A computational pipeline for quantification of pulmonary infections in small animal models using serial PET-CT imaging

EJNMMI Research

Table 3 Evaluation of inter/intra-observer agreements over CT and PET image segmentations is shown

Animals	Observer	Observer	Evaluation of PET	Evaluation of
	agreement rates	agreement rates	segmentation between	interactive
	for CT volume	for PET volume	consecutive threshold	segmentation for
			levels (difference by 5%)	pathological lung
Rabbit	Intra(%) 84.7±6.0	Intra (%) 92.3±14.4	DSC (%) 86.0±3.6	DSC(%) 86.0±7.1
	Inter (%) 80.6±9.5	Inter (%) 89.8±14.7	HD (mm) 13.2±2.3	HD (mm) 14.2±2.1
Ferret	Intra(%) 80.3±4.4	Intra(%) 92.7±12.6	DSC(%) 84.0±5.8	DSC(%) 83.4±8.6
	Inter (%) 77.2±11.2	Inter (%) 81.5±16.2	HD (mm) 10.6±1.6	HD (mm) 4.0±1.4
Mouse	Intra (%) 94.7±1.3	Intra (%) 86.6±2.4	DSC(%) 86.5±1.1	DSC(%) 95.1±2.8
	Inter (%) 89.5±3.7	Inter (%) 83.2±3.6	HD (mm) 10.8±1.8	HD (mm) 0.26±0.48

DSC and HD measurements are also reported for evaluating the interactive and adaptive thresholding-based segmentation algorithms compared to ground truths.