Histofy, a company specialising in developing transparent AI for pathology, today announces a project to develop an AI-powered system that provides precise quantification and profiling of cancer cells in growth overdrive, in order to enhance patient care.
Birmingham, UK (20th March 2023) – Histofy, a UK-based startup developing transparent AI solutions for digital pathology, today announces an industry-academic collaboration with CNAI, a leading AI synthetic data company in Korea, and Korea University to develop MitProfiler, a cloud-based AI solution that profiles mitotic figures in digitised images of cancerous tissue slides. The collaboration is funded by Innovate UK and KAIT under the UK-South Korea Collaborative R&D scheme, which aims to fund collaborative research and development (CR&D) projects focused on industrial research.
Automating Mitosis Detection
Powered by advanced award-winning algorithms, MitProfiler will be able to detect mitotic figures from digitised images of tissue slides and quantify tumour proliferation across entire tissue samples, allowing pathologists to assess tumour proliferation more quickly and accurately. Pathologists typically detect mitotic figures by manually counting mitoses in tumour regions, an inherently subjective and time-consuming task that is prone to observer variability.
"The MitProfiler product is aimed at enhancing workload efficiency of pathologists"
The project will allow rapid adoption of AI-based detection of mitotic figures, helping to drastically reduce the time required for counting. Despite recent developments in computational pathology, there is currently no device on the market that is fully operable in various digital pathology systems across a range of tumour sites.
Histofy leads the project with data collection and refinement of their award-winning algorithm, while CNAI will design and develop the cloud architecture. Histofy will also collaborate with Korea University to research and develop domain-invariant methods for mitotic figure detection, allowing the solution to perform well across institutions with different scanner manufacturers and tissue preparation protocols.
Dr Simon Graham, CTO and co-founder of Histofy said: “Histofy is a team of world leading experts in computational pathology, and we are ready to make AI accessible across a range of solutions for the digital pathology market. The MitProfiler product is aimed at enhancing workload efficiency of pathologists, leading to better patient management.”
Wonseop Lee, CEO and founder of CNAI, said: “We are happy to work with Histofy and Korea University in developing MitProfiler for better patient care. This technology has the potential to revolutionise the field of pathology by improving the accuracy and efficiency of tumour analysis. Our extensive MLOps and biomedical research experience will help us to bring this technology to the market and make an impact that matters.”
Professor Jin Tae Kwak of Korea University said: “We are thrilled to partner with Histofy and CNAI”. This international collaboration will provide an opportunity to create an innovative AI device that can impact patients suffering from serious diseases.”
About CNAI
CNAI is Korea’s first AI synthetic data company with extensive research and business experience. Its expertise in synthetic data for AI vision systems helps large corporations develop AI applications in industries including retail, manufacturing, and healthcare.
About Korea University
Korea University is one of the top institutions of higher education in Korea. The Quantitative Imaging and Informatics Laboratory (QuIIL) at Korea University, led by professor Jin Tae Kwak, is a major contributor of this consortium, which is a research team focussing on the development of advanced AI techniques for medical image analysis.
