Application of diffusion tensor imaging in precise assessment of cervical myelopathy

Abstract

Cervical myelopathy (CM) is the most common cause of spinal dysfunction among the elderly, and it results in considerable disability and health-care costs. CM arises from static and dynamic compression of the cervical spinal cord caused by vertebral and disc degeneration. Surgical decompression is considered as the mainstay of the treatment for CM patients. However, the decision making for surgery is still empirical and the functional outcome is not always satisfactory, even after successful surgery. Hence, there is pressing need for a reliable and precise approach to evaluate the severity of myelopathy and indicate the optimal timing for surgical treatment. Diffusion tensor imaging (DTI) is a promising imaging technique that is capable of detecting the microstructural integrity of the spinal cord. In this project, we used DTI to detect the severity of demyelination in myelopathic spinal cord and to investigate its potential in the diagnosis and prognosis of the CM patients.

In order to define the abnormality of DTI in CM patients, the DTI metrics of healthy subjects were used to establish a database of DTI values of healthy cervical spinal cord. Differences in DTI metrics among different cord levels were also observed. The effects of gender and aging in DTI analysis were investigated, and their minimal influence was found to be negligible in the evaluation of cervical spinal cord disorders.

Level diagnosis is essential for deciding surgical plans for CM patients. However, multilevel CM presents complex neurological signs making level localization difficult. We examined the ability of DTI analysis in level diagnosis of multilevel CM. Orientation entropy-based DTI analysis was found to have higher accuracy and sensitivity than the current anatomic MRI method, and it can be a useful tool to determine the pathological spinal cord level in CM patients.

Meanwhile, we adopted column-specific DTI analysis to detect detailed microstructural impairment. DTI metrics of CM patients were found to significantly differ from those of healthy subjects. Significant correlation between the column-specific FA value and clinical assessment was observed.

Furthermore, the prognostic value of column-specific DTI metrics was investigated. Column-specific DTI analysis presented better prognostic capacity than clinical features and conventional MRI. Preoperative FA values in the dorsal column had the strongest correlation with the recovery rate of CM patients and may contribute to the decision making of surgical treatment.

Finally, we analyzed the postoperative DTI in CM patients and discovered that postoperative DTI metrics exhibited considerable abilities in detecting the degree of neurological improvement in CM patients after surgical treatment. Additionally, to clarify the unexpected DTI values changes after surgical treatment in clinical study, the alteration of DTI metrics before and after surgery was scrutinized in animals. We found that the FA value improved after surgical decompression together with functional recovery, and that remyelination occurred in the decompressed cervical spinal cord, which may account for the neurological improvement after surgical decompression.

In summary, DTI enables us to evaluate the CM patients in a reliable and quantitative approach, which may help in clinical decision and improve the surgical outcome.