Unsupervised Visual Representation Learning Based on Segmentation of Geometric Pseudo-Shapes for Transformer-Based Medical Tasks

Viriyasaranon, Thanaporn; Woo, Sang Myung; Choi, Jang-Hwan

doi:10.1109/jbhi.2023.3237596

Cited by 6 publications

(3 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Because the atypical shape is composed of a random combination of a plurality of simple shapes, it can be easily generated and a variety of complex and differing types of lesions, such as actual tumors, can be formed. Previously, a pseudo-lesion was created in the form of a simple geometric shape, but this simple shape was very weak in its ability to simulate the complex shape of a real tumor [ 18 ]. However, in model observer studies for image quality evaluation, a more realistic tumor or lesion was synthesized and inserted into the CT image.…”

Section: Related Workmentioning

confidence: 99%

“…However, despite the potential of DL, PC diagnosis using DL systems has not yet been actively investigated. Previous studies have demonstrated that DL could reduce the false diagnosis of PC on CT images as a second reader [ 15 , 16 , 17 , 18 , 19 ]. To achieve high-quality results and accurately generalize across multi-centers, CT equipment, and patient ethnicity, a large number of high-quality annotated training datasets are needed to allow deep networks to learn proper visual information for accurate classification.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Annotation-Efficient Deep Learning Model for Pancreatic Cancer Diagnosis and Classification Using CT Images: A Retrospective Diagnostic Study

Viriyasaranon

Chun

Lee

et al. 2023

Cancers

Self Cite

View full text Add to dashboard Cite

The aim of this study was to develop a novel deep learning (DL) model without requiring large-annotated training datasets for detecting pancreatic cancer (PC) using computed tomography (CT) images. This retrospective diagnostic study was conducted using CT images collected from 2004 and 2019 from 4287 patients diagnosed with PC. We proposed a self-supervised learning algorithm (pseudo-lesion segmentation (PS)) for PC classification, which was trained with and without PS and validated on randomly divided training and validation sets. We further performed cross-racial external validation using open-access CT images from 361 patients. For internal validation, the accuracy and sensitivity for PC classification were 94.3% (92.8–95.4%) and 92.5% (90.0–94.4%), and 95.7% (94.5–96.7%) and 99.3 (98.4–99.7%) for the convolutional neural network (CNN) and transformer-based DL models (both with PS), respectively. Implementing PS on a small-sized training dataset (randomly sampled 10%) increased accuracy by 20.5% and sensitivity by 37.0%. For external validation, the accuracy and sensitivity were 82.5% (78.3–86.1%) and 81.7% (77.3–85.4%) and 87.8% (84.0–90.8%) and 86.5% (82.3–89.8%) for the CNN and transformer-based DL models (both with PS), respectively. PS self-supervised learning can increase DL-based PC classification performance, reliability, and robustness of the model for unseen, and even small, datasets. The proposed DL model is potentially useful for PC diagnosis.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Annotation-Efficient Deep Learning Model for Pancreatic Cancer Diagnosis and Classification Using CT Images: A Retrospective Diagnostic Study

Viriyasaranon

Chun

Lee

et al. 2023

Cancers

Self Cite

View full text Add to dashboard Cite

show abstract

“…The trained frameworks can distinguish certain patterns in the image data, but these patterns may not be ideal for the intended task such as semantic segmentation (Chen et al 2019). As a result, most unsupervised learning paradigms for medical image segmentation either perform inadequately or still rely on a few annotated data to fine-tune the trained frameworks (Viriyasaranon et al 2023, Zheng et al 2023.…”

Section: Introductionmentioning

confidence: 99%

Sparse annotation learning for dense volumetric MR image segmentation with uncertainty estimation

Osman,

Li,

Huang

et al. 2023

Phys. Med. Biol.

View full text Add to dashboard Cite

Objective. Training neural networks for pixel-wise or voxel-wise image segmentation is a challenging task that requires a considerable amount of training samples with highly accurate and densely delineated ground truth maps. This challenge becomes especially prominent in the medical imaging domain, where obtaining reliable annotations for training samples is a difficult, time-consuming, and expert-dependent process. Therefore, developing models that can perform well under the conditions of limited annotated training data is desirable. Approach. In this study, we propose an innovative framework called the extremely sparse annotation neural network (ESA-Net) that learns with only the single central slice label for 3D volumetric segmentation which explores both intra-slice pixel dependencies and inter-slice image correlations with uncertainty estimation. Specifically, ESA-Net consists of four specially designed distinct components: (1) an intra-slice pixel dependency-guided pseudo-label generation module that exploits uncertainty in network predictions while generating pseudo-labels for unlabeled slices with temporal ensembling; (2) an inter-slice image correlation-constrained pseudo-label propagation module which propagates labels from the labeled central slice to unlabeled slices by self-supervised registration with rotation ensembling; (3) a pseudo-label fusion module that fuses the two sets of generated pseudo-labels with voxel-wise uncertainty guidance; and (4) a final segmentation network optimization module to make final predictions with scoring-based label quantification. Main results. Extensive experimental validations have been performed on two popular yet challenging magnetic resonance image segmentation tasks and compared to five state-of-the-art methods. Significance. Results demonstrate that our proposed ESA-Net can consistently achieve better segmentation performances even under the extremely sparse annotation setting, highlighting its effectiveness in exploiting information from unlabeled data.

show abstract

Hybrid Deep Learning Model for Pancreatic Cancer Image Segmentation

Bakasa,

Kwenda,

Viriri

2024

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Unsupervised Visual Representation Learning Based on Segmentation of Geometric Pseudo-Shapes for Transformer-Based Medical Tasks

Cited by 6 publications

References 45 publications

Annotation-Efficient Deep Learning Model for Pancreatic Cancer Diagnosis and Classification Using CT Images: A Retrospective Diagnostic Study

Annotation-Efficient Deep Learning Model for Pancreatic Cancer Diagnosis and Classification Using CT Images: A Retrospective Diagnostic Study

Sparse annotation learning for dense volumetric MR image segmentation with uncertainty estimation

Hybrid Deep Learning Model for Pancreatic Cancer Image Segmentation

Contact Info

Product

Resources

About