CNN-based Organ Segmentation in CT Imaging

The CNN-based organ segmentation in CT studies in RECOMIA has been used in multiple studies [7 (link)–12 (link)]. These studies were approved by the Regional Ethical Review Board (#295/08) and were performed following the Declaration of Helsinki. Patients and image acquisition have been described previously [7 (link), 8 (link), 10 , 11 (link)].
A group of experienced radiologists and nuclear medicine physicians manually segmented different organs using the RECOMIA platform. The organs included 77 bones and 23 soft tissue organs (Table 1). Not all organs were annotated in all CT studies, which had to be handled in the training process. A dataset of approximately 13,000 manual organ segmentations in 339 images was used to train the CNNs.
Table 1

List of the 100 different organs segmented throughout the studies grouped by type

Bones	Organs	Soft tissue	Organs
Skull	1	Adrenal gland	2
Mandible	1	Brain	1
Cervical vertebrae	7	Lungs	2
Thoracic vertebrae	12	Trachea	1
Lumbar vertebrae	5	Bronchi	2
Ribs	24	Heart	1
Sacrum and coccyx	1	Aorta	1
Hip bones	2	Ventricle	1
Scapulae	2	Gastrointestinal tract	1
Clavicles	2	Liver	1
Sternum manubrium	1	Gallbladder	1
Sternum body	1	Spleen	1
Humerus	2	Pancreas	1
Radius	2	Kidneys	2
Ulna	2	Urinary bladder	1
Hand	2	Prostate	1
Femur	2	Testes	1
Tibia	2	Musc. gluteus maximus	2
Fibula	2
Patella	2
Foot	2
Total	77		23

A separate test set of 10 patients (5 male/5 female) was used to test the method and obtain data on inter-observer variability. Each test case was segmented independently by two different readers. Ten organs (prostate only for male patients) were segmented in each CT study.
All images used for training, validation, and test had a pixel spacing of 1.36 mm in slices and a distance between slices of 3 mm. Images with different pixel spacing can still be segmented by resampling the images using trilinear interpolation before running the networks. The resulting segmentation is then resampled to the image resolution using the nearest neighbour interpolation.