Surrogate Supervision for Medical Image Analysis: Effective Deep Learning From Limited Quantities of Labeled Data

Tajbakhsh, Nima; Hu, Yufei; Cao, Junli; Yan, Xingjian; Xiao, Yi; Liang, Jianming; Terzopoulos, Demetri; Ding, Xiaowei

doi:10.48550/arxiv.1901.08707

Cited by 4 publications

(6 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Specifically, self-supervised pre-training consists of assigning surrogate or proxy labels to the unlabeled data and then training a randomly initialized network using the resulting surrogate supervision signal. The advantage of model pre-training using unlabeled medical data is that the learned knowledge is related to the target medical task; and thus, can be more effective than transfer learning from a foregin domain (e.g., Tajbakhsh et al (2019) and Ross et al (2018)).…”

Section: Self-supervised Pre-trainingmentioning

confidence: 99%

“…The pre-trained model is then fine-tuned for the task of body part recognition in CT and MR images. Tajbakhsh et al (2019) use prediction of image orientation as the surrogate task where the input image is rotated or flipped and the network is trained to predict such a transformation. The authors show that this surrogate task is highly effective for diabetic retinopathy classification in fundus images and lung lobe segmentation in chest CT scans.…”

Section: Self-supervised Pre-trainingmentioning

confidence: 99%

See 1 more Smart Citation

Embracing imperfect datasets: A review of deep learning solutions for medical image segmentation

Tajbakhsh¹,

Jeyaseelan²,

Li³

et al. 2020

Medical Image Analysis

Self Cite

673

358

View full text Add to dashboard Cite

The medical imaging literature has witnessed remarkable progress in high-performing segmentation models based on convolutional neural networks. Despite the new performance highs, the recent advanced segmentation models still require large, representative, and high quality annotated datasets. However, rarely do we have a perfect training dataset, particularly in the field of medical imaging, where data and annotations are both expensive to acquire. Recently, a large body of research has studied the problem of medical image segmentation with imperfect datasets, tackling two major dataset limitations: scarce annotations where only limited annotated data is available for training, and weak annotations where the training data has only sparse annotations, noisy annotations, or image-level annotations. In this article, we provide a detailed review of the solutions above, summarizing both the technical novelties and empirical results. We further compare the benefits and requirements of the surveyed methodologies and provide our recommended solutions to the problems of scarce and weak annotations. We hope this review increases the community awareness of the techniques to handle imperfect datasets.

show abstract

Section: Self-supervised Pre-trainingmentioning

confidence: 99%

Section: Self-supervised Pre-trainingmentioning

confidence: 99%

Embracing imperfect datasets: A review of deep learning solutions for medical image segmentation

Tajbakhsh¹,

Jeyaseelan²,

Li³

et al. 2020

Medical Image Analysis

Self Cite

673

358

View full text Add to dashboard Cite

show abstract

“…As in this study, [36] looked at self-supervised learning for 3D medical image segmentation. They used rotation prediction to pretrain a dense V-Net for lung lobe segmentation, a task which requires strong supervision and large amounts of labeled data.…”

Section: Discussionmentioning

confidence: 99%

“…Another possibility is to modify the data appearance and try to predict the transformation, e.g. image rotation [36]. In multi-task learning (MTL), additional unsupervised task(s) are performed, either sequentially [8,1] or simultaneously, in order to improve the supervised target task of segmentation.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

3D medical image segmentation with labeled and unlabeled data using autoencoders at the example of liver segmentation in CT images

Sital¹,

Brosch²,

Tio³

et al. 2020

Preprint

View full text Add to dashboard Cite

Automatic segmentation of anatomical structures with convolutional neural networks (CNNs) constitutes a large portion of research in medical image analysis. The majority of CNN-based methods rely on an abundance of labeled data for proper training. Labeled medical data is often scarce, but unlabeled data is more widely available. This necessitates approaches that go beyond traditional supervised learning and leverage unlabeled data for segmentation tasks. This work investigates the potential of autoencoder-extracted features to improve segmentation with a CNN. Two strategies were considered. First, transfer learning where pretrained autoencoder features were used as initialization for the convolutional layers in the segmentation network. Second, multi-task learning where the tasks of segmentation and feature extraction, by means of input reconstruction, were learned and optimized simultaneously. A convolutional autoencoder was used to extract features from unlabeled data and a multi-scale, fully convolutional CNN was used to perform the target task of 3D liver segmentation in CT images. For both strategies, experiments were conducted with varying amounts of labeled and unlabeled training data. The proposed learning strategies improved results in 75% of the experiments compared to training from scratch and increased the dice score by up to 0.040 and 0.024 for a ratio of unlabeled to labeled training data of about 32 : 1 and 12.5 : 1, respectively. The results indicate that both training strategies are more effective with a large ratio of unlabeled to labeled training data.

show abstract

Self-Supervised Learning for Cardiac MR Image Segmentation by Anatomical Position Prediction

Bai

Chen

Tarroni

et al. 2019

Preprint

View full text Add to dashboard Cite

Surrogate Supervision for Medical Image Analysis: Effective Deep Learning From Limited Quantities of Labeled Data

Cited by 4 publications

References 12 publications

Embracing imperfect datasets: A review of deep learning solutions for medical image segmentation

Embracing imperfect datasets: A review of deep learning solutions for medical image segmentation

3D medical image segmentation with labeled and unlabeled data using autoencoders at the example of liver segmentation in CT images

Self-Supervised Learning for Cardiac MR Image Segmentation by Anatomical Position Prediction

Contact Info

Product

Resources

About