| Part Ⅰ MRN imaging of upper extremity nervesObjectiveTo explore the value of axial T1-weighted imaging (T1WI), T2 weighted imaging (T2WI), fat suppression T2 weighted imaging (FS-T2WI), fat suppression proton density-weighted imaging (FS-PDWI), and diffusion-weighted MR neurograph (DW-MRN) sequence in displaying median nerve, ulnar nerve, and radial nerve.Materials and Methods:1. Participant DataThirty-one healthy participants (age range,22-54 years; 16 men and 15 women; mean age33.1±3.4 years; age range,22-54 women) were enrolled from March 2013 to September 20152. MR ProtocolAll participants were examined 3.0T MRI at the hospital, including axial T1WI, T2WI, FS-T2WI, FS-PDWI, and DW-MRN sequence. For each subject, five segments from each upper arm to palm (called middle arm, elbow, middle forearm, wrist, and proximal palm) were scanned.At middle arm, elbow, middle forearm, the protocol of T1WI, T2WI, FS-T2WI, FS-PDWI, and DW-MRN sequence were as follows:TR (repition time, ms)/TE (echo time, ms) 550/20,3000/80,3000/55,3580/26,9000/85, slice thickness=3.5 mm, slice gap=0.5 mm, Acquisition matrix=72×75, FOV= 160×160 mm2; At wrist, and proximal palm:TR (repition time, ms)/TE (echo time, ms) 550/23,3000/80, 3739/70、3577/27、9000/86, slice thickness=3.5 mm, slice gap=0.5 mm, Acquisition matrix=68×75, FOV= 100×100mm2. DW-MRN was obtained with anterior-posterior direction motion-probing gradients (MPGs); No. of signals acquired=12, Half-Fourier factor=0.795, EPI factor=41, Sensitivity-encoding factor=2, b value=800s/mm2.3. Post-processing of DW-MRN ImagesAnalogous to the post-processing of DW-MRN dataset, full-volume maximum intensity projection (MIP) reconstructions were generated to produce three-dimensional (3D) images and Volume editing tool was manually used to remove superimposed structures that showed similar hyperintensity.4. Image AnalysisThe certainty of identifying the nerves were analyzed by 2 experienced neuroradiologists. The imaging quality of the nerves at the wrist and palm were evaluated using a 4-point grading scale:4, excellent (the entire nerve is visible and of excellent signal intensity); 3, good (the entire nerve is visible and of good signal intensity); 2, moderate (the entire nerve is visible and of moderate signal intensity); and 1, poor (the nerve is partially visible). Images were evaluated for nerve signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and assessed the differences in nerve visualization. The imaging parameters for fives equences were similar to that in the part one.5. Statistical AnalysisFriedman and paired Wilcoxon tests were used to assess the differences between the total 5 sequences on the same subject. Significant differences were defined using the corrected P-values (P<0.05). Interobserver agreement for visualization scores was calculated using the Kappa statistic. Kappa values of 0.61-0.80, good agreement; and 0.81-1.00, excellent agreement.Results:1. The nerve visualization:The long trajectories of the major extremity nerves including the median, ulnar, and radial nerve could be easily visualized on DW-MRN.Peripheral nerves including median nerve, ulnar nerve, and radial nerve were seen in all the subjests (31/31,31/31,31/31) in middle arm, elbow, and middle forearm segments, respectively. Radial nerve could not be identified in the wrist and palm with all MRN sequences. Median