Clinical Utility of AI-Enabled CT-to-MRI Translation for Degenerative Spinal Disorders: A Retrospective Reader Study

Area of Science:

• Radiology
• Artificial Intelligence
• Medical Imaging

Background:

• Magnetic Resonance Imaging (MRI) is the gold standard for assessing spinal degenerative disorders.
• MRI accessibility can be limited by delays, unavailability, or contraindications.
• Computed Tomography (CT) findings often guide initial decisions in such cases.

Purpose of the Study:

• To evaluate the clinical utility of Dr.Magic (DRM-S-01), an AI system converting CT to MRI-like images (T^-MRI).
• To assess if T^-MRI improves diagnostic performance for disc-related assessments compared to CT alone.

Main Methods:

• A retrospective reader study involving 92 paired CT/MRI examinations.
• Three radiologists interpreted scans under three conditions: MRI-only, CT-only, and CT augmented with T^-MRI.
• MRI-only interpretation served as the reference standard.

Main Results:

• CT augmented with T^-MRI significantly improved diagnostic accuracy for disc-related assessments compared to CT-only (mean accuracy increased from 39.13% to 73.19%).
• Inter-reader agreement also improved with the addition of T^-MRI (Fleiss' κ: 0.5617 to 0.6621).
• Dr.Magic generated T^-MRI in a median of 10.90 seconds per examination.

Conclusions:

• Augmenting CT interpretation with AI-generated T^-MRI can enhance diagnostic performance and reading consistency.
• This AI-driven approach offers a practical solution when timely MRI is not feasible for spinal disorder assessment.
• The rapid translation time supports the integration of Dr.Magic into clinical CT workflows.