Study population
This study was approved by the institutional review board of Tongji Hospital of Tongji Medical College of Huazhong University of Science and Technology, and written informed consent was obtained from all participants or guardians before enrollment. All methods were carried out in accordance with relevant guidelines and regulations. Fifteen DPN patients (nine males, six females; mean age, 45 years; age range, 26–58 years) that assessed by endocrinologist using the modified diagnostic criteria released by the European Association for the Study of Diabetes (EASD) in 20095 and forty healthy volunteers (eighteen males, twenty two females; mean age, 29 years; age range, 21–55years) underwent MRI(table 1). Written Informed consent was obtained from all subjects. Exclusion criteria were general contraindications for MRI, history of chronic diseases that may affect tibial and common peroneal nerves, history of leg or knee surgery, history of tumor, and systemic vascular disease.
Imaging protocol
Imaging was performed on a 3 T scanner (Skyra, Siemens Healthcare, Erlangen, Germany). All subjects were scanned in a supine position. The knee was placed in a dedicated 15-channel knee coil (Siemens Healthcare) positioned in the scanner’s isocenter.
SMS acquisition with blipped CAIPIRINHA was achieved with dedicated software for research purposes (Siemens Healthcare, Erlangen, Germany). The main parameters are as follows for the axial readout-segmented DTI without SMS: TR 7850 ms; TE 64ms; voxel dimensions, 1.3×1.3×4.0mm3; field of view 252 ×252 mm2; slices 38; generalized autocalibrating partially parallel acquisition 2; bandwidth 885 Hz/Pxz; readout segments, 5; scan time: 8:55min. For the axial SMS-accelerated DTI, the main parameters are as follows: TR, 4000 ms; TE 64 ms; voxel dimensions, 1.3×1.3×4.0mm3; field of view 252 ×252 mm2 ; slices 38; generalized autocalibrating partially parallel acquisition, 2; SMS acceleration factor, 2; bandwidth 885 Hz/Px; readout segments, 5; scan time: 4:54 min. DTI sequences in the study were applied with a b-value of 700 s/mm2 and 12 gradient encoding diffusion directions. After DTI scanning was completed, ADC and FA maps were generated automatically by the machine.
Image quality
Two radiologists with 5 years’ of experience with musculoskeletal imaging scored the image quality of readout-segmented DTI sequences with and without SMS. Scores were assigned for sharpness of nerves, fat-suppression, severity of artifacts, and overall image quality. 5 points denoted the best image quality and 1 point the worst.
Post-processing and measurements
The images were presented for review on a digital workstation (Syngo MultiModality Workplace VE40B, Siemens Healthcare). FA and ADC maps were automatically generated by the scanner (Figure. 1). For FA and ADC measurements, two independent readers placed regions of interest (ROI) on the respective maps in the TN and CPN at the level of the center of knee joint. For patients, ROIs were placed on three levels of most clearly displayed lesions on contiguous slices for ADC and FA measurement, and take the average value as the final measurement result. ROIs were drawn in the area of maximum nerve diameter. Surrounding fat, muscle, and other tissues were avoided. Three selected groups (DTI, ADC map, FA map) of images were presented concurrently with the ROI drawn on the DTI images. The ROI was automatically linked to the ADC and FA maps to measure ADC and FA values. The ROI sizes for TN were 27±6 mm2, and for CPN were15±4 mm2.
Signal-to-noise ratio (SNR) calculation
All DTI sequences with different acceleration factors were acquired twice in 10 volunteers. By subtracting corresponding sequences at b = 0, voxel-based difference images could be generated16,17. On the difference images, the standard deviation of the signal intensity (‘noise’) was measured. The SNR of the TN and CPN at the level of the lower edge of the patella was then calculated for the different sequences as follows:
In addition, the SNR per minute was calculated by dividing the SNR value of each sequence by its acquisition time.
Statistical analysis
Statistical analysis was performed using the SPSS software (SPSS version 25, International Business Machines Corp., Armonk, NY). The inter-reader agreement of the FA and ADC values was assessed using intraclass correlation coefficient (ICC) analysis. An ICC ≤0.20 indicated slight agreement, 0.21–0.40 fair agreement, 0.41–0.60 moderate agreement, 0.61–0.80 substantial agreement, 0.81–1.00 almost perfect agreement18. The paired sample t test was used to statistically analyze the difference between the ADC and FA values of the non-SMS readout-segmented DTI and SMS-accelerated DTI. The independent sample t test was used to assess the differences in ADC and FA values of TN and CPN between the DPN and control groups. Wilcoxon rank sum test was used to compare the difference of image quality in subjective scores between the two DTIs. Receiver operating characteristic curves (ROC) were generated to evaluate the diagnostic efficacy of FA value and ADC values in identifying DPN patients. For all tests, a p-value of <0.05 were considered statistically significant.