Towards Automated Detection and Localization of Radiation-Induced Pneumonitis based on T2-Maps of Lung Cancer Patients

Motivation: Repeated follow-up CT-scans, assessed by a radiologist, are still necessary to infer and monitor potential radiation-induced pneumonitis (RP) in patients after lung radiotherapy. Goal(s): We aimed at developing a workflow based on MRI that would allow the automated identification of RP patients and the detection of the affected lung area. Approach: T2-maps were acquired 2-3 months after the end of MRI-guided radiotherapy for 22 patients and T2 values investigated in high-dose lung regions. Results: T2-values in the high-dose regions showed significant differences between RP and non-RP patients and T2-maps might allow an automatic segmentation of the RP region. Impact: The acquisition of T2-maps for lung tumor patients 2-3 months after the end of radiotherapy could replace the repeated follow-up CT-scans for these patients and could also allow to automatically identify radiation-induced pneumonitis patients and the affected lung area.