Structural & Functional Neuroimaging in Dementia
Neuroimaging is a cornerstone of dementia evaluation, providing critical data on brain atrophy patterns, cerebrovascular pathology, metabolic dysfunction, and pathologic protein deposition. Structural MRI remains the most widely used modality, enabling identification and localization of volume loss, exclusion of non-degenerative causes, and assessment of vascular disease burden. Functional and molecular imaging — FDG-PET, amyloid PET, tau PET, and dopaminergic SPECT — provide disease-specific patterns that improve differential diagnosis. The approval of anti-amyloid monoclonal antibodies (lecanemab, donanemab) has further elevated neuroimaging's role in treatment eligibility determination and safety monitoring.
Bottom Line
- MRI is the primary clinical tool: T1 for atrophy, T2/FLAIR for white matter hyperintensities (WMH), SWI for microbleeds, DWI for white matter integrity
- Atrophy patterns differ by diagnosis: AD → medial temporal/hippocampal; FTD → frontal/anterior temporal; DLB/PDD → relative medial temporal sparing; VCI → WMH + lacunes
- FDG-PET is diagnosis-specific: AD → temporoparietal; FTD → frontal; DLB → occipital with cingulate island sign
- Visual rating scales (MTA, GCA, Fazekas) provide standardized semi-quantitative assessments for clinical practice
- Amyloid PET is required for anti-amyloid therapy eligibility and is CMS-reimbursed
- ARIA monitoring with serial MRI is mandatory during anti-amyloid treatment, requiring multiple scans in the first year
- Biomarker cascade: Amyloid deposition precedes tau, neurodegeneration, and symptoms by up to 20 years
When to Image & Modality Selection
Neuroimaging should be performed in all patients presenting with cognitive impairment to exclude structural causes and refine the differential diagnosis. Guidelines recommend at least one structural brain imaging study during the initial evaluation.
Indications for Brain Imaging in Cognitive Decline
- Initial evaluation: All patients with new-onset cognitive complaints or suspected dementia
- Rapid or atypical decline: Subacute onset, focal deficits, seizures, gait disturbance, or personality change
- Exclusion of reversible causes: Normal pressure hydrocephalus, subdural hematoma, mass lesion, infection, or stroke
- Diagnostic uncertainty: FDG-PET improves certainty by ~50–60% (CMS-approved for this indication)
- Anti-amyloid therapy candidacy: Amyloid PET (or CSF biomarkers) plus baseline MRI within 12 months of treatment
- Safety monitoring: Serial MRI during anti-amyloid therapy to screen for ARIA
| Modality | Primary Clinical Use | Advantages | Limitations |
|---|---|---|---|
| CT | Emergency screening; MRI contraindications | Fast, widely available | Poor gray-white contrast; limited atrophy sensitivity |
| Structural MRI | Atrophy patterns; vascular disease; exclusion of structural causes | High spatial resolution; multiple sequences | Contraindicated with certain implants |
| FDG-PET | Differential diagnosis via metabolic patterns | Disease-specific patterns; CMS-approved | Radiation; cost; limited availability |
| Amyloid PET | Amyloid positivity for diagnosis and treatment eligibility | High sensitivity for Aβ plaques; CMS-reimbursed | Positive in ~25–30% of normal elderly |
| Tau PET | Disease staging; treatment stratification | Recapitulates Braak staging in vivo | Only flortaucipir FDA-approved; limited non-AD sensitivity |
| Dopaminergic SPECT/PET | Differentiating DLB/PDD from AD | Detects presynaptic dopaminergic loss | Cannot distinguish PDD vs DLB vs MSA |
Key MRI Sequences in Dementia
| Sequence | Primary Target | Key Findings in Dementia |
|---|---|---|
| T1-weighted | Gray matter anatomy/atrophy | Hippocampal atrophy (AD), frontal/temporal atrophy (FTD), midbrain atrophy (PSP) |
| T2-weighted | Fluid-containing lesions | WMH, lacunes, hot cross bun sign (MSA) |
| FLAIR | WMH; periventricular disease | Fazekas grading of WMH severity; ARIA-E detection |
| SWI / T2* GRE | Hemosiderin; iron deposition | Microbleeds, superficial siderosis (CAA); ARIA-H; swallow tail sign loss (DLB) |
| DWI | White matter integrity; acute ischemia | Reduced WM integrity in VCI/AD; cortical ribboning in prion disease |
Structural MRI Atrophy Patterns by Dementia Type
The distribution and severity of brain atrophy on structural MRI provides powerful diagnostic information. Characteristic patterns help differentiate among neurodegenerative conditions, though overlap occurs in mixed-pathology cases.
