Auxiliary cancer diagnosis method based on digital pathological images

Abstract

The invention discloses an auxiliary cancer diagnosis method based on digital pathological images. The method comprises the following steps that 1, a section scanner is used for scanning digital pathological sections into images; 2, a digital pathological image database and a digital pathological image label library are built; 3, a digital pathological image automatic diagnosis model and a digital pathological image retrieval model are built; 4, the digital pathological image automatic diagnosis model is adopted for diagnosing an unknown lesion type and digital pathological sections, a diagnosis result is obtained, and meanwhile a plurality of lesion areas are selected; 5, the digital pathological image retrieval model are adopted for searching for similar lesion areas; 6, the diagnosis result, the typical lesion areas and other information are fed back to a doctor for auxiliary diagnosis. The digital pathological image automatic diagnosis model and the digital pathological image retrieval model are combined, previously-diagnosed similar cancer cases are provided for a doctor as references while a computer is used for predicting the lesion type of the digital pathological sections, and the purposes of improving the cancer diagnosis quality and improving the diagnosis capacity of the doctor are achieved.

Description

technical field [0001] The invention belongs to the field of digital image processing, and relates to the technology of pathological image feature extraction, classification and content-based image retrieval, and in particular to a cancer auxiliary diagnosis method based on digital pathological images. Background technique [0002] Digital pathological full slides refer to the use of slice scanning imaging systems to convert pathological slices into high-magnification large-scale digital images that can be displayed, transmitted, and processed by computers. Taking advantage of the characteristics of easy storage, transmission and online viewing of digital pathological slides, the remote diagnosis system of cancer based on pathological images has been developed rapidly. Relying on this cancer remote diagnosis technology, some county-level hospitals in remote areas of the country have already met the conditions for cancer diagnosis. However, according to the statistics of the...

AI Technical Summary

Problems solved by technology

However, according to the statistics of the current remote diagnosis platform, the rate of agreement between the diagnosis opinions of grassroots doctors and pathology experts is less than 60%.

However, such algorithms often can only give diagnostic results, but cannot provide diagnostic evidence to support the diagnostic results.

Images