A Novel and Automated Approach to Classify Radiation Induced Lung Tissue Damage on CT Scans

SIMPLE SUMMARY: Radiation-induced lung damage (RILD) is a common side effect of treating lung cancer with radiotherapy (RT). RILD is visible on CT imaging, and its radiological appearance can vary dramatically from patient to patient as well as across different sub-regions of the lung and treatment volumes. A classification system for RILD able to differentiate radiological damage on a local level would allow us to better understand the underlying patterns of RILD, see how they change over time post irradiation, and link it with clinical outcomes. In this work we propose a five-class morphological lung tissue classification system that can describe parenchymal tissue changes at the voxel level. The classifier was implemented in a fully automated manner using an optimised deep-learning method, then trained and tested using data acquired through a multi-centre clinical trial. The proposed method performed well on an unseen testing dataset. The automated segmentation achieved considerable overlap with manual segmentations (ranging between 26% and 98% for the five classes) and was graded as acceptable by a clinical expert in 88% of cases. This demonstrates it to be suitable for application on a large dataset to help uncover different patterns of changes in the population. ABSTRACT: Radiation-induced lung damage (RILD) is a common side effect of radiotherapy (RT). The ability to automatically segment, classify, and quantify different types of lung parenchymal change is essential to uncover underlying patterns of RILD and their evolution over time. A RILD dedicated tissue classification system was developed to describe lung parenchymal tissue changes on a voxel-wise level. The classification system was automated for segmentation of five lung tissue classes on computed tomography (CT) scans that described incrementally increasing tissue density, ranging from normal lung (Class 1) to consolidation (Class 5). For ground truth data generation, we employed a two-stage data annotation approach, akin to active learning. Manual segmentation was used to train a stage one auto-segmentation method. These results were manually refined and used to train the stage two auto-segmentation algorithm. The stage two auto-segmentation algorithm was an ensemble of six 2D Unets using different loss functions and numbers of input channels. The development dataset used in this study consisted of 40 cases, each with a pre-radiotherapy, 3-, 6-, 12-, and 24-month follow-up CT scans (n = 200 CT scans). The method was assessed on a hold-out test dataset of 6 cases (n = 30 CT scans). The global Dice score coefficients (DSC) achieved for each tissue class were: Class (1) 99% and 98%, Class (2) 71% and 44%, Class (3) 56% and 26%, Class (4) 79% and 47%, and Class (5) 96% and 92%, for development and test subsets, respectively. The lowest values for the test subsets were caused by imaging artefacts or reflected subgroups that occurred infrequently and with smaller overall parenchymal volumes. We performed qualitative evaluation on the test dataset presenting manual and auto-segmentation to a blinded independent radiologist to rate them as ‘acceptable’, ‘minor disagreement’ or ‘major disagreement’. The auto-segmentation ratings were similar to the manual segmentation, both having approximately 90% of cases rated as acceptable. The proposed framework for auto-segmentation of different lung tissue classes produces acceptable results in the majority of cases and has the potential to facilitate future large studies of RILD.