Synthetic magnetic resonance imaging for primary prostate cancer evaluation: Diagnostic potential of a non-contrast-enhanced bi-parametric approach enhanced with relaxometry measurements

Arita, Yuki; Akita, Hirotaka; Fujiwara, Hirokazu; Hashimoto, Masahiro; Shigeta, Keisuke; Kwee, Thomas C.; Yoshida, Soichiro; Kosaka, Takeo; Okuda, S.; Oya, Mototsugu; Jinzaki, Masahiro

doi:10.1016/j.ejro.2022.100403

“…There was no significant association between false negatives ( p = 0.2766) or false positives between experience levels ( p = 0.06051). These results mirror a similar study where expert and basic prostate radiologists found no difference in diagnostic performance between conventional MRI and synthetic MRI [ 29 ]. They also show a similar result to another study where synthetic brain MRIs were created using a different GAN called DCGAN [ 20 ].…”

Section: Resultssupporting

confidence: 89%

Generative Adversarial Networks Can Create High Quality Artificial Prostate Cancer Magnetic Resonance Images

Xu

¹

,

Booven

²

,

Goberdhan

³

et al. 2023

JPM

View full text Add to dashboard Cite

The recent integration of open-source data with machine learning models, especially in the medical field, has opened new doors to studying disease progression and/or regression. However, the ability to use medical data for machine learning approaches is limited by the specificity of data for a particular medical condition. In this context, the most recent technologies, like generative adversarial networks (GANs), are being looked upon as a potential way to generate high-quality synthetic data that preserve the clinical variability of a condition. However, despite some success, GAN model usage remains largely minimal when depicting the heterogeneity of a disease such as prostate cancer. Previous studies from our group members have focused on automating the quantitative multi-parametric magnetic resonance imaging (mpMRI) using habitat risk scoring (HRS) maps on the prostate cancer patients in the BLaStM trial. In the current study, we aimed to use the images from the BLaStM trial and other sources to train the GAN models, generate synthetic images, and validate their quality. In this context, we used T2-weighted prostate MRI images as training data for Single Natural Image GANs (SinGANs) to make a generative model. A deep learning semantic segmentation pipeline trained the model to segment the prostate boundary on 2D MRI slices. Synthetic images with a high-level segmentation boundary of the prostate were filtered and used in the quality control assessment by participating scientists with varying degrees of experience (more than ten years, one year, or no experience) to work with MRI images. Results showed that the most experienced participating group correctly identified conventional vs. synthetic images with 67% accuracy, the group with one year of experience correctly identified the images with 58% accuracy, and the group with no prior experience reached 50% accuracy. Nearly half (47%) of the synthetic images were mistakenly evaluated as conventional. Interestingly, in a blinded quality assessment, a board-certified radiologist did not significantly differentiate between conventional and synthetic images in the context of the mean quality of synthetic and conventional images. Furthermore, to validate the usability of the generated synthetic images from prostate cancer MRIs, we subjected these to anomaly detection along with the original images. Importantly, the success rate of anomaly detection for quality control-approved synthetic data in phase one corresponded to that of the conventional images. In sum, this study shows promise that high-quality synthetic images from MRIs can be generated using GANs. Such an AI model may contribute significantly to various clinical applications which involve supervised machine-learning approaches.

show abstract

“…There was no significant association between false negatives (p=0.2766) or false positives between experience levels (p=0.06051). These results mirror a similar study where expert and basic prostate radiologists found no difference in diagnostic performance between conventional MRI and synthetic MRI 24 . They also show a similar result to another study where brain synthetic brain MRI were created using a different GAN called DCGAN [20].…”

Section: Round 2 Quality Control Testsupporting

confidence: 89%

Generative Adversarial Networks Can Create High Quality Artificial Prostate Cancer Magnetic Resonance Images

