Ultrafast Intelligent\nSensor for Integrated Biological\nFluorescence Imaging and Recognition

Fluorescence imaging and recognition are core technologies\nin targeted\nmedicine, pathological surgery, and biomedicine. However, current\nimaging and recognition systems are separate, requiring repeated data\ntransfers for imaging and recognition that lead to delays and inefficiency,\nhindering the capture of rapidly changing physiological processes\nand biological phenomena. To address these problems, we propose an\nintegrated intelligent sensor for biological fluorescence imaging\nand ultrafast recognition. This sensor integrates an imaging system\nbased on a photodetector array and a recognition system based on neural\nnetworks on a single chip, featuring a highly compact structure, a\ncontinuously adjustable optical response, and reconfigurable electrical\nperformance. The unified architecture of the imaging and recognition\nsystems enables ultrafast recognition (19.63 μs) of tumor margins.\nAdditionally, the special organic materials and bulk heterojunction\nstructure endow the photodetector array with strong wavelength dependence,\nachieving high specific detectivity (3.06 × 10<sup>12</sup> Jones)\nin the narrowband near-infrared range commonly used in biomedical\nimaging (600–800 nm). After training, the sensor can accurately\nrecognize biological fluorescence edges in real time, even under interference\nfrom other colored light noise. Benefiting from its rapidity and high\naccuracy, we demonstrated a simulated surgical experiment showcasing\ntumor edge fluorescence imaging, recognition, and cutting. This integrated\napproach holds the potential to establish a novel paradigm for designing\nand manufacturing intelligent medical sensors.