Automatic Quantification of Enlarged Perivascular Space in Patients With Traumatic Brain Injury Using...

Area of Science:

• Neuroimaging
• Radiology
• Brain imaging analysis

Background:

• Perivascular spaces (PVS) are crucial for brain waste clearance and metabolic homeostasis.
• Enlarged PVS (ePVS) is detectable via MRI but challenging to quantify using standard 2D T2w images.
• Accurate ePVS quantification typically requires high-resolution 3D MRI sequences.

Purpose of the Study:

• To evaluate a deep-learning-based super-resolution (SR) technique for improved ePVS quantification from 2D T2w images.
• To apply this SR method to patients with traumatic brain injury (TBI).

Main Methods:

• Prospective acquisition of 2D and 3D T2w MRI in 26 volunteers for SR validation.
• Retrospective application of SR to 2D T2w images in 41 chronic TBI patients.
• Automatic quantification of ePVS volumes in white matter and brain lobes; Pittsburgh Sleep Quality Index (PSQI) assessment.

Main Results:

• SR T2w images provided ePVS quantification comparable to the 3D T2w standard (r=0.98, p<0.0001).
• SR accurately reflected age-related ePVS burden increase (r=0.80, p<0.0001).
• TBI patients with poor sleep showed a greater age-related ePVS increase, significantly interacting with whole-brain and frontal lobe ePVS (p=0.047, p=0.027).

Conclusions:

• Deep-learning-based SR of 2D T2w MRI enables reliable ePVS quantification, comparable to 3D T2w volumes.
• Age-related ePVS burden is consistently observed across the white matter.
• Impaired glymphatic function due to poor sleep may accelerate ePVS burden increase post-TBI.