Local Spatiotemporal Representation Learning for Longitudinally-consistent Neuroimage Analysis
Nov 28, 2022
Speakers
Mengwei Ren
Speaker · 0 followers
Martin Styner
Speaker · 0 followers
Kelly Botteron
Speaker · 0 followers
Guido Gerig
Speaker · 0 followers
About
Recent self-supervised advances in medical computer vision exploit the global and local anatomical self-similarity for pretraining prior to downstream tasks such as segmentation. However, current methods assume i.i.d. image acquisition, which is invalid in clinical study designs where follow-up longitudinal scans track subject-specific temporal changes. Further, existing self-supervised methods for medically-relevant image-to-image architectures exploit only spatial or temporal self-similarity and do so via a loss applied only at a single image-scale, with naive multi-scale spatiotemporal extensions collapsing to degenerate solutions. To these ends, this paper makes two contributions: (1) It presents a local and multi-scale spatiotemporal representation learning method for image-to-image architectures trained on longitudinal images. It exploits the spatiotemporal self-similarity of learned multi-scale intra-subject image features for pretraining and develops several feature-wise regularizations that avoid degenerate representations; (2) During finetuning, it proposes a surprisingly simple self-supervised segmentation consistency regularization to exploit intra-subject correlation. Benchmarked across various segmentation tasks, the proposed framework outperforms both well-tuned randomly-initialized baselines and current self-supervised techniques designed for both i.i.d. and longitudinal datasets. These improvements are demonstrated across both longitudinal neurodegenerative adult MRI and developing infant brain MRI and yield both higher performance and longitudinal consistency.Recent self-supervised advances in medical computer vision exploit the global and local anatomical self-similarity for pretraining prior to downstream tasks such as segmentation. However, current methods assume i.i.d. image acquisition, which is invalid in clinical study designs where follow-up longitudinal scans track subject-specific temporal changes. Further, existing self-supervised methods for medically-relevant image-to-image architectures exploit only spatial or temporal self-similarity a…
Organizer
NeurIPS 2022
Account · 954 followers
Like the format? Trust SlidesLive to capture your next event!
Professional recording and live streaming, delivered globally.
Sharing
Recommended Videos
Presentations on similar topic, category or speaker
Selective compression learning of Latent Representations for Variable-rate Image Compression
Watch later
Jooyoung Lee, …
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%
Deep Combinatorial Aggregation
Watch later
Yuesong Shen, …
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%
Reinforcement Learning and LTV at Spotify
Watch later
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%
Designing for Impact: Al-enabled Point-of-Care Imaging
Watch later
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%
The Dollar Street Dataset: Images Representing the Geographic and Socioeconomic Diversity of the World
Watch later
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%
In What Ways Are Deep Neural Networks Invariant and How Should We Measure This?
Watch later
Henry Kvinge, …
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%