Diffusion-weighted imaging of the spinal cord

Longitudinal measurements of syrinx size in a rat model of posttraumatic syringomyelia

OBJECTIVE Syringomyelia pathophysiology is commonly studied using rodent models. However, in vivo studies of posttraumatic syringomyelia have been limited by the size of animals and lack of reliable noninvasive evaluation techniques. Imaging the rat spinal cord is particularly challenging because the spinal cord diameter is approximately 1-3 mm, and pathological lesions within the spinal cord parenchyma are even smaller. The standard technique has been histological evaluation, but this has its limitations. The aim of the present study was to determine whether syrinx size could be reliably measured using a preclinical high-field MRI animal system in a rat model of posttraumatic syringomyelia. METHODS The authors used an existing rat model of posttraumatic syringomyelia, which was created using a controlled pneumatic compression device to produce the initial spinal cord injury, followed by a subarachnoid injection of kaolin to produce arachnoiditis. T2-weighted MRI was performed on each animal using a 9.4-T scanner at 7, 10, and 13 weeks after injury. Animals were killed and syrinx sizes were calculated from in vivo MRI and histological studies. RESULTS MRI measurements of syrinx volume and length were closely correlated to histological measurements across all time points (Pearson product moment correlation coefficient r = ± 0.93 and 0.79, respectively). CONCLUSIONS This study demonstrates that high-field T2-weighted MRI can be used to measure syrinx size, and data correlate well with syrinx size measured using histological methods. Preclinical MRI may be a valuable noninvasive technique for tracking syrinx formation and enlargement in animal models of syringomyelia.

Multishot diffusion-weighted MR imaging features in acute trauma of spinal cord

OBJECTIVES

To analyse diffusion-weighted MRI of acute spinal cord trauma and evaluate its diagnostic value.

METHODS

Conventional MRI and multishot, navigator-corrected DWI were performed in 20 patients with acute spinal cord trauma using 1.5-T MR within 72 h after the onset of trauma.

RESULTS

Twenty cases were classified into four categories according to the characteristics of DWI: (1) Oedema type: ten cases presented with variable hyperintense areas within the spinal cord. There were significant differences in the apparent diffusion coefficients (ADCs) between lesions and unaffected regions (t = -7.621, P < 0.01). ADC values of lesions were markedly lower than those of normal areas. (2) Mixed type: six cases showed heterogeneously hyperintense areas due to a mixture of haemorrhage and oedema. (3) Haemorrhage type: two cases showed lesions as marked hypointensity due to intramedullary haemorrhage. (4) Compressed type (by epidural haemorrhage): one of the two cases showed an area of mild hyperintensity in the markedly compressed cord due to epidural haematoma.

CONCLUSIONS

Muti-shot DWI of the spinal cord can help visualise and evaluate the injured spinal cord in the early stage, especially in distinguishing the cytotoxic oedema from vasogenic oedema. It can assist in detecting intramedullary haemorrhage and may have a potential role in the evaluation of compressed spinal cord.

KEY POINTS

• Multishot, navigator-corrected diffusion-weighted MRI is helpful when evaluating spinal cord injury (SCI). • Four types of SCI may be classified according to the DWI characteristics. • DWI differentiates cytotoxic from vasogenic oedema, thereby determining the centre of SCI. • DWI can assist in detecting intramedullary haemorrhage. • DWI can help evaluate the degree of compressed spinal cord.

