Our research combines artificial intelligence, fingerprint analysis, and genetics to develop accessible tools for the early detection of rare developmental disorders such as Kabuki syndrome and Wiedemann-Steiner syndrome. Using deep learning models, including convolutional neural networks and vision transformers, we have demonstrated that subtle fingerprint patterns can distinguish affected individuals from unaffected controls with strong diagnostic performance. Our studies identified previously unknown syndrome-specific fingerprint features and showed that AI can detect clinically meaningful signals invisible to the human eye. By transforming fingerprints into digital biomarkers, we aim to create fast, noninvasive, and globally scalable diagnostic tools that expand access to early genetic screening and support earlier intervention.
Publications:
Marilyn Lionts, Arnhildur Tomasdottir, Viktor I. Agustsson, Yuankai Huo, Hans T. Bjornsson, Lotta M. Ellingsen, “An interpretable vision transformer as a fingerprint-based diagnostic aid for Kabuki and Wiedemann-Steiner syndromes,” Proc. SPIE 13930, Medical Imaging 2026: Imaging Informatics, 139301W;
Viktor Ingi Agustsson, Pall Asgeir Bjornsson, Ashildur Fridriksdottir, Hans Tomas Bjornsson, Lotta Maria Ellingsen, Automated fingerprint analysis as a diagnostic tool for the genetic disorder Kabuki syndrome, Genetics in Medicine Open, Volume 2, 2024, 101884, ISSN 2949-7744,
MYNSTUR Diagnostics 