Ruizhi Pathology Large Model - A multimodal interactive pathology large model jointly launched by Ruijin and Huawei
RuiPath is a clinical-grade, domestically developed, multimodal, interactive pathology model jointly launched by Ruijin Hospital affiliated with Shanghai Jiao Tong University School of Medicine and Huawei. The model covers 90% of all cancer cases in China each year...
What is the Ruizhi Pathology Large Model?
RuiPath is a clinical-grade, domestically developed, multimodal, interactive pathology model jointly launched by Ruijin Hospital affiliated with Shanghai Jiao Tong University School of Medicine and Huawei. The model covers 90% of common cancers affecting all cancer types annually in China, including rare diseases such as pituitary neuroendocrine tumors. Through multimodal data fusion (images, text, etc.), it encompasses multiple fields including clinical, imaging, and pathology, with broad data dimensions and complete labels, accurately reflecting disease characteristics.
Main functions of Ruizhi Pathology Large Model
- Multimodal data fusionRuiPath integrates multiple data modalities such as images and text, covering multiple fields such as clinical, imaging, and pathology. It has a wide range of data dimensions and complete labels, which can comprehensively reflect the characteristics of diseases.
- Highly efficient auxiliary diagnosisThe model can accurately identify lesion areas in advance, and the AI diagnosis time for a single slice is only a few seconds, which changes the doctor's work mode from the traditional "slice-by-slice diagnosis" to "step-by-step review", significantly improving diagnostic efficiency.
- Interactive diagnostic dialogueDoctors can obtain auxiliary diagnostic suggestions through real-time interaction with the model, further improving the accuracy of diagnosis and work efficiency.
- Extensive disease coverageRuiPath covers 90% of the population affected by common cancers in China each year, including rare diseases such as pituitary neuroendocrine tumors.
- Four major innovationsRuiPath has achieved innovations in scenarios and applications, models and algorithms, in-memory computing collaboration, and AI toolchains, thus promoting the intelligent development of pathological diagnosis.
Technical Principles of Ruizhi Pathological Large Model
- Deep learning and knowledge learningRuiPath, based on deep learning algorithms, learns and optimizes itself from massive amounts of historical pathology data. In a short period, it "studied" over 300 pathology diagnostic books and analyzed 1 million digital slides, achieving expert-level diagnostic capabilities.
- Huawei DCS AI SolutionRuiPath is developed based on Huawei's DCS AI solution, supporting data engineering tools and system-level model training and inference acceleration capabilities, significantly improving the efficiency of model training, inference and fine-tuning.
- Interactive diagnostic modeRuiPath can accurately identify lesion areas in advance, with AI diagnosis of a single slice taking only a few seconds. Doctors can review the AI diagnosis results through real-time interaction with the model, upgrading the traditional "slice-by-slice diagnosis" mode to a "step-by-step review" mode, significantly improving diagnostic efficiency.
Application scenarios of Ruizhi pathology large model
- Pathological diagnostic aidRuiPath can accurately identify lesion areas in pathological slides in advance, with AI diagnosis of a single slide taking only a few seconds. The work mode of pathologists has changed from the traditional "slide-by-slide diagnosis" to "step-by-step review". Doctors only need to review the AI's diagnostic results, improving diagnostic efficiency and quality.
- Clinical diagnostic supportRuiPath covers common cancers affecting 90% of cancer patients in China each year, as well as rare diseases such as pituitary neuroendocrine tumors. Its subspecialty knowledge Q&A has an accuracy rate of over 90%, providing doctors with expert-level diagnostic advice.
- Empowering primary healthcareTo address the issue of insufficient numbers and uneven distribution of pathologists in my country, RuiPath can leverage AI technology to enhance the pathological diagnostic capabilities of primary hospitals and reduce misdiagnosis rates.
- Promotion of Medical AI SolutionsBased on Huawei's DCS AI solution, RuiPath provides other healthcare institutions with scalable smart healthcare solutions.
- Medical Education and TrainingRuiPath can be used for medical education and training to help young doctors quickly improve their pathology diagnostic skills.