Brain-ID: Learning Contrast-agnostic Anatomical Representations for Brain Imaging

TL;DR

Brain-ID is introduced, an anatomical representation learning model for brain imaging that achieves state-of-the-art performance in all tasks on different MRI modalities and CT, and more importantly, preserves its performance on low-resolution and small datasets.

Abstract

Recent learning-based approaches have made astonishing advances in calibrated medical imaging like computerized tomography (CT), yet they struggle to generalize in uncalibrated modalities -- notably magnetic resonance (MR) imaging, where performance is highly sensitive to the differences in MR contrast, resolution, and orientation. This prevents broad applicability to diverse real-world clinical protocols. We introduce Brain-ID, an anatomical representation learning model for brain imaging. With the proposed "mild-to-severe" intra-subject generation, Brain-ID is robust to the subject-specific brain anatomy regardless of the appearance of acquired images (e.g., contrast, deformation, resolution, artifacts). Trained entirely on synthetic data, Brain-ID readily adapts to various downstream tasks through only one layer. We present new metrics to validate the intra- and inter-subject robustness of Brain-ID features, and evaluate their performance on four downstream applications, covering contrast-independent (anatomy reconstruction/contrast synthesis, brain segmentation), and contrast-dependent (super-resolution, bias field estimation) tasks. Extensive experiments on six public datasets demonstrate that Brain-ID achieves state-of-the-art performance in all tasks on different MRI modalities and CT, and more importantly, preserves its performance on low-resolution and small datasets. Code is available at https://github.com/peirong26/Brain-ID.

Figures (10)

• Figure 1: Brain-ID's data generator on the fly. Given the brain segmentation labels of a subject, we randomly generate a deformation field, and synthesize intra-subjecct samples featuring various contrast intensities and corruption levels (\ref{['sec: generator']}).
• Figure 2: Brain-ID's contrast-agnostic anatomical representation learning framework.
• Figure 3: (a) Intra-subject and (b) inter-subject robustness of Brain-ID features.
• Figure 4: Qualitative comparisons on downstream tasks of reconstruction (Recon), super-resolution (SR), and segmentation (Seg), between Brain-ID, the baseline SCRATCH, and the state-of-the-art methods CIFLchua2023contrast, SynthSRIglesias2023SynthSRAP (Recon and SR), SAMSEGCerri2020ACM (Seg). The visualized testing examples are taken from: AIBL-FLAIR Fowler2021FifteenYO for Recon, AIBL-T1w Fowler2021FifteenYO for SR, and OASIS-CT LaMontagne2018OASIS3LN for Seg. The mint circles highlight some details.
• Figure 5: SCRATCH, which is well trained on HF T1w scans, produces highly descriptive features for HF T1w images ($1^{\text{st}}$ row), but does not preserve the same high quality useful for downstream tasks when handling LF ($2^{\text{nd}}$ row) or other contrasts ($3^{\text{rd}}$ row).