방사선의학 웹진

바로가기  >

이달의 연구자 바로가기 Click!

Abstract
Purpose: AI-driven scan time reduction is rapidly transforming medical imaging with benefits such as improved patient comfort and enhanced efficiency. A Dual Contrastive Learning Generative Adversarial Network (DCLGAN) was developed to predict full-time PET scans from shorter, noisier scans, improving challenges in imaging patients with movement disorders.

Patients and methods: 18 F FP-CIT PET/CT data from 391 patients with suspected Parkinsonism were used [250 training/validation, 141 testing (hospital A)]. Ground truth (GT) images were reconstructed from 15-minute scans, while denoised images (DIs) were generated from 1-, 3-, 5-, and 10-minute scans. Image quality was assessed using normalized root mean square error (NRMSE), peak signal-to-noise ratio (PSNR), structural similarity index measure (SSIM), visual analysis, and clinical metrics like BP ND and ISR for diagnosis of non-neurodegenerative Parkinson disease (NPD), idiopathic PD (IPD), and atypical PD (APD). External validation used data from 2 hospitals with different scanners (hospital B: 1-, 3-, 5-, and 10-min; hospital C: 1-, 3-, and 5-min). In addition, motion artifact reduction was evaluated using the Dice similarity coefficient (DSC).

Results: In hospital A, NRMSE, PSNR, and SSIM values improved with scan duration, with the 5-minute DIs achieving optimal quality (NRMSE 0.008, PSNR 42.13, SSIM 0.98). Visual analysis rated DIs from scans ≥3 minutes as adequate or higher. The mean BP ND differences (95% CI) for each DIs were 0.19 (-0.01, 0.40), 0.11 (-0.02, 0.24), 0.08 (-0.03, 0.18), and 0.01 (-0.06, 0.07), with the CIs significantly decreasing. ISRs with the highest effect sizes for differentiating NPD, IPD, and APD (PP/AP, PP/VS, PC/VP) remained stable post-denoising. External validation showed 10-minute DIs (hospital B) and 1-minute DIs (hospital C) reached benchmarks of hospital A's image quality metrics, with similar trends in visual analysis and BP ND CIs. Furthermore, motion artifact correction in 9 patients yielded DSC improvements from 0.89 to 0.95 in striatal regions.

Conclusions: The DL-model is capable of generating high-quality 18 F FP-CIT PET images from shorter scans to enhance patient comfort, minimize motion artifacts, and maintain diagnostic precision. Furthermore, our study plays an important role in providing insights into how imaging quality assessment metrics can be used to determine the appropriate scan duration for different scanners with varying sensitivities.

▶ 그림 1. 딥러닝 모델을 통해 복원된 파킨슨 증상 환자들의 PET/CT 영상 비교. 원본(Raw) 영상은 촬영 시간이 짧을수록 노이즈가 심해지지만, 노이즈 제거(Denoised) 영상은 5분 촬영 데이터만으로도 15분 원본 영상과 거의 유사한 고해상도 화질을 보여주며, 진단에 필요한 선조체(Striatum) 섭취 패턴이 명확히 유지되고 있다.

▶ 그림 2. 촬영 도중 움직임이 발생한 환자의 예. 3분 촬영한 영상을 딥러닝 모델로 노이즈를 제거한 결과(왼쪽) 15분 촬영한 영상(오른쪽)에 비해 움직임이 없는 선명한 영상을 보여주고 있다.

Affiliations

Hyunkyung Han 1, Kyobin Choo 2, Tae Joo Jeon 3 4, Sangwon Lee 3, Seungbeom Seo 5, Dongwoo Kim 6, Sun Jung Kim 7, Suk Hyun Lee 8, Mijin Yun 3
1Departments of Artificial Intelligence.
2Computer Science, Yonsei University.
3Department of Nuclear Medicine, Yonsei University College of Medicine.
4Department of Nuclear Medicine, Gangnam Severance Hospital.
5Yonsei University College of Medicine.
6Department of Nuclear Medicine, Hallym University Sacred Heart Hospital, College of Medicine, Hallym University, Seoul.
7Department of Nuclear Medicine, National Health Insurance Service Hospital, Gyeonggi-do.
8Department of Radiology, Division of Nuclear Medicine, Hallym University Kangnam Sacred Heart Hospital, Hallym University College of Medicine, Seoul, Republic of Korea.