Deep Learning-based Quantification of Abdominal Subcutaneous and Visceral Fat Volume on CT Images.

RATIONALE AND OBJECTIVES: Develop a deep learning-based algorithm using the U-Net architecture to measure abdominal fat on computed tomography (CT) images. MATERIALS AND METHODS: Sequential CT images spanning the abdominal region of seven subjects were manually segmented to calculate subcutaneous fat (SAT) and visceral fat (VAT). The resulting segmentation maps of SAT and VAT were augmented using a template-based data augmentation approach to create a large dataset for neural network training. Neural network performance was evaluated on both sequential CT slices from three subjects and randomly selected CT images from the upper, central, and lower abdominal regions of 100 subjects. RESULTS: Both subcutaneous and abdominal cavity segmentation images created by the two methods were highly comparable with an overall Dice similarity coefficient of 0.94. Pearson's correlation coefficients between the subcutaneous and visceral fat volumes quantified using the two methods were 0.99 and 0.99 and the overall percent residual squared error were 5.5% and 8.5%. Manual segmentation of SAT and VAT on the 555 CT slices used for testing took approximately 46 hours while automated segmentation took approximately 1 minute. CONCLUSION: Our data demonstrates that deep learning methods utilizing a template-based data augmentation strategy can be employed to accurately and rapidly quantify total abdominal SAT and VAT with a small number of training images.

Nanocomposite liposomes for pH-controlled porphyrin release into human prostate cancer cells.

It is both challenging and desirable to have drug sensitizers released at acidic tumor pH for photodynamic therapy in cancer treatment. A pH-responsive carrier was prepared, in which fumed silica-attached 5,10,15,20-tetrakis(4-trimethylammoniophenyl)porphyrin (TTMAPP) was encapsulated into 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC) nanocomposite liposomes. The sizes of agglomerates were determined by dynamic light scattering to be 115 nm for silica and 295 nm for silica-TTMAPP-DOPC liposomes. Morphological changes were also found in TEM images, showing liposome formation at pH 8.5 but collapse upon silanol protonation. TTMAPP release is enhanced from 13% at pH 7.5 to 80% at pH 2.3, as determined spectrophotometrically through dialysis membranes. Fluorescence emission of TTMAPP encapsulated in the dry film of liposomes was reduced to half at pH 8.6 when compared to that at pH 5.4, while the production of singlet oxygen was quintupled at pH 5.0 compared to pH 7.6. Upon treatment of human prostate cancer cells with liposomes containing 6.7 µM, 13 µM, 17 µM and 20 µM TTMAPP, the cell viabilities were determined to be 60%, 18%, 20% and 5% at pH 5.4; 58%, 30%, 25% and 10% at pH 6.3; and 90%, 82%, 68% and 35% at pH 7.4, respectively. Light-induced apoptosis in cancerous cells was only observed in the presence of liposomes at pH 6.3 and pH 5.4 but not at pH 7.4, as indicated by chromatin condensation.