| Article Access Statistics|
| Viewed||479 |
| Printed||22 |
| Emailed||0 |
| PDF Downloaded||31 |
| Comments ||[Add] |
Click on image for details.
|Year : 2020 | Volume
| Issue : 5 | Page : 1048-1049
Pattern Recognition: An Aid to Neuroimaging Diagnosis
Professor and Head, Department of Radiology, SCTIMST, Thiruvananthapuram, Kerala, India
|Date of Web Publication||27-Oct-2020|
Dr. C Kesavadas
Professor and Head of Radiology, SCTIMST, Thiruvananthapuram, Kerala
Source of Support: None, Conflict of Interest: None
|How to cite this article:|
Kesavadas C. Pattern Recognition: An Aid to Neuroimaging Diagnosis. Neurol India 2020;68:1048-9
Patients with suspected neurological complication of dengue fever undergo craniospinal imaging using CT or MRI, the latter being better in detecting and characterising the abnormal findings. Though rare, central neurological complications are known with dengue and presentation may be as encephalitis, encephalopathy, meningitis, ischemic/hemorrhagic stroke, acute disseminated encephalomyelitis, cerebellar syndrome & transverse myelitis. In a routine clinical set up, addition of imaging especially MRI findings to clinical and laboratory findings will help in better differentiation of some of these complications. In addition, imaging will give information regarding prognosis based on imaging characteristics, extent and anatomy of the lesion.
Several brain MRI findings have been described with Dengue. This includes lesions that involve both grey and white matter, cerebellar involvement, thalamic, basal ganglia, substantia nigra, brain stem involvement, micro and macro hemorrhages, lesions showing diffusion restriction & findings of posterior reversible leukoencephalopathy. Vyas, et al., in this issue of journal have described the MRI findings in 36 patients with neurological symptoms in serologically proven dengue patients. A pattern recognition approach has been used to evaluate and group the MRI findings so as to differentiate the clinical condition seen as a complication of Dengue.
Pattern recognition is frequently used in medical imaging to classify disease entities., Disease categorisation and grouping is done based on disease process and location. Statistical methods, computational methods and machine learning have been used for pattern recognition. These methods have been used in structural and functional neuroimaging for diagnosis and outcome prediction of neurological diseases. Classification of disease entities is done based on specific imaging features. The methods can also be used for quantitative imaging data.
Vyas, et al., have used specific imaging features to classify the disease as Encephalitis pattern, Encephalopathy pattern, Acute Disseminated Encephalomyelitis, Hemorrhagic diathesis. The approach appears quite simplistic. Based on specific imaging features, a patient's MRI finding was grouped into a particular pattern. For example, if the MRI showed involvement of thalami, basal ganglia, brain stem, cerebellum, cortical gray or subcortical white matter in isolation or combination the MRI pattern was that of encephalitis. However, it may be noted that when the several imaging findings described in literature are grouped into 4 patterns, there is a possibility of either overlap or exclusion of some observed findings. A specific imaging feature such as hippocampal hyperintensity can be seen both in encephalitis and seizure induced change in encephalopathy. Lesions of posterior reversible encephalopathy syndrome seen in cerebellum can mimic lesions in cerebellitis & ADEM.
The merit of pattern recognition lies in its systematic analysis of imaging features that can then be correlated with the clinical and laboratory findings to reach a diagnosis. Differentiating encephalitis, encephalopathy and ADEM can help in improved medical management and prognostication in Dengue fever. The method of pattern recognition used in the study by Vyas, et al., has given a direction towards such a possibility. In conclusion, while applying the various pattern recognition methods in a clinical scenario it is essential to understand the merits and limitations of the used method.
| » References|| |
Li GH, Ning ZJ, Liu YM, Li XH. Neurological Manifestations of Dengue Infection. Front Cell Infect Microbiol. 2017;7:449.
Vanjare HA, Mannam P, Mishra AK, Karuppusami R, Carey RAB, Abraham AM, et al
. Brain Imaging in Cases with Positive Serology for Dengue with Neurologic Symptoms: A Clinicoradiologic Correlation. AJNR Am J Neuroradiol 2018;394:699-703.
Vyas S, Ray N, Maralakunte M, Kumar A, Singh P, Khandelwal N, et al
. Pattern recognition approach to brain MRI findings in patients with dengue fever with neurological complications Neurology India 2020 [Ahead of Print].
Steenweg ME, Vanderver A, Blaser S, Bizzi A, De Koning TJ, Mancini GMS, et al
. Magnetic resonance imaging pattern recognition in hypomyelinating disorders Brain. 2010;13310:2971-82.
Faria AV, Liang Z, Miller MI, Mori S. Brain MRI Pattern Recognition Translated to Clinical Scenarios. Front Neurosci 2017;11:578.