MR Diffusion Anisotropy Imaging of Spinal Cord Mechanical Compression and Injury on the Rat Model. In Vivo

The aim of this study was the application of magnetic resonance diffusion anisotropy imaging (MR DAI) for assessment of spinal cord compression (SCC) and injury (SCI) in rats depending on the time course from the moment of injury. Twenty rats were used, divided into three groups (five with no surgical operation, five with laminectomy only and ten with SCI produced using a dynamic weight - drop model). MR DAI was performed four times (1, 24, 48 and 168 hours after surgery) at 4.7 T with diffusion gradients applied parallel and perpendicular to the spine. Diffusion parameters (lADC, tADC and AI) were calculated for defined regions of white and gray matter. Epidural hematoma which appeared after laminectomy compressed spinal cord and caused a decrease of apparent diffusion coefficient (ADC) values in GM and WM. SCI in WM produced a decrease of lADC and increase in tADC. In GM an increase in both lADC and tADC values after SCI was observed. MR DAI will disclose dynamic changes in water diffusion during the first days after spinal cord contusion.

Diffusion-weighted MR imaging (DWI) in the evaluation of epidural spinal lesions

INTRODUCTION

Epidural spinal cord compression is one of the most critical emergency conditions requiring medical attention and requires prompt and adequate treatment. The aim of our study was to assess the role of diffusion-weighted magnetic resonance (MR) imaging (DWI) in the diagnosis and differentiation of epidural spinal lesions.

METHODS

Three patients with epidural lymphoma, two with sarcoma and three with epidural metastatic disease were imaged on a 1.5T MRI unit. DWI was performed using navigated, interleaved, multi-shot echo planar imaging (IEPI). Three region of interest (ROI)-measurements were obtained on corresponding apparent diffusion coefficient (ADC) maps, and the mean ADC value was used for further analysis. The cellularity of tumors was determined as the N/C ratio (nucleus/cytoplasma ratio) from histological samples. The ADC values and N/C ratios of lesions were compared using a Kruskal-Wallis test.

RESULTS

The mean ADC of the lymphomas was 0.66 x 10(-3) mm2/s, that of the sarcomas was 0.85 x 10(-3) mm2/s and the ADC of the metastatic lesions was 1.05 x 10(-3) mm2/s; however, the differences were not statistically significant. Mean N/C ratios in the lymphoma, sarcomas and metastases were 4:1, 2:1, and 2.6:1, respectively, with a statistically significant difference between the groups (p < 0.025).

CONCLUSION

Although not statistically significant due to the small patient sample, our results clearly show a tendency toward decreased diffusivity in neoplastic lesions with higher cellularity. The data from our study suggest that DWI is a feasible and potentially useful technique for the evaluation of epidural lesions that cause spinal cord compression on a per-patient basis.

ADC measurements at low and high b values: insight into normal brain structure with clinical DWI

PURPOSE

To demonstrate drop in brain ADC measurements from low to high b values; to evaluate the structural information provided based on those changes; and to discuss the anatomical reasons for ADC differences.

METHODS

Four cerebral ROI (precuneus-PRC, hippocampus-HIP, and the genu-GCC and splenium-SCC of the corpus callosum-CC) were drawn for ADC measurements with low (1000) and high (3000) b-value DWI in 50 normal subjects. ANOVA and Bonferroni correction tested ADC differences between areas, between both hemispheres, between GCC and SCC, and between b-value related ADC drop within areas. Pearson test evaluated dependence of interhemispheric and intercallosum ADC measurements obtained with the same b-value, dependence between areas of intrazonal drop, and the interhemispheric and intercallosum dependence of intrazonal drop.

RESULTS

ADCs differed between areas (P<.0001). Interhemispheric ADC only differed in PRC with low b-value (P<.027). No HIP asymmetries occurred regardless the b-value. ADC drop within PRC and HIP was similar but differed (P<.0001) from ADC drop within both CC ROI. ADC drop was also different between GCC and SCC (P<.0001). In PRC and HIP, ADC showed a significant interhemispheric and intrazonal dependence (P<.0001). There was no GCC to SCC ADC dependence. Intrazonal dependence in the CC was only significant in the SCC (P<.001). Interhemispheric dependence of intrazonal drop was significant (PRC P=.007; HIP P<.0001) but failed to reach significance in the CC.

