Combinations of data augmentation methods and deep learning architectures for automatic pancreas segmentation on CT images are proposed and evaluated. Images from a public CT dataset of pancreas segmentation were used to evaluate the models. Baseline U-net and deep U-net were chosen for the deep learning models of pancreas segmentation. Methods of data augmentation included conventional methods, mixup, and random image cropping and patching (RICAP). Ten combinations of the deep learning models and the data augmentation methods were evaluated. Four-fold cross validation was performed to train and evaluate these models with data augmentation methods. The dice similarity coefficient (DSC) was calculated between automatic segmentation results and manually annotated labels and these were visually assessed by two radiologists. The performance of the deep U-net was better than that of the baseline U-net with mean DSC of 0.703–0.789 and 0.686–0.748, respectively. In both baseline U-net and deep U-net, the methods with data augmentation performed better than methods with no data augmentation, and mixup and RICAP were more useful than the conventional method. The best mean DSC was obtained using a combination of deep U-net, mixup, and RICAP, and the two radiologists scored the results from this model as good or perfect in 76 and 74 of the 82 cases.

Purpose To investigate the differences in the fiber architecture and related diffusion tensor imaging (DTI) parameters among different uterine layers of the normal human uterus in vivo at 3 Tesla (T). Materials and Methods DTI of the uterus was performed for nine healthy women of reproductive age on a 3T scanner. A volume of interest (VOI) was drawn for the outer myometrium (OM), junctional zone (JZ), and endometrium (EM). Apparent diffusion coefficient (ADC), fractional anisotropy (FA), and maximum fiber length were compared using paired Student's t-tests. Fibers were classified into four groups and were visually evaluated. Results ADC (×10⁻³ mm²/s) was highest for OM (1.12), followed by EM (0.97) and JZ (0.83) (OM versus JZ, P < 0.0001; JZ versus EM, P = 0.0001; OM versus EM, P = 0.0057). FA was highest for JZ (0.297), followed by OM (0.257) and EM (0.186) (OM versus JZ, P = 0.0002; JZ versus EM, P < 0.0001; OM versus EM, P < 0.0001). Fibers were longest in OM (42.0 mm), followed by JZ (34.2 mm) and EM (20.0 mm). Circularly oriented fibers were observed in 69% in OM and in 70% in JZ. Conclusion DTI of the uterus in vivo revealed layer-wise differences in the microstructure of the uterus. J. Magn. Reson. Imaging 2013;38:1494–1500.