TY  - JOUR
AU  - Huang, L.
AU  - Thummerer, A.
AU  - Papadopoulou, C. I.
AU  - Corradini, S.
AU  - Belka, C.
AU  - Riboldi, M.
AU  - Kurz, C.
AU  - Landry, G.
TI  - Validation of patient-specific deep learning markerless lung tumor tracking aided by 4DCBCT.
JO  - Physics in medicine and biology
VL  - 70
IS  - 6
SN  - 0031-9155
CY  - Bristol
PB  - IOP Publ.
M1  - DKFZ-2025-00529
SP  - 065001
PY  - 2025
AB  - Objective. Tracking tumors with multi-leaf collimators and x-ray imaging can be a cost-effective motion management method to reduce internal target volume margins for lung cancer patients, sparing normal tissues while ensuring target coverage. To realize that, accurate tumor localization on x-ray images is essential. We aimed to develop a systematic method for automatically generating tumor segmentation ground truth (GT) on cone-beam computed tomography (CBCT) projections and use it to help refine and validate our patient-specific AI-based tumor localization model.Approach. To obtain the tumor segmentation GT on CBCT projections, we propose a 4DCBCT-aided GT generation pipeline consisting of three steps: breathing phase extraction and 10-phase 4DCBCT reconstruction, manual segmentation on phase 50
KW  - Lung Neoplasms: diagnostic imaging
KW  - Humans
KW  - Deep Learning
KW  - Cone-Beam Computed Tomography: methods
KW  - Four-Dimensional Computed Tomography: methods
KW  - Image Processing, Computer-Assisted: methods
KW  - 4DCBCT (Other)
KW  - AI (Other)
KW  - lung tumor targeting (Other)
KW  - markerless tracking (Other)
KW  - patient-specific (Other)
LB  - PUB:(DE-HGF)16
C6  - pmid:39978071
DO  - DOI:10.1088/1361-6560/adb89c
UR  - https://inrepo02.dkfz.de/record/299588
ER  -