The IVF-AI project led by HUS Helsinki University Hospital started in November 2024. This two-year collaborative initiative with HPE and CGI aims to revolutionize fertility treatments by utilizing artificial intelligence. The project brings together medicine, artificial intelligence and clinical research with the goal of improving the success rate and cost-effectiveness of fertility treatments.
In the IVF project (in-vitro fertilisation), research and development collaboration is conducted between HUS, HPE (Hewlett Packard Enterprise) and CGI.
HUS’s head of reproductive medicine, Hanna Savolainen-Peltonen, serves as the clinical expert for the project and discusses the growing global need for fertility treatments. Globally, about 15 percent of people wishing to have a child suffer from infertility, making this project highly significant. “From the patient’s perspective, it is crucial to receive realistic information about how likely the treatment is to succeed,” Hanna Savolainen-Peltonen says. The project seeks to find more personalised solutions for treatment selection and predictability.
HPE is leading the AI and Data part of the project by developing comprehensive predictive models using patient and image data. This model aims to predict the general success probability of assisted reproductive treatments and recommend the best line of treatment for each couple through AI-enabled precision medicine. Furthermore, HPE is working on integrating the outputs from the embryo image recognition model with the patient data model to provide a comprehensive solution. “Traditionally, embryologists have manually performed such analyses, but AI can significantly speed up their work by offering a more systematic interpretation of images and numerical data. This integration into the predictive model represents a significant advancement in clinical practice, providing a more efficient and accurate approach to fertility treatments,” states Iveta Lohovska from HPE.
CGI contributes artificial intelligence expertise to the project, necessary for developing medical imaging recognition methods using AI. CGI’s Miikka Kiiski notes: “The aim is to create a technically viable solution that can help couples worldwide.” Artificial intelligence can be used to analyse previous images of embryos and cells if fertility treatments have already been performed
The project is planned to progress through three phases: feasibility, proof-of-value and commercialisation. Soile Hollmen from HPE states that the feasibility phase focuses on assessing the selected concept and determine the development work of the project. A research environment has already been established for the project at HUS Acamedic, where secure data handling is ensured. She adds that the extensive data produced by HUS has already been utilised, and the hope is to discover new insights into embryo analytics.
The ultimate goal of the project is a concrete, commercial tool that supports decision making regarding fertility treatments by providing numerical predictions and assistance in embryo selection. “The aim of this project is for couples to conceive more quickly and for clinicians to save time,” Hanna Savolainen-Peltonen says. Additionally, the project seeks to gain new clinical understanding of embryos and their potential biomarkers.
HUS, HPE and CGI are proud of their collaboration, which is expected to yield significant research results to help couples succeed in infertility treatments. In addition to the practical tool, they will produce scientific publications. The project will also provide important quality assurance for the algorithms.
Key people in the IVF-AI project:
Hanna Savolainen-Peltonen, HUS
Soile Hollmen, HPE
Miikka Kiiski, CGI
Tiina Puurtinen, CHN.
