Neurophet said a prospective study validating its brain‐degeneration analysis software, Neurophet AQUA, has been published in the Annals Academy of Medicine, Singapore, showing strong agreement with human reads and established automated pipelines while cutting processing time to minutes.

The single-site study enrolled 90 adults between April 2022 and July 2023 at the Dementia Research Centre, Lee Kong Chian School of Medicine, Nanyang Technological University. 

Participants were classified as cognitively normal (n=30), mild cognitive impairment (n=40) or dementia (n=20). Researchers compared AQUA’s automated ratings of white matter hyperintensities (WMH) and gray matter volumes with expert visual ratings and with two standard toolkits, LST and CAT12.

AQUA’s WMH ratings correlated with human visual reads for total, periventricular and deep WMH (ρ=0.66, 0.50 and 0.57; P<0.0001). AQUA’s WMH volumes also tracked closely with LST and CAT12 (ρ=0.84–0.85; P<0.0001), and those associations held after adjusting for age, sex and diagnosis. 

On gray matter, AQUA detected expected atrophy patterns across the dementia spectrum, with significant differences in the hippocampus and posterior cingulate cortex; pair-wise tests separated dementia from normal cognition and from mild incognitive impairment.

The team reported run times under five minutes for gray-matter segmentation and under 20 minutes for combined WMH and gray-matter analysis, compared with several hours for conventional workflows.

Neurophet said the results support AQUA’s use for faster, more objective dementia workups and longitudinal monitoring in routine practice. “Our study demonstrated clinical effectiveness of Neurophet AQUA in a Singapore cohort, an important step in validating our technology globally,” co-CEO Kim Dong-hyun said.