Assessment of peripheral neuropathy in type 2 diabetes by diffusion tensor imaging

To evaluate the diagnostic accuracy of diffusion tensor imaging (DTI) in diabetic peripheral neuropathy (DPN) for patients with type 2 diabetes and detect the correlations with electrophysiology.

A total of 27 patients with type 2 diabetes with DPN, 24 patients with type 2 diabetes without peripheral neuropathy (NDPN), as well as 32 healthy controls (HC) were enrolled in this study. Clinical examinations and neurophysiologic tests were used to determine the presence of DPN. Fractional anisotropy (FA) and apparent diffusion coefficient (ADC) of peripheral nerves, including the tibial nerve (TN) and common peroneal nerve (CPN), were calculated. Receiver operating characteristic (ROC) analysis was performed for FA and ADC values. Pearson's correlation coefficient was used to assess the correlation between DTI and electrophysiology parameters in the patient group.

The tibial and common peroneal nerve FAs were lowest (P=0.003, 0.001, respectively) and ADC was highest (P=0.004, 0.005, respectively) in the DPN group. The FA value of the axonal injury group was lower than that in the demyelination group (P=0.035, 0.01, respectively), while the ADC value was higher (P=0.02, 0.01, respectively). In the DPN group, FA value was positively correlated with motor conduction velocity (MCV) (tibial nerve: r=0.420, P=0.007; common peroneal nerve: r=0.581, P<0.001) and motor amplitude (MA) (tibial nerve: r=0.623, P<0.001; common peroneal nerve: r=0.513; P=0.001), while ADC values was negatively correlated with MCV (tibial nerve: r=-0.320, P=0.044; common peroneal nerve: r=-0.569; P<0.001), and MA (tibial nerve: r=-0.491, P=0.001; common peroneal nerve: r=-0.524; P=0.001).

With a lower FA value and higher ADC value, DTI accurately discriminated DPN. The DTI multi-parameter quantitative analysis of peripheral nerves differentiated DPN axonal injury from the demyelinating lesion, and hence, could be applied in the diagnosis of DPN.