Roche Diagnostics and Deep Informatics collaborate to jointly launch a pathology AI-assisted interpretation algorithm

On September 6th, Roche Diagnostics China announced its official partnership with Deep Informatics. They will jointly launch a pathology artificial intelligence (AI) assisted interpretation algorithm product, further promoting the digitalization and intelligent transformation of pathology diagnosis in China, providing strong support for the development of personalized precision diagnosis and treatment.

With the development of precision oncology, pathological diagnosis, known as the 'gold standard' for tumor diagnosis, urgently needs to overcome challenges in efficiency and quality to better meet the demands of individualized precision diagnosis and treatment for both clinical purposes and patients. Artificial intelligence (AI) technology has opened up a new path for improving the quality and efficiency of pathological diagnosis.

Building on the core strengths of its existing pathology diagnostic products, Roche Diagnostics is focusing on creating a comprehensive smart pathology solution by using AI pathology analysis modules as innovative drivers, thereby leading innovation and transformation in pathology diagnostics.

With the successful collaboration between Roche and Deep Informatics, a brand-new integrated solution, covering everything from digital scanning to AI-assisted diagnosis, is officially being implemented under the concept of 'diagnosis and scanning integration.' Leveraging Roche Diagnostics' VENTANA DP 200 and DP 600 digital pathology slide scanners, which provide high-definition, full-field digital images, and Deep Informatics's pathology AI-assisted interpretation algorithms, this integrated solution achieves a seamless fusion of hardware in the pathology laboratory and intelligent algorithms. This will significantly enhance the precision and efficiency of pathology slide interpretation for pathologists.

According to announcement, the collaborative effort includes AI-assisted pathology algorithms for HER2, ER, PR, Ki-67, and PD-L1. These algorithms, which hold significant medical value, are expected to enable intelligent slide reading, adding valuable tools for the pathological diagnosis and precision medication of breast cancer and non-small cell lung cancer. Among these, HER2, as a crucial driver gene in breast cancer, serves as a primary efficacy monitor and prognostic indicator for targeted therapies. In recent years, the introduction of antibody-drug conjugates targeting HER2 has provided new treatment options for patients with low HER2 expression metastatic breast cancer. Simultaneously, the interpretation of low HER2 expression poses new challenges for pathologists.

The AI-assisted pathology algorithms for HER2, ER, PR, Ki-67, and PD-L1 in this collaboration aim to assist pathologists in more accurately, efficiently, and objectively quantifying the molecular subtypes of breast cancer patients. They provide robust support for precise quantitative statistics, particularly in the interpretation of HER2 expression, especially when it is low.

The AI-assisted interpretation algorithm product developed through this collaboration will also be integrated into Roche Diagnostics' Digital Pathology Open Platform, becoming a crucial component of Roche Diagnostics' integrated digital smart pathology solution in China.

Michael Rivers

Vice President of Digital Pathology at Roche Tissue Diagnostics

"Artificial intelligence has tremendous application potential and promising prospects in the field of pathology. AI-assisted interpretation can effectively enhance the consistency and efficiency of pathological diagnoses, especially in cases involving difficult and semi-quantitative interpretations of immunohistochemical biomarkers, where its practical value is particularly significant. With the close collaboration between the two parties, this not only brings new medical insights to pathologists but also provides tangible benefits for precise diagnosis and treatment of patients."

Lin Yang

Founder and Chairman of Deep Informatics

"As a leading provider of digitalized intelligent pathology solutions in China, Diyinga has always been committed to creating efficient and precise diagnostic solutions by combining AI algorithms with pathological diagnosis, providing solid support for the innovative development of the pathology discipline. This collaboration with Roche Diagnostics enables us to seamlessly integrate 'Chinese intelligence' and 'global innovation,' jointly promoting the development of a new ecosystem of intelligent pathology in China."

Richard Yiu

General Manager of Roche Diagnostics China

"This collaboration signifies a significant step for Roche Diagnostics in deepening its commitment 'In China, For China.' Digital innovation is a focal point for Roche Diagnostics and a critical direction for the future development of pathological diagnostics. We believe that through our combined efforts, we will provide a more efficient and precise overall solution for pathological diagnostics, empowering personalized and precise diagnosis and treatment for more Chinese patients. In the future, Roche Diagnostics will continue to embrace an open spirit and collaborate with local innovation to co-create a diagnosis and treatment ecosystem, shaping the future of intelligent pathology in China."​​