Xu

¹

,

Booven

²

,

Goberdhan

³

et al. 2022

Preprint

View full text Add to dashboard Cite

Purpose: Recent integration of open-source data to machine learning models, especially in the medical field, has opened new doors to study disease progression and/or regression. However, the limitation of using medical data for machine learning approaches is the specificity of data to a particular medical condition. In this context, most recent technologies like generative adversarial networks (GAN) could be used to generate high quality synthetic data that preserves the clinical variability. Materials and Methods: In this study, we used 139 T2-weighted prostate magnetic resonant images (MRI) from various sources as training data for Single Natural Image GAN (SinGAN), to make a generative model. A deep learning semantic segmentation pipeline trained the model to segment the prostate boundary on 2D MRI slices. Synthetic images with a high-level segmentation boundary of the prostate were filtered and used in the quality control assessment by participating scientists with varying degree of experience (more than 10 years, 1 year, or no experience) to work with MRI images. Results: The most experienced participating group correctly identified conventional vs synthetic images with 67% accuracy, the group with 1 year of experience correctly identified the images with 58% accuracy, and group with no prior experience reached 50% accuracy. Nearly half (47%) of the synthetic images were mistakenly evaluated as conventional images. Interestingly, a blinded quality assessment by a board-certified radiologist to differentiate conventional and synthetic images was not significantly different in context of the mean quality of synthetic and conventional images. Conclusions: This study shows promise that high quality synthetic images from MRI can be generated using GAN. Such an AI model may contribute significantly to various clinical applications which involves supervised machine learning approaches.

show abstract

“…It may be that the outer space between the cells was compressed, which limited the diffusion of water molecules, the MRI signal became higher, and the ADC value decreased. With the increase of PCa cell invasion, the adenoid structure gradually became solid and flaky [ 14 , 15 ]. Ogura et al [ 16 ] pointed out that the higher the degree of PCa malignancy, the lower the ADC value, which was consistent with the results of this study.…”

Section: Discussionmentioning

confidence: 99%

Analysis of Apparent Diffusion Coefficient Value and Dynamic Contrast-Enhanced Magnetic Resonance Imaging Parameters of Prostate Cancer Patients after Diagnosis and Treatment with Magnetic Resonance Imaging

Gu¹

2022

Computational and Mathematical Methods in Medicine

0

View full text Add to dashboard Cite

This research was aimed at exploring the changes in the apparent diffusion coefficient (ADC) and dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) parameters of prostate cancer (PCa) patients. Sixty PCa patients from the hospital were recruited as the research object, and dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) scans were performed to determine the shape, scope, and enhancement characteristics of prostate lesions and their relationship with surrounding tissues. The quantitative parameters of ADC and DCE-MRI were measured. There were 4 patients (6.67%) with a Gleason score of 6 and 15 patients (25%) with a 4 + 3 score. The ADC with Gleason = 6 is 0.81 ± 0.08 × 10 − 3 s / m m 2 , the ADC with Gleason = 3 + 4 is 0.74 ± 0.07 × 10 − 3 s / m m 2 , the ADC with Gleason = 4 + 3 is 0.73 ± 0.05 × 10 − 3 s / m m 2 , the ADC with Gleason = 9 is 0.65 ± 0.06 × 10 − 3 s / m m 2 , and the ADC with Gleason = 10 is 0.59 ± 0.07 × 10 − 3 s / m m 2 . As the Gleason score increased, the ADC decreased and the permeation parameter transfer constant increased. When the ADC was combined with the permeability parameter transfer constant, the AUC of Gleason = 6 points and Gleason = 7 points was greatly different ( P < 0.05 ). The 95% CI of the ADC combined permeability parameter transport constant when Gleason = 6 points and Gleason = 7 points was 0.898-0.934, the sensitivity was 75.4%, and the specificity was 86.2%. The ADC value was negatively correlated with Gleason score. The ADC value combined with VTC value has good diagnostic performance in evaluating the invasion of PCa, which is very important for making treatment plan and evaluating prognosis.

show abstract

Synthetic magnetic resonance imaging for primary prostate cancer evaluation: Diagnostic potential of a non-contrast-enhanced bi-parametric approach enhanced with relaxometry measurements

Cited by 5 publications

References 35 publications

Generative Adversarial Networks Can Create High Quality Artificial Prostate Cancer Magnetic Resonance Images

Generative Adversarial Networks Can Create High Quality Artificial Prostate Cancer Magnetic Resonance Images

Generative Adversarial Networks Can Create High Quality Artificial Prostate Cancer Magnetic Resonance Images

Analysis of Apparent Diffusion Coefficient Value and Dynamic Contrast-Enhanced Magnetic Resonance Imaging Parameters of Prostate Cancer Patients after Diagnosis and Treatment with Magnetic Resonance Imaging

Contact Info

Product

Resources

About