| Diagnosis | Characteristic Atrophy Pattern | Distinguishing Features |
|---|---|---|
| Late-onset AD | Medial temporal (hippocampus, entorhinal), lateral temporal, parietal | Hippocampal atrophy ~4.7%/yr; progressive medial temporal → global |
| Early-onset AD | Lateral temporal, inferior/medial parietal with relative medial temporal sparing | More severe global atrophy at same cognitive level vs late-onset |
| PCA | Occipital, visual association cortex, posterior parietal/temporal | Faster occipital atrophy rates than logopenic aphasia |
| Logopenic aphasia | Left temporoparietal (middle/superior temporal, inferior parietal) | Dominant hemisphere predominant; left sylvian fissure widening |
| bvFTD | Frontal, temporal, anterior cingulate, insula (bilateral) | Tau subtype: frontal > posterior; TDP-43: markedly asymmetric |
| Semantic dementia | Ventrolateral anterior temporal pole (left > right) | Right-dominant variant → prosopagnosia |
| Nonfluent aphasia | Left inferior frontal (Broca area), insula, frontal operculum | Widening of left sylvian fissure; progression along frontal aslant tract |
| DLB / PDD | Insula, cingulate, temporal-occipital, orbitofrontal | Relative medial temporal sparing (vs AD); DLB > PDD severity |
| VCI | WMH, lacunes, gray/white matter atrophy proportional to vascular burden | Atrophy correlates with WMH severity; scattered vascular lesions |
| PSP | Midbrain > pons; posterior frontal, caudate, brainstem, cerebellum | Hummingbird sign (sagittal); Mickey Mouse sign (axial) |
| CBD | Asymmetric frontal/parietal (motor/sensory cortex), corpus callosum, basal ganglia | Contralateral to most affected body side; MR parkinsonism index |
| MSA | Cerebellum, pons, thalamus, substantia nigra | Hot cross bun sign (cruciform T2 pontine hyperintensity) |
Visual Rating Scales
Standardized visual rating scales allow clinicians to assess atrophy and white matter disease semi-quantitatively without volumetric software. These scales are widely used in clinical practice and referenced in treatment eligibility criteria.
Medial Temporal Atrophy (MTA) Scale — Scheltens
Assessed on coronal T1-weighted images through the hippocampus, evaluating choroidal fissure width, temporal horn width, and hippocampal height. Age-adjusted abnormal thresholds: MTA ≥1.5 if <75 years; MTA ≥2 if ≥75 years. Most useful for supporting AD diagnosis but not specific — medial temporal atrophy occurs in other dementias and normal aging.
| Grade | Choroidal Fissure | Temporal Horn | Hippocampal Height | Interpretation |
|---|---|---|---|---|
| 0 | Normal | Normal | Normal | No atrophy |
| 1 | Widened | Normal | Normal | Questionable |
| 2 | Moderately widened | Widened | Decreased | Mild atrophy |
| 3 | Widely open | Moderately widened | Moderately decreased | Moderate atrophy |
| 4 | Widely open | Widely open | Severely decreased | Severe atrophy |
Global Cortical Atrophy (GCA) Scale — Pasquier
GCA should be assessed regionally (frontal, temporal, parietal) to identify pattern-specific atrophy. Disproportionate frontal atrophy suggests FTD; posterior/parietal atrophy suggests PCA; global symmetric atrophy is common in late-stage AD.
