Shannon entropy as a stable diffusion tensor parameter for evaluating normal spinal cord regions in dogs

Area of Science:

• Veterinary Neurology
• Biomedical Imaging
• Quantitative MRI

Background:

• Diffusion tensor imaging (DTI) is utilized in veterinary medicine for diagnosing spinal cord conditions.
• Apparent diffusion coefficient (ADC) and fractional anisotropy (FA) values are key DTI metrics but are influenced by anatomical variations and imaging parameters, making lesion detection challenging.
• Shannon entropy, a measure of information variability, is explored as a potential new biomarker for structural integrity.

Purpose of the Study:

• To evaluate Shannon entropy as a novel indicator for assessing canine spinal cord structural integrity.
• To compare the utility of Shannon entropy against traditional ADC and FA values in a canine model.

Main Methods:

• Six healthy beagle dogs underwent magnetic resonance imaging (MRI) at 3 Tesla.
• Apparent diffusion coefficient (ADC), fractional anisotropy (FA), and Shannon entropy values were calculated for cervical and thoracolumbar spinal cord segments.
• Neurological examinations confirmed the absence of spinal cord disease in all subjects.

Main Results:

• ADC and FA values varied significantly between cervical and thoracolumbar regions, indicating anatomical dependence and lack of fixed normal values.
• Shannon entropy values showed no significant differences across intervertebral levels.
• Shannon entropy demonstrated potential for differentiating normal tissue from lesions, independent of MRI acquisition variables.

Conclusions:

• Shannon entropy is a more robust indicator of canine spinal cord structural integrity compared to ADC and FA.
• This metric shows promise for lesion detection in veterinary neurology, overcoming limitations associated with traditional DTI parameters.