nerve and ulnar nerve were seen in all the subjests (31/31, 31/31,31/31) in the wrist. The branches of median nerve were seen in 0/31,2/31, 3/31,8/31,28/31 in the palm. The branches of ulnar nerve were seen in 3/31,5/31, 6/31,7/31,29/31 in the palm. The number of visualized palm branches of median nerve on DW-MRN was higher than those on T1WI、T2WI、FS-T2WI、FS-PDWI sequence (x2 value were 23.70,16.70,18.19,18.50, P<0.05). The number of visualized palm branch of ulnar nerve on DW-MRN were higher than those on T1WI、 T2WI、FS-T2WI、FS-PDWI sequence (x2 value were 20.60,23.31,23.46,23.50, P<0.05).2. Nerve SNR:The SNR of median nerve on T1WI、T2WI、FS-T2WI、FS-PDWI, and DW-MRN sequence at middle arm were 3.67±0.17,4.70±0.15,4.72±0.20,4.74±0.15 and 4.74±0.19; The SNR of median nerve on T1WI、T2WI、FS-T2WI、FS-PDWI, and DW-MRN sequence at elbow were 3.66±0.13,4.70±0.16,4.71±0.19,4.72±0.15 and 4.73±0.17; The SNR of median nerve on T1WI, T2WI、FS-T2WI、FS-PDWI, and DW-MRN sequence at middle forearm were 3.66±0.09,4.73±0.09,4.73±0.14, 4.73±0.16 and 4.73±0.19; The SNR of median nerve on T1WI、T2WI、 FS-T2WI、 FS-PDWI, and DW-MRN sequence at wrist were 3.65±0.04,4.71±0.10,4.71±0.11, 4.71±0.13 and 4.72±0.19. The SNR of median nerve on T2WI、FS-T2WI、FS-PDWI, and DW-MRN sequence at proximal palm were 3.88±0.12,3.88±0.16,3.88±0.19 and 3.89±0.18.The SNR of ulnar nerve on T1WI、T2WI、FS-T2WI、FS-PDWI, and DW-MRN sequence at middle arm were 3.63±0.13,4.69±0.14,4.71±0.19,4.73±0.18 and 4.73±0.19; The SNR of median nerve on T1WI、T2WI、FS-T2WI、FS-PDWI, and DW-MRN sequence at elbow were 3.66±0.11,4.70±0.13,4.71±0.16,4.72±0.11, 4.73±0.16; The SNR of median nerve on T1WI、T2WI、FS-T2WI、FS-PDWI, and DW-MRN sequence at middle forearm were3.66±0.01,4.73±0.02,4.73±0.13, 4.73±0.14,4.73±0.17; The SNR of median nerve on T1WI、T2WI、FS-T2WI、 FS-PDWI, and DW-MRN sequence at wrist were 3.65±0.01,4.71±0.09,4.71±0.06, 4.71±0.10 and 4.72±0.09. The SNR of median nerve on T2WI、FS-T2WI、FS-PDWI, and DW-MRN sequence at proximal palm were 3.88±0.10,3.88±0.13,3.88±0.17 and 3.89±0.12.The SNR of radial nerve on T1WI、T2WI、FS-T2WI、FS-PDWI, and DW-MRN sequence at middle arm were 3.61±0.14,4.68±0.12,4.70±0.17,4.73±0.16 and 4.73±0.12; The SNR of median nerve on T1WI、T2WI、FS-T2WI、FS-PDWI, and DW-MRN sequence at elbow were 3.65±0.18,4.69±0.15,4.70±0.17,4.71±0.18 and 4.72±0.21.3. Nerve CNR:The CNR of median nerve on T1WI、T2WI、FS-T2WI、 FS-PDWI, and DW-MRN sequence at middle arm were 0.59±0.07,0.66±0.08,0.67±0.06,0.68±0.09 and 0.69±0.07; The CNR of median nerve on T1WI、T2W、FS-T2WI、 FS-PDWI, and DW-MRN sequence at elbow were 0.58±0.08,0.65±0.07,0.66±0.05,0.67±0.06 and 0.68±0.08; The CNR of median nerve on T1WI、T2WI、FS-T2WI、FS-PDWI, and DW-MRN sequence at middle forearm were 0.58±0.02,0.67±0.06,0.67±0.09, 0.67±0.11,0.68±0.06; The CNR of median nerve on T1WI、T2WI、FS-T2WI、 FS-PDWI, and DW-MRN sequence at wrist were 0.57±0.06,0.67±0.06,0.67±0.07, 0.67±0.08 and 0.68±0.09. The CNR of median nerve on T2WI、FS-T2WI、FS-PDWI, and DW-MRN sequence at proximal palm were 0.55±0.01,0.55±0.02,0.55±0.04 and 0.55±0.08. The CNR of median nerve on FS-PDWI was higher than FS-T2WI sequence(x2 value were 179.13,173.16, P<0.05)。