Integrating time-dependent diffusion MRI and intravoxel incoherent motion for predicting NPI and molecular subtypes in breast cancer

To investigate the potential of combining time-dependent diffusion MRI (td-dMRI) with intravoxel incoherent motion (IVIM) to predict the Nottingham Prognostic Index (NPI) and molecular subtypes of breast cancer. A total of 103 breast cancer patients who underwent td-DWI sequences were included. Quantitative diffusion parameters from td-dMRI (ADC0Hz, ADC25Hz, ADC50Hz, Diameter, fin, Dex, and Cellularity) and IVIM model (D, D*, f) were measured and compared using the Mann-Whitney U-test in NPI grades and molecular subtypes. Binary logistic regression analysis was conducted to combine parameters, and the discriminative power of individual and combined models was assessed using Receiver Operating Characteristic curves with Area Under the Curve. The high-grade NPI group exhibited significantly higher D* and lower ADC50Hz, f and Dex values compared to the low-grade group (p < 0.05). The combined model achieved the highest AUC (0.863). The luminal subtype showed increased fin and Cellularity, along with decreased ADC0Hz and ADC25Hz values compared to non-Luminal subtypes (p < 0.05). The D and fin values were significantly different between HER2-enriched subtype and other subtypes (p < 0.05). The ADC0Hz, ADC25Hz, Diameter and Cellularity values differed significantly between triple-negative subtype and other subtypes (p < 0.05). The integration of td-dMRI and IVIM parameters offers a promising noninvasive strategy for preoperative differentiation of NPI grades and molecular subtypes in breast cancer. The combination of td-dMRI and IVIM parameters may enhance the precision of prognostic estimation and the formulation of personalized treatment strategies in clinical practice. • Preoperative prediction of breast cancer NPI and molecular subtypes is crucial. • IVIM and td-dMRI can assist in distinguishing NPI and molecular subtypes. • The integration of IVIM and td-dMRI could facilitate accurate prognostic predictions.