Visual and Quantitative Evaluation of Amyloid Brain PET Image Synthesis with Generative Adversarial Network

Kang, Hyeon; Park, Jang-Sik; Cho, Kook; Kang, Do‐Young

doi:10.3390/app10072628

Cited by 19 publications

(9 citation statements)

References 50 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We used quantitative and qualitative techniques to analyze the results and demonstrate the quality of the synthetic images generated by the GAN 58 , 59 . In the qualitative evaluation, two methods were applied: (i) t-SNE visualization and (ii) quality assessment by clinical experts.…”

Section: Methodsmentioning

confidence: 99%

Colonoscopic image synthesis with generative adversarial network for enhanced detection of sessile serrated lesions using convolutional neural network

Yoon

Kong

Kim

et al. 2022

Sci Rep

View full text Add to dashboard Cite

Computer-aided detection (CADe) systems have been actively researched for polyp detection in colonoscopy. To be an effective system, it is important to detect additional polyps that may be easily missed by endoscopists. Sessile serrated lesions (SSLs) are a precursor to colorectal cancer with a relatively higher miss rate, owing to their flat and subtle morphology. Colonoscopy CADe systems could help endoscopists; however, the current systems exhibit a very low performance for detecting SSLs. We propose a polyp detection system that reflects the morphological characteristics of SSLs to detect unrecognized or easily missed polyps. To develop a well-trained system with imbalanced polyp data, a generative adversarial network (GAN) was used to synthesize high-resolution whole endoscopic images, including SSL. Quantitative and qualitative evaluations on GAN-synthesized images ensure that synthetic images are realistic and include SSL endoscopic features. Moreover, traditional augmentation methods were used to compare the efficacy of the GAN augmentation method. The CADe system augmented with GAN synthesized images showed a 17.5% improvement in sensitivity on SSLs. Consequently, we verified the potential of the GAN to synthesize high-resolution images with endoscopic features and the proposed system was found to be effective in detecting easily missed polyps during a colonoscopy.

show abstract

Section: Methodsmentioning

confidence: 99%

Colonoscopic image synthesis with generative adversarial network for enhanced detection of sessile serrated lesions using convolutional neural network

Yoon

Kong

Kim

et al. 2022

Sci Rep

View full text Add to dashboard Cite

show abstract

“…The authors of [ 53 ] proposed a GAN for computed tomography (CT) to magnetic resonance image (MRI) data synthesis and translation. A conditional GAN was used for PET image synthesis [ 54 ]. A deep convolutional GAN (DCGAN) was recommended for image synthesis and the detection of liver cancer on X-ray and CT images [ 55 ].…”

Section: Related Workmentioning

confidence: 99%

TilGAN: GAN for Facilitating Tumor-Infiltrating Lymphocyte Pathology Image Synthesis With Improved Image Classification

2021

View full text Add to dashboard Cite

Tumor-infiltrating lymphocytes (TILs) act as immune cells against cancer tissues. The manual assessment of TILs is usually erroneous, tedious, costly and subject to inter- and intraobserver variability. Machine learning approaches can solve these issues, but they require a large amount of labeled data for model training, which is expensive and not readily available. In this study, we present an efficient generative adversarial network, TilGAN , to generate high-quality synthetic pathology images followed by classification of TIL and non-TIL regions. Our proposed architecture is constructed with a generator network and a discriminator network. The novelty exists in the TilGAN architecture, loss functions, and evaluation techniques. Our TilGAN -generated images achieved a higher Inception score than the real images (2.90 vs. 2.32, respectively). They also achieved a lower kernel Inception distance (1.44) and a lower Fréchet Inception distance (0.312). It also passed the Turing test performed by experienced pathologists and clinicians. We further extended our evaluation studies and used almost one million synthetic data, generated by TilGAN , to train a classification model. Our proposed classification model achieved a 97.83% accuracy, a 97.37% F1-score, and a 97% area under the curve. Our extensive experiments and superior outcomes show the efficiency and effectiveness of our proposed TilGAN architecture. This architecture can also be used for other types of images for image synthesis.

show abstract

“…Second, t-SNE defines a similar probability distribution over the features in the low-dimensional map for visualization. The details are as follows [52].…”

Section: ) Visualization Using T-snementioning

confidence: 99%

Synthesizing Foot and Ankle Kinematic Characteristics for Lateral Collateral Ligament Injuries Detection

et al. 2020

View full text Add to dashboard Cite

Deep learning has been applied in healthcare, where features of patients are read by the computer to assist with diagnosis and treatment. In sports medicine, kinematic characteristics of injuries need to be defined. Patients' data are then acquired amplified to training deep learning models. In this study, we tracked motions of lower extremities in patients with lateral collateral ligament injuries of the ankle. Key kinematic characteristics of injuries were identified by comparing patients to normal individuals. The deep convolutional generative adversarial networks (DCGANs) was employed to synthesize a modest-sized labeled dataset to avoid the problems raised from using large-scale manual labeling data. We then fed a combination of real and synthesized data to train long short-term memory (LSTM) networks to detect patients with ligament injuries. The results showed that combined data yielded a better outcome, measured by classification accuracy and f1-score, than solely using the patient data or with a large quantity of synthesized single range of motion feature. INDEX TERMS kinematic characteristics, lateral collateral ligament injuries (LCL injuries) of the ankle, injuries detection, deep convolutional generative adversarial networks (DCGANs), long short-term memory (LSTM) networks.

show abstract

Visual and Quantitative Evaluation of Amyloid Brain PET Image Synthesis with Generative Adversarial Network

Cited by 19 publications

References 50 publications

Colonoscopic image synthesis with generative adversarial network for enhanced detection of sessile serrated lesions using convolutional neural network

Colonoscopic image synthesis with generative adversarial network for enhanced detection of sessile serrated lesions using convolutional neural network

TilGAN: GAN for Facilitating Tumor-Infiltrating Lymphocyte Pathology Image Synthesis With Improved Image Classification

Synthesizing Foot and Ankle Kinematic Characteristics for Lateral Collateral Ligament Injuries Detection

Contact Info

Product

Resources

About