Machine-learned target volume delineation of F-FDG PET images after one cycle of induction chemotherapy.

Biological tumour volume (GTV) delineation on F-FDG PET acquired during induction chemotherapy (ICT) is challenging due to the reduced metabolic uptake and volume of the GTV. Automatic segmentation algorithms applied to F-FDG PET (PET-AS) imaging have been used for GTV delineation on F-FDG PET imaging acquired before ICT. However, their role has not been investigated in F-FDG PET imaging acquired after ICT. In this study we investigate PET-AS techniques, including ATLAAS a machine learned method, for accurate delineation of the GTV after ICT. Twenty patients were enrolled onto a prospective phase I study (FiGaRO). PET/CT imaging was acquired at baseline and 3 weeks following 1 cycle of induction chemotherapy. The GTV was manually delineated by a nuclear medicine physician and clinical oncologist. The resulting GTV was used as the reference contour. The ATLAAS original statistical model was expanded to include images of reduced metabolic activity and the ATLAAS algorithm was re-trained on the new reference dataset. Estimated GTV contours were derived using sixteen PET-AS methods and compared to the GTV using the Dice Similarity Coefficient (DSC). The mean DSC for ATLAAS, 60% Peak Thresholding (PT60), Adaptive Thresholding (AT) and Watershed Thresholding (WT) was 0.72, 0.61, 0.63 and 0.60 respectively. The GTV generated by ATLAAS compared favourably with manually delineated volumes and in comparison, to other PET-AS methods, was more accurate for GTV delineation after ICT. ATLAAS would be a feasible method to reduce inter-observer variability in multi-centre trials.