on Cybernetics · doi:10.1109/tcyb.2021.3109881 · Cited by 59
VanBerlo, B., Hoey, J., & Wong, A. M. (2024) · BMC Medical Imaging · doi:10.1186/s12880-024-01253-0 · Cited by 61
Kidney Tumor Classification on CT images using Self-supervised Learning
Özbay, E., Ozbay, F. A., & Gharehchopogh, F. S. (2024) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2024.108554 · Cited by 52
Beauferris, Y., Teuwen, J., Karkalousos, D., Moriakov, N., Caan, M. W. A., Yiasemis, G., Rodrigues, L., Lopes, A., Pedrini, H., Rittner, L., et al. (2022) · Frontiers in Neuroscience · doi:10.3389/fnins.2022.919186 · Cited by 51
Mahapatra, D., Ge, Z., & Reyes, M. (2022) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2022.3163232 · Cited by 49
Artificial intelligence in multiparametric magnetic resonance imaging: A review
Li, C., Li, W., Liu, C., Zheng, H., Cai, J., & Wang, S. (2022) · Medical Physics · doi:10.1002/mp.15936 · Cited by 52
Metadata
- Authors
- Chen, L., Bentley, P., Mori, K., Misawa, K., Fujiwara, M., & Rueckert, D.
- Affiliations
- Imperial College London (GB)
- Nagoya University (JP)
- Aichi Cancer Center (JP)
- Nagoya University Hospital (JP)
- Corresponding author
- Chen, Liang
- Venue
- Medical Image Analysis
- Publisher
- Elsevier BV
- Published
- 2019-12
- Volume
- 58
- Pages
- 101539
- ISSN
- 1361-8415 (ISSN-L)
- PMC
- PMC7613987
- OpenAlex
- W2964744899
- Semantic Scholar
- d6e78069dbbb35592f013187ef3432286def8f18
- Cited by
- 575
- References
- 73
- Citations by year
- 2026: 44 · 2025: 85 · 2024: 99 · 2023: 133 · 2022: 105 · 2021: 74 · 2020: 34 · 2019: 1
- FWCI
- 24.53
- Type
- journal article
- Language
- en
- Open access
- green
- www.ncbi.nlm.nih.gov
- Locations
- Medical Image Analysis (publishedVersion)
- PubMed (publishedVersion)
- PubMed Central (acceptedVersion)
- Spiral (Imperial College London) (submittedVersion, cc-by-nc-nd)
- Funders
- Wellcome Trust (102431)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Domain Adaptation and Few-Shot Learning
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Medical Image Segmentation Techniques
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Keywords
- self-supervised learning, unlabeled data, CT images, brain tumor segmentation, multimodal MRI, image classification, image localization, image segmentation
- MeSH terms
- Abdomen, Supervised Machine Learning, Brain Neoplasms, Humans, Magnetic Resonance Imaging, Tomography, X-Ray Computed, Ultrasonography, Prenatal
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
