Regional brain volume changes in Hakim's disease versus Alzheimer's and mild cognitive impairment

Abstract

Idiopathic normal-pressure hydrocephalus (Hakim's disease) is characterized by ventricular enlargement and disproportionately enlarged subarachnoid space hydrocephalus, leading to localized brain deformation. Differentiating regional brain volume changes in Hakim's disease from those in Alzheimer's disease, Hakim's disease with Alzheimer's disease, and mild cognitive impairment provides insights into disease-specific mechanisms. This study aimed to identify disease-specific patterns of brain volume changes in Hakim's disease, Alzheimer's disease, Hakim's disease with Alzheimer's disease, and mild cognitive impairment and compare them with those in cognitively healthy individuals using an advanced artificial intelligence-based brain segmentation tool. The study included 970 participants, comprising 52 patients with Hakim's disease, 256 with Alzheimer's disease, 25 with Hakim's disease with Alzheimer's disease, 163 with mild cognitive impairment, and 474 healthy controls. The intracranial spaces were segmented into 100 brain and 7 CSF subregions from 3D T1-weighted MRIs using brain subregion analysis. The volume ratios of these regions were compared among the groups using Glass's Δ, referencing 400 healthy controls aged ≥50 years. Hakim's disease exhibited significant volume reduction in the supramarginal gyrus of the parietal lobe and the paracentral gyrus of the frontal lobe. Alzheimer's disease exhibited prominent volume loss in the hippocampus and temporal lobe, particularly in the entorhinal cortex, fusiform gyrus, and inferior temporal gyrus. Hakim's disease with Alzheimer's disease showed significant volume reductions in the supramarginal gyrus of the parietal lobe, similar to Hakim's disease, whereas temporal lobe volumes were relatively preserved compared with those in Alzheimer's disease. Patients with mild cognitive impairment aged ≥70 years had comparable regional brain volume ratios with healthy controls in the same age group. The Hakim's disease and Hakim's disease with Alzheimer's disease groups were characterized by volume reductions in the frontal and parietal lobes caused by disproportionately enlarged subarachnoid space hydrocephalus-related compression compared with temporal lobe atrophy observed in the Alzheimer's disease group. These disease-specific morphological changes highlight the need for longitudinal studies to clarify the causes of compression and atrophy.

Keywords: Alzheimer’s disease; Hakim’s disease; artificial intelligence; idiopathic normal-pressure hydrocephalus; mild cognitive impairment.

Graphical Abstract

Figure 1

Segmented region volume ratio of cerebrospinal fluid spaces. Each graph has violin plots for the distribution of the segmented volume ratio and a circle with line graphs for the mean volume ratio in each decade stratified according to disease. Hakim’s disease (HD, N = 52), Alzheimer’s disease (AD, N = 256), HD with AD (HD + AD, N = 25), mild cognitive impairment (MCI, N = 163), and normal volunteers (N = 474). (A) Ventricle; (B) Subarachnoid space; (C) Upper part of lateral ventricle; (D) Inferior horn of lateral ventricle; (E) Third ventricle; and (F) Fourth ventricle.

Figure 2

Segmented region volume ratio of large subregions. Each graph has violin plots for the distribution of the segmented volume ratio and a circle with line graphs for the mean volume ratio in each decade stratified according to disease. Hakim’s disease (HD, N = 52), Alzheimer’s disease (AD, N = 256), HD with AD (HD + AD, N = 25), mild cognitive impairment (MCI, N = 163), and normal volunteers (N = 474). (A) Cortical grey matter; (B) Subcortical grey matter; (C) Cerebral white matter; and (D) Cerebellum.

Figure 3

Segmented region volume ratio of mid-subregions. Each graph has violin plots for the distribution of the segmented volume ratio and a circle with line graphs for the mean volume ratio in each decade stratified according to disease. Hakim’s disease (HD, N = 52), Alzheimer’s disease (AD, N = 256), HD with AD (HD + AD, N = 25), mild cognitive impairment (MCI, N = 163), and normal volunteers (N = 474). (A) Frontal cortex; (B) Parietal cortex; (C) Temporal cortex; and (D) Occipital cortex.

Figure 4

Segmented region volume ratio of specific brain subregions. Each graph has violin plots for the distribution of the segmented volume ratio and a circle with line graphs for the mean volume ratio in each decade stratified according to disease. Hakim’s disease (HD, N = 52), Alzheimer’s disease (AD, N = 256), HD with AD (HD + AD , N = 25), mild cognitive impairment (MCI, N = 163), and normal volunteers (N = 474). (A) Supramarginal gyrus; (B) Pars triangularis; (C) Paracentral gyrus; (D) Superior frontal gyrus; (E) Entorhinal cortex; (F) Fusiform gyrus; (G) Insula cortex; (H) Hippocampus; and (I) Inferior temporal gyrus.

Figure 5

Superficial distribution of the mean Glass’s Δ. Superior, inferior, outer lateral, and inner lateral surface views showing the distribution of the mean Glass’s Δ in the Alzheimer’s disease (AD, N = 256) (A), Hakim’s disease (HD, N = 52) (B), HD with AD (HD + AD, N = 25) (C), and mild cognitive impairment groups (N = 163) (D).

Figure 6

Mean Glass’s Δ on axial views surrounding the hippocampus and temporal lobe. A shows the names of each automatically segmented subregion observed within the cross-section. The mean Glass’s Δ in the Alzheimer’s disease (AD, N = 256) (B), Hakim’s disease (HD, N = 52) (C), HD with AD (HD + AD, N = 25) (D), and mild cognitive impairment groups (N = 163) (D).