Our solution is able to noninvasively reconstruct 3D skin microstructuremicroangiography with micrometer-scale resolutions for diagnoses of skin disorders. Moreover, various skin parameters can be quantitatively evaluated from 3D images according to the proposed algorithms, including the thickness of stratum corneum, the thickness of epidermis, water concentration, collagen concentration,skin roughness. The developed technique can be applied to different clinical environments including hospitalscosmetic clinics for diagnosesanalyses of skin.

In order to develop an AI model that can accuratelycompletely segment coronary arteries, our team, the TW-CVAI, has established a training dataset composed of strictly verified annotations of coronary lumen boundaries in coronary CT angiography (CCTA). We designed a deep learning model, two-channel 3D-UNet, with a priori prerequisite (vesselness prior) to facilitate identification of vascular structures. The final model, the TaiCAD-Net, greatly shortens the CCTA interpretation time from 6 hours to 10 minutes, with the overall segmentation accuracy of 86 by Dice similarity coefficient.