Improved performance and robustness of multi-task representation learning with consistency loss between pretexts for intracranial hemorrhage identification in head CT

Kyung, Sunggu; Shin, Keewon; Jeong, Heejin; Kim, Kiduk; Park, Jooyoung; Cho, Kyungjin; Lee, Jeong Hyun; Hong, Gil-Sun; Kim, Namkug

doi:10.1016/j.media.2022.102489

Cited by 12 publications

(7 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the upstream, effective representation learning was performed by multi-task learning (classification, segmentation, reconstruction) and differences in the specific head of the consistency loss mitigation target are added. For downstream, feature extractor trained upstream is combined with 3D operator (classifier or divider) to implement specific tasks [16]. Wu et al proposed a combination of an attention-based convolutional neural network and a variational Gaussian process for multiple instance learning method for predicting intracranial hemorrhage slices [24].…”

Section: A Related Intracranial Hemorrhage Segmentation Methodsmentioning

confidence: 99%

“…Because of the clinical significance and the intrinsic challenges, the task of automatic intracranial hemorrhage segmentation has attracted extensive attention in the past few years. Recently, deep learning-based ICH segmentation models that segment ICH regions and quantify hematoma volume have been performed to effectively diagnose ICH and have achieved competitive results [6], [16]- [20]. However, all those abovementioned ICH segmentation methods ignore the anisotropic nature of the NCCT volume by simply performing 2D or 3D convolutional networks, and they were evaluated on different in-house hemorrhage segmentation datasets with distinct metrics, making it highly challenging to improve segmentation performance and perform objective comparisons among these methods.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

The state-of-the-art 3D anisotropic intracranial hemorrhage segmentation on non-contrast head CT: The INSTANCE challenge

Li¹,

Luo²,

Wang³

et al. 2023

Preprint

View full text Add to dashboard Cite

Automatic intracranial hemorrhage segmentation in 3D non-contrast head CT (NCCT) scans is significant in clinical practice. Existing hemorrhage segmentation methods usually ignores the anisotropic nature of the NCCT, and are evaluated on different in-house datasets with distinct metrics, making it highly challenging to improve segmentation performance and perform objective comparisons among different methods. The 2022 intracranial hemorrhage segmentation on non-contrast head CT (INSTANCE 2022) was a grand challenge held in conjunction with the 2022 International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI). It is intended to resolve the above-mentioned problems and promote the development of both intracranial hemorrhage segmentation and anisotropic data processing. The INSTANCE released a training set of 100 cases with ground-truth and a validation set with 30 cases without ground-truth labels that were available to the participants. A held-out testing set with 70 cases is utilized for the final evaluation and ranking. The methods from different participants are ranked based on four metrics, including Dice Similarity Coefficient (DSC), Hausdorff Distance (HD), Relative Volume Difference (RVD) and Normalized Surface Dice (NSD). A total of 13 teams submitted distinct solutions to resolve the challenges, making several baseline models, pre-processing strategies and anisotropic data processing techniques available to future researchers. The winner method achieved an average DSC of 0.6925, demonstrating a significant growth over our proposed baseline method. To the best of our knowledge, the proposed INSTANCE challenge releases the first intracranial hemorrhage segmentation benchmark, and is also the first challenge that intended to resolve the anisotropic problem in 3D medical image segmentation, which provides new alternatives in these research fields.

show abstract

Section: A Related Intracranial Hemorrhage Segmentation Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

The state-of-the-art 3D anisotropic intracranial hemorrhage segmentation on non-contrast head CT: The INSTANCE challenge

Li¹,

Luo²,

Wang³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Many studies have focused on AI in radiology because of the complexities involved in training owing to high-dimensional data and the limitations of data acquisition. These radiology studies demonstrated the superiority of the MTL over other models; for instance, MTL has shown better performance than alternative approaches in various tasks, such as segmenting thoracic organs from computed tomography (CT) slices [ 85 ]; object detection, segmentation, and classification in breast cancer diagnosis using full-field digital mammogram datasets [ 86 ]; identification and segmentation of coronavirus disease 2019 (COVID-19) lesions from chest CT images [ 87 ]; and identification of hemorrhage and segmentation in head and neck CT images [ 88 ] ( Fig. 3 ).…”

