The three futures of bone marrow imaging: a radiologist’s guide

Abstract Evaluating normal and pathological bone marrow (BM) remains a challenging task for radiologists. Magnetic resonance imaging (MRI) has demonstrated to be the most effective modality for BM assessment, offering superior accuracy over conventional X-ray and computed tomography (CT). However, the traditional morphological MRI approach, relying on T1-weighted and other sequences, may not be sufficient in fully characterizing BM pathology or providing predictive and monitoring capabilities in an era of quantitative and personalized medicine. This review categorizes BM imaging modalities into 3 conceptual stages: past-future, present-future, and future-future. The “past-future” includes well-established but underutilized MRI sequences in the field of BM assessment, such as DIXON, DWI, and PWI. The “present-future” encompasses recent advances such as AI integration for image acquisition and emerging CT modalities like dual energy CT (DECT). Finally, the “future-future” anticipates transformative technologies including photon-counting CT (PCCT), ultra-high field MRI (5T and 7T), MRI fingerprinting, radiomics, and advanced AI applications. Each stage highlights the evolving toolkit available to radiologists for accurate BM characterization, early detection, and effective monitoring. Embracing these innovations will empower radiologists to provide more precise, patient-specific care, shaping the future of BM imaging.