Precise diagnosis of cancer relies on comprehensive interpretation of histology images by experienced pathologists. The preparation, digitalisation, and interpretation could be weeks, involving a massive amount of time and resources. Here, we propose the application of advanced deep learning technologies to full-spectral autofluorescence lifetime microscopy to significantly decrease the time and resources used for lung cancer diagnosis.
Autofluorescence lifetime images are stitched together for rapid visual recognition of lung cancer per the lifetime contrast across various emission wavelengths. The corresponding intensity images are translated to virtual histology images for precise recognition by pathologists.
Hospitals, particularly those departments involving cancer diagnosis