The CNR of ulnar nerve on T1WI、T2WI、FS-T2WI、FS-PDWI, and DW-MRN sequence at middle arm were 0.58±0.05,0.66±0.03,0.67±0.02,0.68±0.07 and 0.69±0.02; The CNR of median nerve on T1WI、T2WI、FS-T2WI、FS-PDWI, and DW-MRN sequence at elbow were 0.58±0.04,0.65±0.06,0.66±0.01,0.67±0.03 and 0.68±0.06; The CNR of median nerve on T1WI、T2WI、FS-T2WI、FS-PDWI, and DW-MRN sequence at middle forearm were 0.58±0.01,0.67±0.03,0.67±0.07, 0.67±0.10 and 0.68±0.05; The CNR of median nerve on T1WI、T2WI、FS-T2WI、 FS-PDWI, and DW-MRN sequence at wrist were 0.57±0.02,0.67±0.03,0.67±0.05, 0.67±0.07 and 0.68±0.06. The CNR of median nerve on T2WI、FS-T2WI、FS-PDWI, and DW-MRN sequence at proximal palm were 0.54±0.01,0.54±0.03,0.54±0.05 and 0.54±0.07. The CNR of ulnar nerve on FS-PDWI was higher than FS-T2WI sequence(x2 value were 173.11、172.02,P<0.05)。The CNR of radial nerve on T1WI、 T2WI、FS-T2WI、FS-PDWI, and DW-MRN sequence at middle arm were 0.57±0.04,0.65±0.03,0.66±0.08,0.68±0.04 and 0.68±0.05; The CNR of median nerve on T1WI、T2WI、FS-T2WI、 FS-PDWI, and DW-MRN sequence at elbow were 0.57±0.09,0.64±0.07,0.65±0.04,0.66±0.08 and 0.67±0.05.DW-MRN performed better (P<0.01) for nerve SNR and CNR than conventional MR images in the five segments. MRN provided improved SNR and CNR of upper extremity nerves on the middle arm, elbow, middle forearm, and wrist segments as compared to palm segments (P<0.05). No significant difference was found in the SNR and CNR for upper extremity nerves in the conventional MRI on the five segments (P>0.05).Conclusions:1. FS-PDWI rather than FS-T2WI was chosen for comparison to DW-MRN due to superior image quality of the FS-PDWI in the wrist and palm.2. DW-MRN provided improved visualization of extremity nerves branches as compared to conventional MR images in the palm.Part Ⅱ:The clinical value of MRN in the diagnosis of peripheral neuropathy at upper extremityObjectiveTo explore the clinical value of MRN in the diagnosis of peripheral neuropathy at upper extremity compared with ultrasound.Materials and Methods:1. PatientsFrom January 2013 to May 2016, a total of 55 patients (31 men,24 women; age range,3-69 years; mean age 36.3±2.7 years) with peripheral neuropathy at upper extremity were enrolled in the study. The peripheral neuropathy at upper extremity is defined as neuropathy from the upper arm to palm (including median, ulnar and radial nerve) in this study.2. MR Imaging ProtocolAll the patients underwent MRN and Ultrasound examinations. MRN was performed using a 8-channel knee coil and dedicated eight-channel phased array wrist coil. MRN was performed at 3.0 T by using T1WI, T2WI, and FS-T2WI, fat suppression proton density-weighted imaging (FS-PDWI), and diffusion-weighted MR neurography (DW-MRN). The following acquisition parameters were used including wrist and palm:TR/TE/TI 9000/85/260ms, Half-Fourier factor 0.795, EPI factor 41, Sensitivity-encoding factor, b value 800s/mm2, FOV 100x100 mm2, Acquisition matrix 68x75, slice thickness 3.0 mm, slice gap 0, No. of signals acquired 12. DW-MRN was obtained with anterior-posterior direction motion-probing gradients (MPGs). The axial FS-PDWI, axial FS-T2WI and axial T1WI were obtained in similar anatomical positions and the slice thicknesses was as DW-MRN, and the parameters were TR/TE,3577/27ms,3739/70ms and 550/23 ms, respectively. The following acquisition parameters were used from the upper arm to wrist:the imaging parameters for DW-MRN were similar, except for FOV 160×160 mm2, acquisition matrix 72×75, imaging time 227 seconds. The axial FS-PDWI imaging, axial T2-weighted SPAIR imaging and axial TIWI