CONCLUSION

Low and high b-value measurements show different diffusion behaviours within different tissues, especially in a highly anisotropic structure as the corpus callosum. This fact can provide valuable information about brain structure and different diffusion compartments in clinical DWI.

Diffusion-weighted MR imaging (DWI) in spinal cord ischemia

INTRODUCTION

Spinal cord infarction is a rare clinical diagnosis characterized by a sudden onset of paralysis, bowel and bladder dysfunction, and loss of pain and temperature perception, with preservation of proprioception and vibration sense. Magnetic resonance imaging (MRI) usually demonstrates intramedullary hyperintensity on T2-weighted MR images with cord enlargement. However, in approximately 45% of patients, MR shows no abnormality. Diffusion-weighted MR imaging (DWI) has been widely used for the evaluation of a variety of brain disorders, especially for acute stroke. Preliminary data suggest that DWI has the potential to be useful in the early detection of spinal infarction.

METHODS

We performed DWI, using navigated, interleaved, multishot echo planar imaging (IEPI), in a series of six patients with a clinical suspicion of acute spinal cord ischemia.

RESULTS

In all patients, high signal was observed on isotropic DWI images with low ADC values (0.23 and 0.86x10(-3) cm(2)/s), indicative of restricted diffusion.

CONCLUSION

We analyzed the imaging findings from conventional MR sequences and diffusion-weighted MR sequences in six patients with spinal cord infarction, compared the findings with those in published series, and discuss the value of DWI in spinal cord ischemia based on current experience. Although the number of patients with described DWI findings totals only 23, the results of previously published studies and those of our study suggest that DWI has the potential to be a useful and feasible technique for the detection of spinal infarction.

Value of diffusion-weighted MR imaging in acute cervical cord injury as a predictor of outcome

INTRODUCTION

We analyzed the findings of diffusion-weighted (DW) imaging using the single-shot fast spin-echo sequence in acute cervical cord injury and evaluated the usefulness of this method for predicting the prognosis.

METHODS

Our patient group comprised 14 patients examined 2 h to 3 days after injury. First, we visually evaluated the DW imaging findings in all patients. Apparent diffusion coefficient (ADC) maps were also assessed in 13 patients. Second, we assessed follow-up magnetic resonance (MR) examinations obtained in six patients whose DW images showed hyperintensity. Third, we reviewed the functional outcome at discharge.

RESULTS

The lesions showed hyperintensity in ten patients, and no abnormal signal was noted in the remaining four patients. The ADC maps showed restricted diffusion in all patients with hyperintensity on DW imaging except in one patient for whom the ADC map was unavailable. Repeated MR examinations obtained in six of the ten patients showed either myelomalacia or exacerbation. Seven of the ten patients (70%) required assistance and the other three were independent. Among the four patients without hyperintensity on DW imaging, three (75%) were independent and only one required assistance.

CONCLUSION

DW imaging in acute cervical cord injury often reveals restricted diffusion. This finding may predict an unfavorable functional prognosis.

In vivo serial diffusion tensor imaging of experimental spinal cord injury

In vivo longitudinal diffusion tensor imaging (DTI) of rodent spinal cord injury (SCI) was carried out over a period of eight weeks post-injury. A balanced, rotationally invariant, alternating gradient polarity icosahedral diffusion encoding scheme was used for an unbiased estimation of the DTI metrics. The fractional anisotropy (FA), diffusivities along (longitudinal), and perpendicular (transverse) to the fiber tracts, were estimated for the ventral, dorsal, and lateral white matter. In all the three regions, the DTI metrics were observed to be significantly different in injured cords relative to the uninjured controls close to the epicenter of the injury. However, these differences gradually disappeared away from the epicenter. The spatio-temporal changes in the DTI metrics showed a recovery pattern that is region specific. Although the temporal trends in the tissue recovery in rostral and caudal sections seem to be similar, overall the DTI metrics were observed to be closer to the normal tissue values in the caudal relative to the rostral sections (rostral-caudal asymmetry).