TY  - JOUR
AU  - Rabe, Moritz
AU  - Meliadò, Ettore F
AU  - Marschner, Sebastian N
AU  - Belka, Claus
AU  - Corradini, Stefanie
AU  - van den Berg, Cornelis A T
AU  - Landry, Guillaume
AU  - Kurz, Christopher
TI  - Patient-specific uncertainty calibration of deep learning-based autosegmentation networks for adaptive MRI-guided lung radiotherapy.
JO  - Physics in medicine and biology
VL  - 70
IS  - 10
SN  - 0031-9155
CY  - Bristol
PB  - IOP Publ.
M1  - DKFZ-2025-01015
SP  - 105018
PY  - 2025
AB  - Objective.Uncertainty assessment of deep learning autosegmentation (DLAS) models can support contour corrections in adaptive radiotherapy (ART), e.g. by utilizing Monte Carlo Dropout (MCD) uncertainty maps. However, poorly calibrated uncertainties at the patient level often render these clinically nonviable. We evaluated population-based and patient-specific DLAS accuracy and uncertainty calibration and propose a patient-specific post-training uncertainty calibration method for DLAS in ART.Approach.The study included 122 lung cancer patients treated with a low-field MR-linac (80/19/23 training/validation/test cases). Ten single-label 3D-U-Net population-based baseline models (BM) were trained with dropout using planning MRIs (pMRIs) and contours for nine organs-at-riks (OARs) and gross tumor volumes (GTVs). Patient-specific models (PS) were created by fine-tuning BMs with each test patient's pMRI. Model uncertainty was assessed with MCD, averaged into probability maps. Uncertainty calibration was evaluated with reliability diagrams and expected calibration error (ECE). A proposed post-training calibration method rescaled MCD probabilities for fraction images in BM (calBM) and PS (calPS) after fitting reliability diagrams from pMRIs. All models were evaluated on fraction images using Dice similarity coefficient (DSC), 95th percentile Hausdorff distance (HD95) and ECE. Metrics were compared among models for all OARs combined (n = 163), and the GTV (n = 23), using Friedman and posthoc-Nemenyi tests (α = 0.05).Main results.For the OARs, patient-specific fine-tuning significantly (p < 0.001) increased median DSC from 0.78 (BM) to 0.86 (PS) and reduced HD95 from 14 mm (BM) to 6.0 mm (PS). Uncertainty calibration achieved substantial reductions in ECE, from 0.25 (BM) to 0.091 (calBM) and 0.22 (PS) to 0.11 (calPS) (p < 0.001), without significantly affecting DSC or HD95 (p > 0.05). For the GTV, BM performance was poor (DSC = 0.05) but significantly (p < 0.001) improved with PS training (DSC = 0.75) while uncertainty calibration reduced ECE from 0.22 (PS) to 0.15 (calPS) (p = 0.45).Significance.Post-training uncertainty calibration yields geometrically accurate DLAS models with well-calibrated uncertainty estimates, crucial for ART applications.
KW  - Humans
KW  - Deep Learning
KW  - Uncertainty
KW  - Radiotherapy, Image-Guided: methods
KW  - Calibration
KW  - Lung Neoplasms: radiotherapy
KW  - Lung Neoplasms: diagnostic imaging
KW  - Magnetic Resonance Imaging
KW  - Organs at Risk: radiation effects
KW  - Radiotherapy Planning, Computer-Assisted: methods
KW  - MR-linac (Other)
KW  - Monte Carlo dropout (Other)
KW  - adaptive radiotherapy (Other)
KW  - autosegmentation (Other)
KW  - deep learning (Other)
KW  - epistemic uncertainty (Other)
KW  - uncertainty calibration (Other)
LB  - PUB:(DE-HGF)16
C6  - pmid:40340988
DO  - DOI:10.1088/1361-6560/add640
UR  - https://inrepo02.dkfz.de/record/301473
ER  -