Cerebral Small Vessel Disease Biomarkers Detection on MRI-Sensor-Based Image and Deep Learning

Area of Science:

• Medical Imaging
• Artificial Intelligence in Medicine
• Neurology

Background:

• Magnetic Resonance Imaging (MRI) provides detailed brain structure visualization.
• Cerebrovascular disease is a leading cause of death and stroke globally and in Taiwan.
• Small blood vessel lesions are common contributors to stroke.

Purpose of the Study:

• To establish a computer-aided diagnosis system for identifying small blood vessel lesions in brain MRI images.
• To utilize Convolutional Neural Networks (CNNs) and deep learning for analyzing brain vascular occlusion.
• To assist clinicians in determining stroke probability and severity.

Main Methods:

• Analysis of MRI data from 50 patients (30 stroke, 17 occlusion without stroke, 3 dementia).
• Application of deep learning and CNNs to detect and analyze cerebral small vessel lesions.
• Development of a system to output labeled images with lesion location, range, and size.

Main Results:

• The system identifies cerebral small vessel lesions in brain MRI images.
• Outputs include lesion coordinates, range, area size, and potential stroke risk.
• A 3D display of brain small vessels is synthesized to aid diagnosis.

Conclusions:

• The developed computer-aided diagnosis system enhances the detection of small blood vessel lesions.
• It assists clinicians in making faster and more accurate stroke-related diagnoses.
• The system provides precise lesion localization for improved patient care.