were obtained in similar anatomical positions and slice thicknesses as DW-MRN, and the parameters were TR/TE 3577/27ms,3000/80ms,550/20 ms, slice gap 0.5mm, respectively.3. UltrasoundA GE Vivid7 ultrasound system and 17-5 MHz broadband linear array probe were used for ultrasonographic examination. The full course of the upper extremity nerve in the arm was examined with high-frequency ultrasound. The upper extremity nerve was first located in cross section at arm and then with continuous cross-sectional scanning to observe the full course of the nerve in the arm with high-frequency ultrasound. The number and extent of the lesion were observed.4. Statistical AnalysisSurgery was performed in 55 patients. Compared with surgical findings, the diagnostic efficiency of MRN and ultrasound were evaluated. The accuracy of MRN and ultrasound were compared by using the non-parametric Chi-squared test. Values of P<0.05 were deemed to indicate statistical significance.ResultA total of 55 cases were confirmed by surgical or clinical follow-up findings. The final diagnosis of the fifty-five patients were carpal tunnel syndrome (CTS, n=10), guyon canal syndrome (n=3), cubital tunnel syndrome (n=11), nerve injury (n= 15), non-traumatic nerve fascicle torsion (n=7), schwannoma (n=5), neurofibroma (n=3), and lipofibroma hamartoma (n=1) by surgical findings which were as the reference standard.Surgical pathology or clinical follow-up confirmed 25 cases of entrapment disease, 25 cases diagnosed using by MRN,21 cases diagnosed using by ultrasound. The sensitivity and specificity were 100% and 100% by MRN. The sensitivity and specificity were 87.5% and 99.4% by ultrasound. The ultrasound missed diagnosis in 3 cases.Surgical pathology or clinical follow-up confirmed 15 cases of injury,13 cases diagnosed by MRN,13 cases diagnosed by ultrasound. The sensitivity and specificity were 86.7% and 97.5% by MRN. The sensitivity and specificity were 86.7% and 97.5% by ultrasound. One case was missed by MRN and ultrasound. One case was misdiagnosised by MRN and ultrasound.Surgical pathology or clinical follow-up confirmed 8 cases of tumor,5 cases diagnosed using by MRN,4 cases diagnosed using by ultrasound. The sensitivity and specificity were 80.0% and 97.1% by MRN. The sensitivity and specificity were 83.3% and 98.0% by ultrasound. Three cases were misdiagnosed by MRN. One case was missed and three cases were misdiagnosed by ultrasound.Surgical pathology or clinical follow-up confirmed 7 cases of nerve fascicle torsion, 7 cases diagnosed using by MRN,5 cases diagnosed using by ultrasound. The sensitivity and specificity were 100% and 100% by MRN. The sensitivity and specificity were 83.3% and 97.9% by ultrasound. One case was missed and one case was misdiagnosed by ultrasound.Conclusion1. MRN has higher sensitivity to upper limb of the entrapment disease compared with ultrasound, especially for the deep location of neuropathy and functional disorder of neuropathy with no morphology change. MRN may therefore add diagnostic value as a highly sensitive technique for detection of peripheral neuropathy at upper extremity.2. MRN and ultrasound on upper limb peripheral nerve injury in the early pathological changes of imaging were bad. We believe that MRN become a commonly necessary supplement technique for peripheral neuropathy at upper extremity. |
PDF Full Text Request