Uncertainty-Aware Lumbar Stenosis Classifier
Journal project where I developed a deep learning classifier that grades central-canal stenosis severity from intervertebral volumes on lumbar sagittal MRI, and — crucially — quantifies how confident each prediction is.
Two complementary uncertainty-quantification strategies, Monte-Carlo dropout and test-time augmentation, turn the grading network into a stochastic predictor with mean and variance, so low-confidence cases can be flagged for clinical review rather than silently mis-graded.
Published in JOR Spine, 2026.
Authors
Maria Monzon
(she/her)
Computer Vision & Medical AI Researcher
PhD candidate at ETH Zurich developing robust and trustworthy deep learning for medical image analysis — spine and cardiac MRI, multimodal biomedical data, and uncertainty quantification. Previously a computer-vision researcher at BASF, where I deployed models to production in regulated, GLP-certified environments. I care about efficient code and reproducible research.