Tolga Çukur and colleagues have published “I2I-Mamba: Multi-modal medical image synthesis via selective state space modeling” in IEEE Transactions on Biomedical Engineering (January 2026).
Many diagnoses rely on several types of medical images, such as different MRI contrasts or CT, but not all of them are always acquired. The paper introduces I2I-Mamba, an AI method that synthesizes a missing image type from the available ones, using state space models that capture both local detail and long-range context in the image. Reliable synthesis could reduce the need for extra scans, saving time for patients and clinics.
Authors: Ömer F. Atlı, Bilal Kabaş, Fuat Arslan, Arda C. Demirtas, Mahmut Yurt, Onat Dalmaz, Tolga Çukur