Mycosis is an extremely common disease. According to epidemiological study statistics, there are about more than 150 million people with superficial fungal infections in China, and about 2.5 million new deep fungal infections occur each year. Medical fungal infection test methods include: Pathogen culture, direct microscopic examination, histopathology, serology and molecular detection and other methods. Among them, pathogen culture, direct microscopic examination and histopathology are based on fungal morphological differences for diagnosis, and are also the most basic clinical detection methods and diagnostic gold standard.
Artificial fungal morphological detection and diagnosis takes a lot of time for inspectors and microbial pathologists. The average time for each fungal morphological detection and diagnosis of a fungal inspector with more than 5 years of work experience is 30-50 minutes. The clinical detection and diagnosis efficiency is low, the reporting cycle is long, and it is difficult to meet the clinical needs.With the vigorous advancement of the national tiered diagnosis and treatment system, the transformation of medical AI applications has developed rapidly and diversified in application scenarios such as medical imaging, pathological diagnosis and drug development. However, the accumulation of data in the field of medical fungal identification is difficult and professional, and it is still in the blank stage. Based on a large number of existing medical fungal image information and close clinical cooperation, this team aims to develop a medical fungal artificial intelligence identification project to achieve rapid screening and identification of fungus-related pictures. In addition, the auxiliary diagnostic system will be combined with the relevant hardware (microscope, clean operating table, incubator, etc.) in the process of fungal diagnosis to realize the dynamic identification during fungal culture while reducing the manual operation process in the process of fungal imaging diagnosis, and provide a complete set of laboratory detection solutions for systemic fungal pathogens for primary hospitals.
According to the current situation of medical fungal detection and diagnosis in China, this project is oriented to major clinical application needs, combines medical fungal disciplines and modern high-tech integration of artificial intelligence technology for scientific and technological innovation, develops medical fungal artificial intelligence auxiliary diagnosis system from three major directions: software, hardware and platform, and provides an effective tool for improving the medical fungal detection and diagnosis ability of medical institutions at all levels in China (especially grassroots medical institutions) as a whole. The contents of project construction include: Medical fungal AI identification system, intelligent fungal diagnosis system, fungal disease collaborative cloud platform. The model of "medical fungal AI identification system + intelligent fungal diagnosis system + fungal disease cooperative cloud platform" is an integrated solution for nosocomial fungal infection, realizing social value and forming a "closed-loop" business model of medical fungi, which is of great economic value.
