The HeaortaNet is developed by the TW-CVAI team. The HeaortaNet is a deep learning model trained by 70,000 axial images from 200 patients with verified annotations of the pericardiumaorta. It shortens the time for data processing from 60 minutes, by manual segmentation of both pericardiumaorta, to 0.4 seconds. The segmentation accuracy, as assessed by dice similarity coefficient, is 94.8 for the pericardium,91.6 for the aorta. The imaging-based Cardiovascular Risk Prediction module was constructed by analyzing data from the National Health Insurance Databank.

Two parts in this integrated creative invention: deep processing for frozen aquatic meal productshigh-valued utilization for byproducts. The deep processing technology for fish meal is the commodities without thawingwashing process while cooking. The technology also can remove fishy smells. Byproducts are full of rich nutrition but usually were used as animal feeding food, sold by low prices,evenly discarded. This technology is developed to extract nutrition factors in fish bone/scales by using an autoclave system with a series of heatingpressure procedures.

This invention targets at value-added increase for fish bone/scale by-products through a series of manufacturing units of steaming, pressure/ temperature,drying. The core design of pressure/temperature in terms of powdered machining equipment consists of a high pressure chamber, heating components, superheated steam,operations experiments as well.

The technology realizes a perovskite nanocrystal-polymer composite material, which has extremely high stability at room temperaturehigh temperature. It can be used for 3D printing, shaped into fluorescent microfilaments,can also be used for white light-emitting diodes. The application of this air-stabile perovskite nanocrystal-polymer composite material in the 3D printing industry, texti

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.

By combining the thermal evaporation and spin-coating techniques, we are able to demonstrate synthesis of 3 cm x 3 cm and 1.5 cm x 1.5 cm organometallic halide perovskite light-emitting diodes with electroluminescence at 538 nm and 405 nm, respectively. Other than the advantage in large device area, to our best knowledge, the demonstrated violet LEDs have produced the shortest emission wavelength available from current hybrid perovskites. The luminance of our green LEDs reach a maximum value of 9967 cd/m2 with an EQE of 0.88% and color purity of 95%. Moreover, it is possible to generate white light emission by combining the violet perovskite LEDs with visible light phosphor.

(1) Ｇel-type solid electrolytes and All solid-state electrolytes for Lithium Ion Battery (LIBs) (2) Fabrication of dye-sensitized solar cells by printing process and its applications on Internet-of-Things (IoT) - MOST 2019 Award (3) Silicon-carbon coating anode materials for LIBs: SiLican startup company growth and global business layout (4) Power enhanced Intermediate Temperature Solid State Fuel Cell (SOFC) (5) Worldwide highest conversion efficiency (19.4%) perovskite solar cells (6) Industrial linking and Product design : LevioPole and Smart motorcycle suitcase