Our work "Tetris-inspired detector with neural network for radiation mapping", led by our team members Ryotaro, former group member Shangjie, and NSE alumnus Jayson, has been accepted by Nature Communications. Conventional radiation detection often uses a "grid" of detector arrays of pixels, and inspired by the famous game “Tetris,” the Tetris-shaped detector is asymmetric and is thereby more efficient in detector pixel arrangement and can reach high resolution even with as few as 4 detector pixels. That is a great save of detector number of pixels and complexity. We even did a single-blind field testing, and with machine-learning data analysis, the method was able to locate the radioactive sources in the actual envrionment!