| Grade | Description |
|---|---|
| 0 | No cortical atrophy — normal sulci |
| 1 | Mild — slight widening of sulci |
| 2 | Moderate — volume loss of gyri |
| 3 | Severe — knife-blade gyri |
Fazekas White Matter Hyperintensity Scale
| Grade | Periventricular WMH | Deep WMH |
|---|---|---|
| 0 | Absent | Absent |
| 1 | Caps or pencil-thin lining | Punctate foci |
| 2 | Smooth halo | Beginning confluence |
| 3 | Irregular, extending into deep WM | Large confluent areas |
Clinical Significance of Fazekas Grade 3
- Fazekas ≥3 is an exclusion criterion for anti-amyloid monoclonal antibody therapy (lecanemab, donanemab)
- Severe WMH increases risk for ARIA, particularly ARIA-H (hemorrhage)
- Must be assessed on FLAIR sequences and documented before treatment initiation
- Progressive parietal WMH has been associated with increased incidence of AD onset
FDG-PET Metabolic Patterns
FDG-PET measures regional glucose metabolism and provides disease-specific hypometabolism patterns for differential diagnosis. CMS reimburses FDG-PET for diagnostic uncertainty because it improves certainty by ~50–60%. The distinction between AD temporoparietal and FTD frontal patterns is one of the most clinically useful applications.
FDG-PET Hypometabolism by Diagnosis
- AD: Lateral temporoparietal and posterior cingulate/precuneus; frontal involvement in later stages
- bvFTD: Frontal lobe (orbitofrontal, anterior cingulate); spreads to parietal and temporal lobes
- Semantic dementia: Asymmetric anterior temporal (left > right), especially temporal pole
- Progressive nonfluent aphasia: Left inferior frontal (Broca area) and peri-insular
- DLB / PDD: Occipital, parietal, and frontal with cingulate island sign (relative sparing of posterior cingulate vs surrounding areas); medial temporal metabolism preserved
- PSP: Prefrontal, caudate, pallidum, mesencephalon, subthalamic nucleus, thalamus
- CBD: Asymmetric posterior frontal, sensorimotor cortex, thalamus, basal ganglia
- MSA: Cerebellar and pontine hypometabolism with minimal cortical involvement
- VCI: Focal, asymmetric, or scattered near cortical/subcortical arteries or watershed regions
Amyloid & Tau PET
Amyloid PET
Amyloid PET tracers detect Aβ deposition and are essential for determining treatment eligibility. Key clinical points:
- AD patients show widespread cortical amyloid deposition in frontal, parietal, and temporal association cortices
- Approximately 25–30% of cognitively normal older adults are amyloid-positive (strongly associated with APOE*ε4)
- Longitudinal accumulation is slow, reaching an asymptotic state at high deposition levels
- Generally negative in FTD, PSP, CBD, and VCI unless comorbid AD pathology is present
- In DLB, >50% are amyloid-positive; in CAA, occipital-predominant deposition is characteristic
Tau PET
Tau PET (flortaucipir, the only FDA-approved tracer) recapitulates Braak staging in vivo. In AD, deposition progresses from mesial temporal → lateral temporal/parietal → frontal cortex. The TRAILBLAZER-ALZ 2 trial demonstrated greatest donanemab efficacy in patients with low to medium tau burden, supporting tau PET for treatment stratification. Current tau tracers have limited sensitivity to non-AD tauopathies; signal in bvFTD and semantic dementia may represent off-target binding to TDP-43 or monoamine oxidase.
