Brain FDG-PET for Dementia and Epilepsy

Brain FDG-PET converts regional glucose metabolism into an image of synaptic function, and it earns its clinical value in two settings where anatomy alone falls short: separating the neurodegenerative dementias from one another, and localizing the seizure focus in focal epilepsy. The physics is only half the story. Whether the metabolic pattern is read correctly depends on disciplined patient preparation, a genuinely resting uptake environment, a fixed reconstruction recipe, and comparison against a properly matched normal database.¹²

Fluorine-18 fluorodeoxyglucose (FDG) is a glucose analogue. It is transported into neurons and glia and phosphorylated to FDG-6-phosphate, which is then metabolically trapped. Because cerebral glucose use is dominated by synaptic activity, the resulting image is a map of where the brain is — and is not — working. In dementia, disease-specific patterns of hypometabolism appear years before structural atrophy is obvious; in epilepsy, the epileptogenic zone frequently shows interictal hypometabolism that can guide surgical planning.¹⁵⁸

This article walks through what brain FDG-PET measures, how to prepare and scan the patient, the disease-specific patterns that drive interpretation, the semiquantitative analysis that supports the reader, and the quality control that keeps a subtle study defensible. DRPS supports PET/CT programs across Florida, Maryland, Virginia, Washington DC, California, Nevada, Pennsylvania, New York, New Jersey, and Delaware through its PET/CT and nuclear medicine physics and accreditation support services.

Introduction

Brain FDG-PET is a functional study: it is read for pattern, not just for a single number. Unlike a lesion-hunting oncology scan, the neurological FDG study asks whether the distribution of glucose metabolism matches a healthy brain or one of several recognizable disease signatures. That framing changes everything downstream — preparation, acquisition, reconstruction, and analysis are all built to preserve a faithful, low-noise map of relative regional uptake.¹

In dementia, FDG-PET is used chiefly for differential diagnosis after a clinical work-up, not as a first-line screen. Guideline and appropriateness documents position it to help distinguish Alzheimer disease from frontotemporal dementia, dementia with Lewy bodies, and other causes when the clinical picture is atypical or overlapping.²³⁴ In epilepsy, interictal FDG-PET contributes to presurgical localization of the seizure focus, adding value especially when MRI is normal or discordant with the scalp EEG.¹⁸

The common thread is that FDG-PET reveals dysfunction that structure alone can miss. But the same sensitivity that makes it useful also makes it unforgiving: a talkative uptake room, an uncontrolled blood glucose, or a reconstruction change between the patient scan and the reference database can all introduce apparent "abnormality" that has nothing to do with the patient's brain.

Topic Explanation

From glucose analogue to metabolic map

FDG follows glucose into the cell through glucose transporters and is phosphorylated by hexokinase. Unlike glucose, FDG-6-phosphate is not a good substrate for the next step of glycolysis, so it accumulates in proportion to metabolic demand over the uptake period. The brain is a high-consumer organ with little metabolic reserve, and gray matter uptake is normally several-fold higher than white matter, producing the high-contrast cortical images that neuro FDG-PET depends on.¹

Because the signal reflects synaptic activity, anything that changes brain state during uptake changes the picture. Muscle tension, anxiety, active conversation, bright light, and recent seizures all leave metabolic footprints. This is why brain FDG protocols are stricter about the uptake environment than whole-body oncology protocols, where a modest amount of physiologic variability is tolerable.¹

Where it fits clinically

For cognitive decline, brain FDG-PET is one of several molecular tools. Amyloid and tau PET interrogate protein pathology directly; FDG interrogates downstream neuronal dysfunction and provides a topographic pattern. Our companion piece on amyloid and tau brain PET covers the pathology-specific tracers, and DaTscan (I-123 ioflupane) SPECT addresses the dopaminergic question in suspected Lewy body disease and parkinsonian syndromes. FDG sits alongside these, contributing the metabolic pattern that often carries the differential.²⁷

For epilepsy, FDG-PET is part of a multimodal presurgical battery that includes video-EEG, high-resolution MRI, and sometimes magnetoencephalography and ictal perfusion SPECT. Its role is localization and lateralization of the epileptogenic zone through interictal hypometabolism, which is typically broader than the true seizure-onset zone but still clinically informative.¹⁸

Key Technical Principles

Patient preparation drives image validity

Brain FDG-PET preparation centers on producing a controlled resting metabolic state:

• Fasting. The patient fasts (commonly 4–6 hours) so that circulating glucose does not competitively suppress brain FDG uptake.¹
• Blood glucose control. Elevated blood glucose competes with FDG at the transporter and reduces the target-to-background contrast; procedure guidance advises confirming reasonably controlled glucose (a commonly cited target is below about 7 mmol/L, roughly 126 mg/dL) and rescheduling or interpreting cautiously when markedly elevated.¹
• Resting uptake environment. FDG is injected in a dim, quiet room. The patient avoids talking, reading, chewing, and movement during uptake, with eyes-open or eyes-closed handled consistently across the program to keep visual cortex activation reproducible.¹
• Uptake time. Brain studies use a defined uptake interval, commonly on the order of 30 minutes, held consistent so that comparisons to a normal database and to prior studies are valid.¹

A worked look at administered activity and dose

Administered activity is chosen to balance count statistics against radiation dose. Using the effective dose coefficient for F-18 FDG of approximately $0.019{\text{mSv⋅MBq}}^{-1}$ from ICRP Publication 128, a typical adult brain administered activity translates into effective dose as:¹¹

Across a typical adult brain activity range of roughly 125–250 MBq, the radiopharmaceutical effective dose is on the order of 2.4–4.7 mSv, before adding the dose from any CT acquired for attenuation correction and localization.¹¹¹ Activity is kept as low as reasonably achievable consistent with the count statistics needed for a low-noise cortical map.

