Self-supervised learning leads to improved performance in biparametric prostate MRI classification

Abstract

Background and objective: Develop two-dimensional self-supervised learning (SSL) models which can be used in volumetric imaging and demonstrate their application in volumetric prostate bi-parametric MRI (bpMRI) classification tasks.

Methods: Prostate multiparametric MRI (mpMRI) data from 12 distinct European centers were used to train two SSL methods. We transfer these models to classification tasks in volumetric prostate bpMRI using 3 attention-based multiple instance learning (MIL) methods with T2-weighted (T2) or bpMRI studies. Three prostate cancer (PCa) tasks were considered: PCa diagnosis (D-PCa), clinically significant PCa (csPCa) diagnosis (D-csPCa), and virtual biopsy to confirm csPCa (VB). All approaches were compared with a fully supervised learning (FSL) baseline. Performance was assessed using the area under the receiver operating curve (AUC) and using both 5-fold cross-validation and a hold-out test set, and attention scores were analyzed. Finally, sensitivity analyses were performed for training and pre-training dataset size, data domain (MRI vs. natural images), and architecture.

Results: Two 2D SSL methods were trained using 6,798 studies (1,722,978 DICOM images) and their downstream performance was assessed on 3D tasks (n=1,622, n=1,615 and n=1,295 bmMRI studies for D-PCa, D-csPCa and VB, respectively). We show these models are comparable or better than FSL baseline models trained on the same data: AUC_SSL=0.82 and AUC_FSL=0.75 for bpMRI D-PCa (p=0.017), AUC_SSL=0.73 and AUC_FSL=0.68 for T2 D-csPCa (p=0.043) and AUC_SSL=0.73 and AUC_FSL=0.65 for bpMRI VB, while other models showed no differences (p>0.05). Learning curve analyses show that SSL-based models required fewer training data to perform similarly, while sensitivity analyses showed that large amounts of domain-specific pre-training data are essential for optimal performance. Attention scores correlate with lesion location.

Conclusion: Data with no annotations was used to train SSL models which were more data efficient and performed better than FSL models, highlighting the importance of large-scale data collection efforts in biomedical imaging.

Keywords: Multiple-instance learning; Prostate multi-parametric MRI; Self-supervised learning.