Dopaminergic & Other Molecular Imaging
Dopamine transporter (DaT) SPECT detects presynaptic dopaminergic loss characteristic of Lewy body diseases. Both presynaptic and postsynaptic dopaminergic targets show reduced binding in DLB, PDD, and MSA, providing excellent differentiation from AD where dopaminergic systems are relatively preserved. However, DaT SPECT cannot distinguish among parkinsonian syndromes (PDD vs DLB vs MSA vs PSP).
The swallow tail sign on SWI — a hyperintense signal from intact nigrosome-1 dopaminergic cells in the substantia nigra — is lost in DLB and Parkinson disease. A novel alpha-synuclein PET tracer (ACI-12589) has shown promising uptake in cerebellar white matter and middle cerebellar peduncles in MSA, sensitively distinguishing MSA from other synucleinopathies and neurodegenerative disorders.
Vascular Imaging: STRIVE-2 Criteria (2023)
The updated Standards for Reporting Vascular Changes on Neuroimaging (STRIVE-2) define seven MRI markers of cerebral small vessel disease: recent small subcortical infarcts, lacunes, WMH, enlarged perivascular spaces, cerebral microbleeds, cortical superficial siderosis, and cortical cerebral microinfarcts. In VCI, these findings are much more severe and widespread than in normal aging. Mixed AD-vascular pathology is the most common finding at autopsy in dementia patients, with greater WMH burden and white matter degeneration than either pathology alone.
Neuroimaging in Anti-Amyloid Therapy
The approval of lecanemab and donanemab has made neuroimaging essential for both treatment eligibility and safety monitoring. Baseline MRI within 12 months of treatment initiation is required.
MRI Exclusion Criteria for Anti-Amyloid Therapy (Lecanemab)
- >4 microhemorrhages (≤10 mm) or ≥1 macrohemorrhage (>10 mm)
- Any superficial siderosis or evidence of vasogenic edema
- ≥2 lacunar infarcts or stroke involving a major vascular territory
- Severe subcortical hyperintensities (Fazekas ≥3)
- Evidence of Aβ angiitis, CAA-related inflammation, or other major intracranial pathology
ARIA Monitoring
Amyloid-related imaging abnormalities (ARIA) are the most significant side effect of anti-amyloid antibodies. Risk is highest with comorbid cerebrovascular disease, CAA, and APOE*ε4/ε4 homozygosity. ARIA-E (vasogenic edema/sulcal effusion) appears as hyperintense signal on T2/FLAIR. ARIA-H (micro- or macrohemorrhage) appears as hypointense hemosiderin deposits on SWI/T2* GRE. Most cases are asymptomatic or mild and resolve within 4 months after treatment suspension.
Required safety MRI for lecanemab: before the 5th infusion (~9–10 weeks), 7th infusion (~13–14 weeks), 14th infusion (~27–28 weeks), and 26th infusion (~1 year). Additional safety MRI is indicated for symptoms such as headache, confusion, dizziness, vomiting, loss of consciousness, or gait disturbance.
Landmark Neuroimaging Signs
| Sign | Finding | Associated Condition |
|---|---|---|
| Hummingbird sign | Midbrain atrophy on midsagittal view | PSP |
| Mickey Mouse / morning glory sign | Midbrain atrophy on axial view | PSP |
| Hot cross bun sign | Cruciform T2 hyperintensity in pons | MSA (nonspecific but characteristic) |
| Swallow tail sign (loss) | Absent nigrosome-1 hyperintensity on SWI | DLB, Parkinson disease |
| Cingulate island sign | Preserved posterior cingulate metabolism on FDG-PET | DLB / PDD (distinguishes from AD) |
| Knife-blade atrophy | Severe cortical thinning, markedly widened sulci | Advanced FTD; severe AD |
Longitudinal Monitoring & Biomarker Cascade
Serial neuroimaging provides important prognostic information and helps monitor disease progression across the AD continuum.