Section: Overcoming the Challengesmentioning

confidence: 99%

Overcoming the Challenges in the Development and Implementation of Artificial Intelligence in Radiology: A Comprehensive Review of Solutions Beyond Supervised Learning

Hong,

Jang,

Kyung

et al. 2023

Korean J Radiol

Self Cite

View full text Add to dashboard Cite

Artificial intelligence (AI) in radiology is a rapidly developing field with several prospective clinical studies demonstrating its benefits in clinical practice. In 2022, the Korean Society of Radiology held a forum to discuss the challenges and drawbacks in AI development and implementation. Various barriers hinder the successful application and widespread adoption of AI in radiology, such as limited annotated data, data privacy and security, data heterogeneity, imbalanced data, model interpretability, overfitting, and integration with clinical workflows. In this review, some of the various possible solutions to these challenges are presented and discussed; these include training with longitudinal and multimodal datasets, dense training with multitask learning and multimodal learning, self-supervised contrastive learning, various image modifications and syntheses using generative models, explainable AI, causal learning, federated learning with large data models, and digital twins.

show abstract

“…The simplest form of semi-supervised learning is transfer learning (Raina et al, 2007 ; Torrey and Shavlik, 2010 ) where the model's weights are initialized with those learned using the unlabeled/weakly-labeled data on a pretext task. A few studies have explored semi-supervised methods for ICH segmentation (Wang et al, 2020 ; Kyung et al, 2022 ). To the best of our knowledge, only Wang et al ( 2020 ) examined the use of external unlabeled public data, from RSNA (Flanders et al, 2020 ).…”

Section: Introductionmentioning

confidence: 99%

Label-efficient deep semantic segmentation of intracranial hemorrhages in CT-scans

Spahr,

Ståhle,

Wang

et al. 2023

Front. Neuroimaging

View full text Add to dashboard Cite

Intracranial hemorrhage (ICH) is a common finding in traumatic brain injury (TBI) and computed tomography (CT) is considered the gold standard for diagnosis. Automated detection of ICH provides clinical value in diagnostics and in the ability to feed robust quantification measures into future prediction models. Several studies have explored ICH detection and segmentation but the research process is somewhat hindered due to a lack of open large and labeled datasets, making validation and comparison almost impossible. The complexity of the task is further challenged by the heterogeneity of ICH patterns, requiring a large number of labeled data to train robust and reliable models. Consequently, due to the labeling cost, there is a need for label-efficient algorithms that can exploit easily available unlabeled or weakly-labeled data. Our aims for this study were to evaluate whether transfer learning can improve ICH segmentation performance and to compare a variety of transfer learning approaches that harness unlabeled and weakly-labeled data. Three self-supervised and three weakly-supervised transfer learning approaches were explored. To be used in our comparisons, we also manually labeled a dataset of 51 CT scans. We demonstrate that transfer learning improves ICH segmentation performance on both datasets. Unlike most studies on ICH segmentation our work relies exclusively on publicly available datasets, allowing for easy comparison of performances in future studies. To further promote comparison between studies, we also present a new public dataset of ICH-labeled CT scans, Seq-CQ500.

show abstract

Improved performance and robustness of multi-task representation learning with consistency loss between pretexts for intracranial hemorrhage identification in head CT

Cited by 12 publications

References 33 publications

The state-of-the-art 3D anisotropic intracranial hemorrhage segmentation on non-contrast head CT: The INSTANCE challenge

The state-of-the-art 3D anisotropic intracranial hemorrhage segmentation on non-contrast head CT: The INSTANCE challenge

Overcoming the Challenges in the Development and Implementation of Artificial Intelligence in Radiology: A Comprehensive Review of Solutions Beyond Supervised Learning

Label-efficient deep semantic segmentation of intracranial hemorrhages in CT-scans

Contact Info

Product

Resources

About