Portable Low-Field Magnetic Resonance Imaging in People With Human Immunodeficiency Virus

Area of Science:

• Neurology
• Medical Imaging
• Artificial Intelligence

Background:

• Aging population with HIV (PWH) faces risks of accelerated aging and dementia.
• Portable, low-field MRI (LF-MRI) offers potential for enhanced access and routine monitoring of PWH.
• Machine learning (ML) algorithms can analyze brain imaging data for atrophy and white matter hyperintensities (WMH).

Purpose of the Study:

• Evaluate the feasibility of LF-MRI for PWH.
• Apply an ML segmentation algorithm to assess brain atrophy and WMH in PWH compared to people without HIV (PWoH).
• Investigate regional brain volume differences between PWH and control groups.

Main Methods:

• Acquired LF-MRI (64 mT) data from HIV-positive individuals on antiretroviral therapy.
• Used ML algorithm WMH-SynthSeg for brain region segmentation via FreeSurfer.
• Compared intracranial volume-corrected brain regions between PWH and PWoH cohorts (VC and MCI groups), adjusting for age and sex.

Main Results:

• LF-MRI detected reduced caudate, putamen, and white matter volumes in PWH compared to PWoH with vascular comorbidities (p < 0.05).
• Hippocampal volume was comparable between PWH and PWoH (p ≥ 0.05).
• No significant differences in white matter hyperintensities (WMH) were observed between cohorts (p > 0.05).

Conclusions:

• LF-MRI is a feasible imaging technique in outpatient settings for PWH.
• ML algorithms effectively detect regional atrophy and WMH using LF-MRI data.
• LF-MRI holds promise for frequent monitoring and early detection of neurological changes in at-risk populations.