Camila Gonzalez
Postdoctoral Scholar, Psychiatry
Bio
Camila González is a postdoctoral scholar at the Computational Neuroscience Laboratory at Stanford University. Her aim is to develop methods that allow the continual training of deep learning models for neuroimaging applications, suitable for dynamic settings with ongoing data collection. She has received multiple distinctions, including the MICCAI Young Scientist Award and the Best Presentation Award at the EuSoMII annual meeting. Her work was featured in outlets such as the Computer Vision News magazine and the AI-Ready Healthcare podcast. Outside her research, she presided the MICCAI student board for two years and was DEI chair in the first edition of the ContinualAI Unconference.
Professional Education
-
Doctor of Philosophy, Technical University of Darmstadt, Medical Image Computing (2023)
-
Master of Science, Technical University of Darmstadt, Computer Science (2020)
Community and International Work
-
DEI Chair for the ContinualAI (un)conference
Partnering Organization(s)
ContinualAI
Ongoing Project
Yes
Opportunities for Student Involvement
Yes
-
President of the MICCAI Student Board (MSB)
Partnering Organization(s)
MICCAI Society
Location
International
Ongoing Project
Yes
Opportunities for Student Involvement
Yes
All Publications
-
Lifelong nnU-Net: a framework for standardized medical continual learning.
Scientific reports
2023; 13 (1): 9381
Abstract
As the enthusiasm surrounding Deep Learning grows, both medical practitioners and regulatory bodies are exploring ways to safely introduce image segmentation in clinical practice. One frontier to overcome when translating promising research into the clinical open world is the shift from static to continual learning. Continual learning, the practice of training models throughout their lifecycle, is seeing growing interest but is still in its infancy in healthcare. We present Lifelong nnU-Net, a standardized framework that places continual segmentation at the hands of researchers and clinicians. Built on top of the nnU-Net-widely regarded as the best-performing segmenter for multiple medical applications-and equipped with all necessary modules for training and testing models sequentially, we ensure broad applicability and lower the barrier to evaluating new methods in a continual fashion. Our benchmark results across three medical segmentation use cases and five continual learning methods give a comprehensive outlook on the current state of the field and signify a first reproducible benchmark.
View details for DOI 10.1038/s41598-023-34484-2
View details for PubMedID 37296233
View details for PubMedCentralID PMC10256748
-
Med-NCA: Robust and Lightweight Segmentation with Neural Cellular Automata
Information Processing in Medical Imaging
2023: 705--716
View details for DOI 10.1007/978-3-031-34048-2_54
- Continual hippocampus segmentation with transformers IEEE/CVF Conference on Computer Vision and Pattern Recognition 2022: 3711--3720
-
Distance-based detection of out-of-distribution silent failures for Covid-19 lung lesion segmentation
Medical image analysis
2022; 82: 102596
View details for DOI 10.1016/j.media.2022.102596
-
Disentanglement enables cross-domain hippocampus segmentation
IEEE International Symposium on Biomedical Imaging
2022: 1--5
View details for DOI 10.1109/isbi52829.2022.9761560
-
M3d-CAM: A PyTorch library to generate 3D attention maps for medical deep learning
Bildverarbeitung für die Medizin
2021: 217--222
View details for DOI 10.1007/978-3-658-33198-6_52
- Self-supervised out-of-distribution detection for cardiac CMR segmentation Medical Imaging with Deep Learning 2021: 205--218
-
Detecting when pre-trained nnu-net models fail silently for covid-19 lung lesion segmentation
Medical Image Computing and Computer Assisted Intervention
2021: 304--314
View details for DOI 10.1007/978-3-030-87234-2_29