Integrated assessment of diabetic nephropathy by multi-parametric magnetic resonance imaging with non-enhanced T1 mapping, diffusion tensor imaging, and diffusion kurtosis imaging.

BACKGROUND

This study investigated the diagnostic utility of multi-parametric magnetic resonance imaging (mp-MRI) for early diabetic kidney disease (DKD). We aimed to assess renal impairment by combining parameters from non-enhanced T1 mapping, diffusion tensor imaging (DTI), and diffusion kurtosis imaging (DKI).

METHODS

We prospectively recruited 63 patients with type 2 diabetes, who were divided into three groups based on urine albumin-creatinine ratio: Group 0 (simple diabetes, n = 11), Group 1 (early DKD, n = 20), and Group 2 (advanced DKD, n = 32). All patients underwent renal mp-MRI. Parameters were correlated with clinical indicators (e.g., eGFR, Scr) and, in 11 patients, with histopathology scores from renal biopsy. Receiver operating characteristic (ROC) analysis was used to evaluate diagnostic performance.

RESULTS

MRI parameters correlated significantly with clinical indicators and key pathology scores. For distinguishing Group 0 from Group 1, mean diffusivity (MD) from DTI (AUC = 0.948) and mean kurtosis (MK) from DKI (AUC = 0.945) were the most effective single parameters. For distinguishing Group 1 from Group 2, MK (AUC = 0.934) and T1 (AUC = 0.878) showed the highest performance. A combined model integrating T1, DTI, and DKI to distinguish Group 0 from Group 1 achieved an AUC of 0.982, with 95.0 % sensitivity and 81.8 % specificity.

CONCLUSION

The combined use of T1 mapping, DTI, and DKI provides a non-invasive method for diagnosing early DKD that outperforms any single MRI parameter.