Dementia & Movement-Disorder Imaging: Which Test When
Choosing among FDG-PET, amyloid PET, tau PET, and DAT-SPECT across the cognitive and parkinsonian differential
Quick Answer
Molecular imaging in cognitive and movement disorders answers different questions with different tracers, and choosing well depends on the specific clinical uncertainty. FDG-PET maps regional hypometabolism and is best for distinguishing dementia subtypes by pattern (e.g. Alzheimer disease vs frontotemporal dementia). Amyloid PET answers a binary biological question — is amyloid pathology present — most useful to rule Alzheimer pathology out in atypical or early-onset cases. Tau PET adds pathology that tracks stage and clinical severity more closely than amyloid. DAT-SPECT (¹²³I-ioflupane) images the nigrostriatal dopaminergic terminal, separating degenerative parkinsonism and dementia with Lewy bodies from non-degenerative mimics. Match the tracer to whether the question is pattern, amyloid presence, tau burden, or dopaminergic integrity.
The decision in one place
Ask what you actually need to know. Pattern of neurodegeneration → FDG-PET. Is Alzheimer amyloid present (especially to exclude it) → amyloid PET. Is there tau pathology / how advanced → tau PET. Is the dopaminergic system degenerating (PD/DLB vs essential tremor or drug-induced) → DAT-SPECT. Using amyloid or tau PET without a specific management question — or in typical late-onset dementia where the answer won't change care — is low-value.
Head-to-head
| Test | What it measures | Best question | Caveat |
|---|---|---|---|
| FDG-PET | Regional glucose metabolism | Subtype by pattern (AD temporoparietal/posterior cingulate vs FTD frontotemporal vs DLB occipital) | Nonspecific; patterns overlap; not pathology-specific |
| Amyloid PET | Fibrillar β-amyloid plaques | Presence/absence of amyloid — strong to exclude AD in atypical/early-onset cases | Amyloid can be positive in cognitively normal elderly; positivity ≠ cause |
| Tau PET | Paired-helical-filament tau | Tau burden/topography, tracks stage and severity | Availability; off-target binding; largely specialist/research–clinical use |
| DAT-SPECT (¹²³I-ioflupane) | Presynaptic dopamine transporter | Degenerative parkinsonism / DLB vs non-degenerative (essential tremor, drug-induced, psychogenic) | Abnormal in PD, MSA, PSP alike — does not subtype among degenerative parkinsonisms |
How to sequence
By clinical problem:
- Typical late-onset dementia, subtype unclear: FDG-PET for the metabolic pattern (AD vs FTD vs DLB).
- Atypical, early-onset, or diagnostically uncertain cognitive impairment where confirming/excluding AD changes management: amyloid PET (per appropriate-use criteria) — a negative scan effectively excludes AD as the cause.
- Suspected dementia with Lewy bodies vs Alzheimer disease: DAT-SPECT (reduced striatal uptake supports DLB) and/or the occipital hypometabolism pattern on FDG.
- Parkinsonism where the question is degenerative vs not (e.g. tremor of uncertain cause, drug-induced, psychogenic): DAT-SPECT — but remember it cannot distinguish PD from MSA/PSP.
- Staging AD pathology / trial or specialist context: tau PET, which correlates with clinical stage better than amyloid.
Key selection concepts
Amyloid positivity is not the same as amyloid causation — plaques accumulate in many cognitively normal older adults, so amyloid PET is most powerful as a rule-out (a negative scan makes AD very unlikely). DAT-SPECT confirms a degenerative dopaminergic process but does not name which one — it separates PD/DLB/atypical parkinsonism (all abnormal) from essential tremor, drug-induced, and psychogenic parkinsonism (normal). Using each test for the question it can actually answer is the core skill.
Pitfalls in selection
- Appropriate-use criteria exist for amyloid PET precisely because indiscriminate use in typical dementia rarely changes management.
- Medication effects on DAT (some stimulants, cocaine) and technical/normal-variant issues affect interpretation.
- Mixed pathology is common in older patients; a single positive test rarely tells the whole story.
Bottom line
Choose by the question: FDG for pattern, amyloid PET for amyloid presence (best as a rule-out), tau PET for burden/stage, and DAT-SPECT for dopaminergic integrity in the parkinsonian/DLB differential.
Self-Check
Q1. What question is amyloid PET best suited to answer, and why is it strongest as a rule-out?
Answer: Whether fibrillar β-amyloid is present. It is strongest as a rule-out because a negative scan makes Alzheimer pathology very unlikely, whereas positivity also occurs in cognitively normal elders (positivity ≠ causation).
Q2. What does DAT-SPECT distinguish, and what can it not distinguish?
Answer: It separates degenerative dopaminergic parkinsonism/DLB (abnormal) from non-degenerative mimics (normal) — essential tremor, drug-induced, psychogenic. It cannot distinguish PD from MSA or PSP (all abnormal).
Q3. Which test best separates dementia subtypes by pattern, and give an example pattern.
Answer: FDG-PET — e.g. temporoparietal/posterior cingulate hypometabolism in Alzheimer disease vs frontotemporal in FTD vs occipital involvement in DLB.
Q4. When is amyloid PET considered low-value?
Answer: In typical late-onset dementia or whenever the result won't change management — hence the appropriate-use criteria restricting it to cases where confirming/excluding AD alters care.
Key References
- Amyloid PET appropriate-use criteria (SNMMI/Alzheimer's Association) and society guidance on FDG brain PET.
- EANM/SNMMI guidelines for ¹²³I-ioflupane DAT-SPECT; reviews of tau PET in clinical staging.
Evidence & sources
https://nuc.med/dementia-imaging-selection/ Report an issue