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

[OBJECTIVES] 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.

[METHODS] A total of 103 breast cancer patients who underwent td-DWI sequences were included. Quantitative diffusion parameters from td-dMRI (ADC, ADC, ADC, Diameter, f, D, 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.

[RESULTS] The high-grade NPI group exhibited significantly higher D* and lower ADC, f and D values compared to the low-grade group ( < 0.05). The combined model achieved the highest AUC (0.863). The luminal subtype showed increased f and Cellularity, along with decreased ADC and ADC values compared to non-Luminal subtypes ( < 0.05). The D and f values were significantly different between HER2-enriched subtype and other subtypes ( < 0.05). The ADC, ADC, Diameter and Cellularity values differed significantly between triple-negative subtype and other subtypes ( < 0.05).

[CONCLUSION] 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.

[CRITICAL RELEVANCE STATEMENT] 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.

[SUPPLEMENTARY INFORMATION] The online version contains supplementary material available at 10.1186/s12880-025-02128-8.