Skip to main content
. 2022 Jun 27;5:78. doi: 10.1038/s41746-022-00626-5

Fig. 1. System architecture and clinical scheme of ARISES.

Fig. 1

A T1D subject is equipped with CGM and the wristband to measure glucose levels and vital signs, both of which communicate with the ARISES app via Bluetooth connectivity and provide input data for the deep learning models. The wearable devices in the system are marked by *. The data collected in phase I are used to train a population model with meta-learning, which is then fine-tuned in phase II to develop personalized models.