EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Automated segmentation of lesions and organs at risk on [68Ga]Ga-PSMA-11 PET/CT images using self-supervised learning with Swin UNETR

Elmira Yazdani, Najme Karamzadeh-Ziarati, Seyyed Saeid Cheshmi, Mahdi Sadeghi, Parham Geramifar, Habibeh Vosoughi, Mahmood Kazemi Jahromi, Saeed Reza Kheradpisheh

Cancer Imaging · 29.02.2024 · 24 (1)

DOI: 10.1186/s40644-024-00675-x ↗

Kurz gefasst

Fachartikel in der Zeitschrift Cancer Imaging, online erschienen am 29.02.2024. Verfasst von Elmira Yazdani und 7 weiteren Autor:innen. Beteiligte Einrichtung: Medical Physics Department, School of Medicine, Iran University of Medical Sciences, Tehran, 14155-6183, Iran. Hauptthemen laut MEDLINE: Positron Emission Tomography Computed Tomography, Prostatic Neoplasms, Castration-Resistant. Die Arbeit wurde bisher 49-mal zitiert (Crossref) und nennt 59 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Iran University of Medical Sciences.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background

Prostate-specific membrane antigen (PSMA) PET/CT imaging is widely used for quantitative image analysis, especially in radioligand therapy (RLT) for metastatic castration-resistant prostate cancer (mCRPC). Unknown features influencing PSMA biodistribution can be explored by analyzing segmented organs at risk (OAR) and lesions. Manual segmentation is time-consuming and labor-intensive, so automated segmentation methods are desirable. Training deep-learning segmentation models is challenging due to the scarcity of high-quality annotated images. Addressing this, we developed shifted windows UNEt TRansformers (Swin UNETR) for fully automated segmentation. Within a self-supervised framework, the model’s encoder was pre-trained on unlabeled data. The entire model was fine-tuned, including its decoder, using labeled data.

Methods

In this work, 752 whole-body [ 68 Ga]Ga-PSMA-11 PET/CT images were collected from two centers. For self-supervised model pre-training, 652 unlabeled images were employed. The remaining 100 images were manually labeled for supervised training. In the supervised training phase, 5-fold cross-validation was used with 64 images for model training and 16 for validation, from one center. For testing, 20 hold-out images, evenly distributed between two centers, were used. Image segmentation and quantification metrics were evaluated on the test set compared to the ground-truth segmentation conducted by a nuclear medicine physician.

Results

The model generates high-quality OARs and lesion segmentation in lesion-positive cases, including mCRPC. The results show that self-supervised pre-training significantly improved the average dice similarity coefficient (DSC) for all classes by about 3%. Compared to nnU-Net, a well-established model in medical image segmentation, our approach outperformed with a 5% higher DSC. This improvement was attributed to our model’s combined use of self-supervised pre-training and supervised fine-tuning, specifically when applied to PET/CT input. Our best model had the lowest DSC for lesions at 0.68 and the highest for liver at 0.95.

Conclusions

We developed a state-of-the-art neural network using self-supervised pre-training on whole-body [ 68 Ga]Ga-PSMA-11 PET/CT images, followed by fine-tuning on a limited set of annotated images. The model generates high-quality OARs and lesion segmentation for PSMA image analysis. The generalizable model holds potential for various clinical applications, including enhanced RLT and patient-specific internal dosimetry.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Positron Emission Tomography Computed TomographyProstatic Neoplasms, Castration-ResistantMaleHumansGallium RadioisotopesOrgans at RiskTissue DistributionSupervised Machine LearningImage Processing, Computer-Assisted

Stichwörter der Autor:innen

Deep-learningLesion detectionNeural networkPET/CTPSMAProstate cancerSegmentationSelf-supervised learningSwin UNETR

Substanzen

PSMA-11Gallium Radioisotopes

Autor:innen und Institutionen

  • Medical Physics Department, School of Medicine, Iran University of Medical Sciences, Tehran, 14155-6183, Iran.
  • Fintech in Medicine Research Center, Iran University of Medical Sciences, Tehran, Iran.
  • Research Center for Nuclear Medicine, Tehran University of Medical Sciences, Tehran, Iran.
  • Department of Computer and Data Sciences, Faculty of Mathematical Sciences, Shahid Beheshti University, Tehran, Iran.
  • Medical Physics Department, School of Medicine, Iran University of Medical Sciences, Tehran, 14155-6183, Iran. sadeghi.m@iums.ac.ir.
  • Fintech in Medicine Research Center, Iran University of Medical Sciences, Tehran, Iran. sadeghi.m@iums.ac.ir.

Förderung

  • Iran University of Medical Sciences

Zitieren

APA

Yazdani, E., Karamzadeh-Ziarati, N., Cheshmi, S. S., Sadeghi, M., Geramifar, P., Vosoughi, H., Jahromi, M. K., & Kheradpisheh, S. R. (2024). Automated segmentation of lesions and organs at risk on [68Ga]Ga-PSMA-11 PET/CT images using self-supervised learning with Swin UNETR. Cancer Imaging, 24(1). https://doi.org/10.1186/s40644-024-00675-x

Vancouver

Yazdani E, Karamzadeh-Ziarati N, Cheshmi SS, Sadeghi M, Geramifar P, Vosoughi H, et al. Automated segmentation of lesions and organs at risk on [68Ga]Ga-PSMA-11 PET/CT images using self-supervised learning with Swin UNETR. Cancer Imaging. 2024;24(1). doi:10.1186/s40644-024-00675-x

BibTeX
@article{Yazdani2024,
  author = {Yazdani, Elmira and Karamzadeh-Ziarati, Najme and Cheshmi, Seyyed Saeid and Sadeghi, Mahdi and Geramifar, Parham and Vosoughi, Habibeh and Jahromi, Mahmood Kazemi and Kheradpisheh, Saeed Reza},
  title = {Automated segmentation of lesions and organs at risk on [68Ga]Ga-PSMA-11 PET/CT images using self-supervised learning with Swin UNETR},
  journal = {Cancer Imaging},
  year = {2024},
  volume = {24},
  number = {1},
  doi = {10.1186/s40644-024-00675-x}
}
Kostenlos starten →