Prognostic Value of Serial Neuroimaging
- Hippocampal atrophy rate: AD ~4.7%/yr vs ~1.5%/yr in normal aging; greater baseline entorhinal/hippocampal atrophy predicts faster MCI → AD conversion
- MCI converters: Intermediate atrophy rates between AD and normals; temporoparietal FDG-PET hypometabolism predicts future dementia conversion
- Preclinical AD: Accelerated atrophy in Aβ-positive cognitively normal adults, especially those who subsequently convert to MCI or AD
- Biomarker cascade (Jack model): Amyloid (~20 yr pre-symptom) → tau (~15 yr) → functional changes → atrophy (~5–10 yr) → cognitive decline → dementia
- Tau PET progression: Slower than amyloid accumulation but faster in cognitively impaired individuals; correlates with cognitive decline trajectory
- Autosomal dominant AD (DIAN): Amyloid abnormalities ~20 years before expected onset, glucose hypometabolism ~15 years, cortical thinning ~5–10 years; precuneus earliest region affected across all three biomarkers
Revised A/T/N Biological Framework (2024)
The 2024 update expands the original A/T/N framework to include inflammation (I), synucleinopathy (S), and vascular damage (V). Under these criteria, any amyloid-positive individual is classified as having AD. Core 2 biomarkers (tau PET) and neurodegeneration markers (MRI, FDG-PET) are used for staging disease severity and predicting the trajectory of cognitive decline.
| Category | Biomarker | Role |
|---|---|---|
| A (Amyloid) — Core 1 | Amyloid PET; CSF/plasma Aβ42 | Defines the disease |
| T1 (Secreted tau) — Core 1 | CSF/plasma pTau 217, 181, 231 | Defines the disease |
| T2 (Tau proteinopathy) — Core 2 | Tau PET | Staging severity |
| N (Neurodegeneration) | Structural MRI/CT; FDG-PET | Staging (nonspecific to AD) |
| I (Inflammation) | CSF GFAP | Nonspecific to AD |
| V (Vascular) | MRI (infarction, WMH) | Copathology marker |
| S (Synuclein) | CSF alpha-synuclein seed amplification assay | Copathology marker |
References
- Risacher SL. Neuroimaging in dementia. Continuum (Minneap Minn). 2024;30(6, Dementia):1761-1789.
- Jack CR, Andrews JS, Beach TG, et al. Revised criteria for diagnosis and staging of Alzheimer's disease. Alzheimers Dement. 2024;20(8):5143-5169.
- Duering M, Biessels GJ, Brodtmann A, et al. Neuroimaging standards for research into small vessel disease — advances since 2013. Lancet Neurol. 2023;22(7):602-618.
- Rabinovici GD, Gatsonis C, Apgar C, et al. Association of amyloid PET with subsequent change in clinical management. JAMA. 2019;321(13):1286-1294.
- Sims JR, Zimmer JA, Evans CD, et al. Donanemab in early symptomatic Alzheimer disease: the TRAILBLAZER-ALZ 2 trial. JAMA. 2023;330(6):512-527.
- Cummings J, Apostolova L, Rabinovici GD, et al. Lecanemab: appropriate use recommendations. J Prev Alzheimers Dis. 2023;10(3):362-377.
- Hampel H, Elhage A, Cho M, et al. Amyloid-related imaging abnormalities (ARIA): radiological, biological and clinical characteristics. Brain. 2023;146(11):4414-4424.
- Harper L, Barkhof F, Scheltens P, et al. An algorithmic approach to structural imaging in dementia. J Neurol Neurosurg Psychiatry. 2014;85(6):692-698.
- Gordon BA, Blazey TM, Su Y, et al. Spatial patterns of neuroimaging biomarker change in autosomal dominant Alzheimer's disease: a longitudinal study. Lancet Neurol. 2018;17(3):241-250.
- Smith R, Capotosti F, Schain M, et al. The alpha-synuclein PET tracer [18F] ACI-12589 distinguishes multiple system atrophy from other neurodegenerative diseases. Nat Commun. 2023;14(1):6750.