Search

Public imaging datasets are critical for the development and evaluation of automated tools in cancer imaging. Unfortunately, many do not include annotations or image-derived features, complicating their downstream analysis. Artificial intelligence-based annotation tools have been shown to achieve acceptable performance and thus can be used to automatically annotate large datasets. As part of the effort to enrich public data available within NCI Imaging Data Commons (IDC), here we introduce AI-generated annotations for two collections of computed tomography images of the chest, NSCLC-Radiomics, and the National Lung Screening Trial. Using publicly available AI algorithms we derived volumetric annotations of thoracic organs at risk, their corresponding radiomics features, and slice-level annotations of anatomical landmarks and regions. The resulting annotations are publicly available within IDC, where the DICOM format is used to harmonize the data and achieve FAIR principles. The annotations are accompanied by cloud-enabled notebooks demonstrating their use. This study reinforces the need for large, publicly accessible curated datasets and demonstrates how AI can be used to aid in cancer imaging.

Automatic segmentation of tumours and organs at risk can function as a useful support tool in radiotherapy treatment planning as well as for validating radiomics studies on larger cohorts. In this paper, we developed robust automatic segmentation methods for the delineation of gross tumour volumes (GTVs) from planning Computed Tomography (CT) and FDG-Positron Emission Tomography (PET) images of head and neck cancer patients. The data was supplied as part of the MICCAI 2020 HECKTOR challenge. We developed two main volumetric approaches: A) an end-to-end volumetric approach and B) a slice-by-slice prediction approach that integrates 3D context around the slice of interest. We exploited differences in the representations provided by these two approaches by ensembling them, obtaining a Dice score of 66.9% on the held out validation set. On an external and independent test set, a final Dice score of 58.7% was achieved.