Quantitative validation and application of clinical biomarkers in 3T WB-MRI

This work investigates the quantitative validation and clinical application of two MRI-derived biomarkers in 3T Whole-Body MRI (WB-MRI): the Apparent Diffusion Coefficient (ADC), reflecting water mobility and microstructural density, and the Fat Fraction (FF), representing tissue fat content. The objectives were to assess their robustness, define reference values in healthy tissues and evaluate their diagnostic potential in Multiple Myeloma (MM). Technical validation was performed using a water phantom for ADC accuracy and a custom-built fat–water phantom for FF quantification. In vivo characterisation in 20 healthy females across seven anatomical sites — parenchymal organs (liver, spleen, pancreas, kidneys) and skeletal regions (vertebrae, pelvic bone, femoral head) — examined repeatability, variability and radiomic feature stability. ADC showed greater consistency in parenchymal organs, whereas FF proved more stable in adipose-rich skeletal tissues. The clinical study analysed 107 MM patients (38 Smoldering, the early asymptomatic stage; 46 active; 23 Relapsed/Refractory) against the 20 healthy controls. In the pelvic bone the replacement of normal marrow by malignant plasma cells was associated with increased ADC and decreased FF. FF consistently showed higher discriminative power in distinguishing healthy from infiltrated marrow even at early stages, while ADC reached high significance mainly in advanced disease. Radiomic descriptors further enhanced classification performance, with FF-based models achieving Area Under the Curve (AUC) values above 0.95 and ADC providing complementary information in multiparametric models. The results demonstrate that ADC and FF can be robustly quantified in WB-MRI. FF emerges as the more stable and diagnostically sensitive biomarker for bone marrow infiltration in MM. The proposed framework supports the integration of quantitative biomarkers into non-invasive tools for diagnosis, staging and longitudinal monitoring of MM patients.