Highlight B (HB)

Applications of AI in healthcare

FCAI Highlight Program B creates AI tools to tackle real-world problems in healthcare together with expert collaborators from the respective fields.

Application B1: AI for genetics (contact: Samuli Ripatti, University of Helsinki)

We will create AI to analyze multivariate but structured genotype and phenotype data. The FinnGen project combines genetic data and electronic health records for 500,000 Finns. In collaboration with FinnGen we apply the AI tools to find genes modifying disease risk, progression, and comorbidities.

Application B2: Computational vaccines  (contact: Jukka Corander, University of Helsinki) 

We will develop an AI-driven R&D tool for digital engineering of bacterial vaccines, which uses population genomic surveillance data combined with experiments to make probabilistic predictions of the campaign effects for candidate vaccines and to identify optimal formulations. The tool will significantly accelerate development of new vaccines and has large implications for global human health.

Application B3: Healthcare resource allocation (contact: Pekka Marttinen, Aalto University)

We will create AI for prediction of healthcare services and train it on nation-wide healthcare register data, in collaboration with the National Institute for Health and Welfare (THL). The platform can predict healthcare costs of individuals and will be used to allocate resources to healthcare providers in a fair and efficient way. It will also be used to assess and compare treatment practices across the country to identify the most effective ones.

Examples of publications:

• Tianyu Cui, Yogesh Kumar, Pekka Marttinen, Samuel Kaski. 2022. Deconfounded Representation Similarity for Comparison of Neural Networks. Thirty-sixth Conference on Neural Information Processing Systems (NeurIPS 2022).

• Severi Rissanen and Pekka Marttinen. 2021. A Critical Look at the Consistency of Causal Estimation with Deep Latent Variable Models. Thirty-fifth Conference on Neural Information Processing Systems (NeurIPS 2021), pp. 4207-4217.

• Caroline Colijn, Jukka Corander, Nicholas J. Croucher. 2020. Designing ecologically optimized pneumococcal vaccines using population genomics. Nature Microbiology, 5.

• Yogesh Kumar, Salo Henri, Tuomo Nieminen, Kristian Vepsäläinen, Sangita Kulathinal, Pekka Marttinen. 2020. Predicting utilization of healthcare services from individual disease trajectories using RNNs with multi-headed attention. Proceedings of Machine Learning Research, 116.

Program poster (pdf)

Coordinating professor: Pekka Marttinen – pekka.marttinen at aalto.fi

People

The groups of following PIs take part in this Highlight. If you would like to join the Highlight, please contact the coordinating professor.

• Jukka Corander, University of Helsinki, University of Oslo

• Andrea Ganna, genetics, register-based research, machine learning, FIMM, University of Helsinki

• Antti Honkela, machine learning, privacy, University of Helsinki

• Unto Häkkinen, health economics, National Institute for Health and Welfare

• Alexander Ilin, deep learning, Aalto University

• Giulio Jacucci, InfoVis, University of Helsinki

• Kimmo Kaski, deep learning in imaging, social networks, Aalto University

• Samuel Kaski, user interaction, Aalto University

• Sangita Kulathinal, register-based research, statistics, University of Helsinki

• Pekka Marttinen, machine learning, Aalto University – coordinating professor

• Samuli Ripatti, FIMM, University of Helsinki

• Karoliina Snell, social sciences, University of Helsinki

• Petri Ylikoski, social sciences, University of Helsinki