Structure-preserving Image-quality Enhancement for 3D Synthetic FLAIR Using a 3D U-Net with Content and Style Losses
Abstract
PURPOSE: Synthetic MRI can generate multiple contrasts from a single acquisition, yet synthetic fluid-attenuated inversion recovery (FLAIR) generally shows lower quality than conventional FLAIR. We aimed to improve 3D synthetic FLAIR image quality using deep learning, without losing the scan-time advantage of synthetic MRI. METHODS: We studied 55 adults with inflammatory demyelinating diseases who underwent 3T MRI. For each participant, five 3D quantification using an interleaved Look-Locker acquisition sequence with T2 preparation (3D-QALAS) source images and a conventional 3D-FLAIR image were acquired, and a synthetic FLAIR image was generated from the 3D-QALAS data. We trained a deep learning model in which a 3D U-shaped convolutional network (U-Net)-based attention network predicted voxel-wise weights for generating FLAIR images from the five 3D-QALAS source images, using conventional 3D-FLAIR images as the reference. The model was trained with a combined loss function of mean squared error, content loss, and style loss. Agreement with the reference was assessed using image similarity and error metrics, lesion overlap using the Dice similarity coefficient, and overall image quality and focal lesion visibility using blinded radiologist ratings. Synthetic FLAIR and the deep learning-generated FLAIR images were compared using 2-sided Wilcoxon signed-rank tests; P < 0.05 was considered statistically significant. RESULTS: The deep learning-generated FLAIR images showed significantly higher agreement with the reference image than synthetic FLAIR images (all P < 0.001). Lesion overlap was higher with the deep learning-generated FLAIR images (median [interquartile range]: 0.642 [0.528-0.711] versus 0.487 [0.344-0.641]). Qualitatively, the deep learning-generated FLAIR images improved overall image quality and focal lesion visibility, but reader scores remained lower than those for the reference conventional 3D-FLAIR images (all P < 0.001). CONCLUSION: A deep learning-based approach applied to five 3D-QALAS source images improved the image quality of 3D synthetic FLAIR. These improvements may increase the clinical utility of 3D synthetic MRI for neuroradiologic assessment.
Related
- AI-driven quality enhancement of medical datasets: a closed-loop framework of synthesis, annotation, quality control, and integration
- Quality of Douyin videos on childhood asthma : A cross-sectional content analysis
- Quality and reliability of sarcopenia-related short videos on Douyin : A cross-sectional content analysis
- How well do large language models answer postoperative lumbar fusion questions? A blinded comparative analysis of accuracy, quality, readability, and safety-related content
- Comparative Evaluation of Large Language Model Interfaces in Third Molar Surgery Complication Scenarios: Response Quality, Clinical Content, Potential Clinical Risk, Readability, and Externally Observable Response Latency—A Cross-Sectional Comparative Benchmark Study
BibTeX
@article{kamio2026structurepreserving,
title = {{Structure-preserving Image-quality Enhancement for 3D Synthetic FLAIR Using a 3D U-Net with Content and Style Losses}},
author = {Satoru Kamio and Akifumi Hagiwara and Yujiro Otsuka and Moto Nakaya and Reina Hosoi and Yasunobu Hoshino and Yuji Tomizawa and Yuika Tsukamoto and Naohisa Hara and Junko Kikuta and Christina Andica and Toshiaki Akashi and Akihiko Wada and Hiroshi Kusahara and Ryutaro Yano and Taku Hatano and Osamu Abe and Koji Kamagata},
journal = {Magnetic Resonance in Medical Sciences},
year = {2026},
month = aug,
doi = {10.2463/mrms.mp.2026-0101},
url = {https://www.jstage.jst.go.jp/article/mrms/25/4/25_mp.2026-0101/_article},
note = {CC BY},
}
RIS
TY - JOUR
TI - Structure-preserving Image-quality Enhancement for 3D Synthetic FLAIR Using a 3D U-Net with Content and Style Losses
AU - Satoru Kamio
AU - Akifumi Hagiwara
AU - Yujiro Otsuka
AU - Moto Nakaya
AU - Reina Hosoi
AU - Yasunobu Hoshino
AU - Yuji Tomizawa
AU - Yuika Tsukamoto
AU - Naohisa Hara
AU - Junko Kikuta
AU - Christina Andica
AU - Toshiaki Akashi
AU - Akihiko Wada
AU - Hiroshi Kusahara
AU - Ryutaro Yano
AU - Taku Hatano
AU - Osamu Abe
AU - Koji Kamagata
PY - 2026
DA - 2026/08/01
JO - Magnetic Resonance in Medical Sciences
DO - 10.2463/mrms.mp.2026-0101
UR - https://www.jstage.jst.go.jp/article/mrms/25/4/25_mp.2026-0101/_article
AB - PURPOSE: Synthetic MRI can generate multiple contrasts from a single acquisition, yet synthetic fluid-attenuated inversion recovery (FLAIR) generally shows lower quality than conventional FLAIR. We aimed to improve 3D synthetic FLAIR image quality using deep learning, without losing the scan-time advantage of synthetic MRI. METHODS: We studied 55 adults with inflammatory demyelinating diseases who underwent 3T MRI. For each participant, five 3D quantification using an interleaved Look-Locker acquisition sequence with T2 preparation (3D-QALAS) source images and a conventional 3D-FLAIR image were acquired, and a synthetic FLAIR image was generated from the 3D-QALAS data. We trained a deep learning model in which a 3D U-shaped convolutional network (U-Net)-based attention network predicted voxel-wise weights for generating FLAIR images from the five 3D-QALAS source images, using conventional 3D-FLAIR images as the reference. The model was trained with a combined loss function of mean squared error, content loss, and style loss. Agreement with the reference was assessed using image similarity and error metrics, lesion overlap using the Dice similarity coefficient, and overall image quality and focal lesion visibility using blinded radiologist ratings. Synthetic FLAIR and the deep learning-generated FLAIR images were compared using 2-sided Wilcoxon signed-rank tests; P < 0.05 was considered statistically significant. RESULTS: The deep learning-generated FLAIR images showed significantly higher agreement with the reference image than synthetic FLAIR images (all P < 0.001). Lesion overlap was higher with the deep learning-generated FLAIR images (median [interquartile range]: 0.642 [0.528-0.711] versus 0.487 [0.344-0.641]). Qualitatively, the deep learning-generated FLAIR images improved overall image quality and focal lesion visibility, but reader scores remained lower than those for the reference conventional 3D-FLAIR images (all P < 0.001). CONCLUSION: A deep learning-based approach applied to five 3D-QALAS source images improved the image quality of 3D synthetic FLAIR. These improvements may increase the clinical utility of 3D synthetic MRI for neuroradiologic assessment.
SN - 1880-2206
SN - 1347-3182
ER -