| abstract |
The invention discloses a kidney cancer pathological image analysis method, system and intelligent microscope device based on CoAtNet deep learning, including kidney cancer pathological data collection, image slicing, background filtering, data enhancement, kidney cancer region detection, and kidney cancer subtype classification , renal cancer grading and renal cancer prognosis analysis. Through the combination of artificial intelligence technology and microscope, the present invention can be used to assist pathologists to complete end-to-end renal cancer diagnosis, and the analysis model of renal cancer pathological images can meet the purposes of high recognition rate, strong real-time performance and complete functions at the same time. It helps to alleviate the shortage of talents in pathology department in my country and the long training cycle; it is beneficial to reduce the probability of misjudgment and fatigue of pathologists. |