Reconstruction and correction chain

The quantitative fidelity of a brain study depends on the full correction chain: attenuation correction, scatter and randoms correction, and dead-time correction. Our overview of PET/CT attenuation correction covers the CT-based μ-map and its pitfalls, including patient motion between the CT and PET that misregisters the correction. For the brain, motion is a particular concern because a small head shift smears cortical detail and can create artifactual asymmetry.¹

Two physical effects deserve emphasis in neuro work:

• Partial volume effect. Finite spatial resolution causes apparent uptake in small or atrophic structures to be underestimated as activity spills out into adjacent tissue. In an atrophic brain, this can exaggerate apparent hypometabolism, blurring the line between atrophy and true metabolic loss. See our detailed treatment of the PET partial volume effect.¹
• Reconstruction consistency. Iteration number, subsets, post-filter, and matrix must match whatever was used to build the normal database. A change in smoothing alone can shift Z-scores enough to alter a borderline read.¹⁵

Semiquantitative analysis

Visual reading is anchored by semiquantitative tools that compare the patient to an age-matched healthy reference:

• Standardized uptake value ratio (SUVr). A target region's uptake is normalized to a reference region assumed to be relatively spared (for example, the pons, cerebellum, or whole brain), giving a unitless ratio robust to global scaling.
• Statistical parametric mapping (SPM) and 3D stereotactic surface projections (3D-SSP). These spatially normalize the brain to a template and express deviation from normal as voxel-wise Z-scores or surface maps, making subtle, distributed hypometabolism visible.⁵
• Asymmetry index (AI). For lateralized disease such as temporal lobe epilepsy, left–right asymmetry is quantified to flag the abnormal side.

Guideline panels explicitly recommend semiquantitative assessment to assist visual reading, because it improves reproducibility and sensitivity without replacing interpreter judgment.²⁵

A worked asymmetry example: suppose a homologous temporal region reads ${\text{SUVr}}_L=1.05$ and ${\text{SUVr}}_R=1.35$. The asymmetry index is:

An asymmetry of this magnitude, concordant with the clinical and EEG picture, supports left temporal hypometabolism as the likely epileptogenic side — the direction and concordance matter more than the absolute SUVr.⁸

Disease-specific metabolic patterns

The table is a reading aid, not a substitute for the full guideline descriptions. Overlap is real, mixed pathology is common, and pattern reading is always integrated with clinical data, MRI, and where indicated pathology-specific tracers.²⁴

Clinical Impact

Differential diagnosis of dementia

The central contribution of FDG-PET in cognitive decline is separating overlapping syndromes. Panels convened by the European Association of Nuclear Medicine and the European Academy of Neurology found FDG-PET useful for distinguishing Alzheimer disease from frontotemporal lobar degeneration and dementia with Lewy bodies, for evaluating atypical presentations, and for supporting a diagnosis when the clinical picture is ambiguous — while cautioning against its use for pre-clinical screening in asymptomatic individuals.² United States appropriateness guidance similarly positions FDG-PET as a problem-solving tool in patients already evaluated by a dementia specialist, particularly to separate Alzheimer disease from frontotemporal dementia.³⁴

Because the metabolic pattern often precedes marked atrophy, FDG-PET can contribute earlier diagnostic confidence, which matters as disease-modifying strategies place a premium on accurate, early characterization.¹⁰

Presurgical localization in epilepsy

In drug-resistant focal epilepsy, accurate localization of the epileptogenic zone is the gateway to potentially curative surgery. Interictal FDG-PET frequently shows hypometabolism at or around the seizure focus, and in temporal lobe epilepsy the ipsilateral temporal lobe is the classic finding. When PET is co-registered with MRI and combined with complementary modalities, localization improves: one temporal lobe epilepsy series reported lobar localization in about 94.5% of patients when FDG PET/MRI was combined with magnetoencephalography, higher than either modality alone.⁸ Metabolic patterns can also carry prognostic information; in mesial temporal lobe epilepsy, the extent and location of hypometabolism have been linked to underlying pathology and surgical outcome.⁹

The practical lesson is that FDG-PET is a localizing input, not a standalone verdict. It is read in the context of EEG, MRI, and the surgical hypothesis, and the injection must occur during a documented interictal period, because a recent or unrecognized seizure can convert hypometabolism into misleading focal hypermetabolism.¹

Practical Optimization Tips

• Standardize the uptake room. Enforce a dim, quiet environment, minimize conversation, and document the eyes-open or eyes-closed convention for every brain patient. Small, consistent choices protect the visual and sensorimotor cortices from spurious activation.¹
• Confirm and record blood glucose. Measure glucose before injection, record it, and have a written rule for how elevated values are handled, because the value directly scales brain contrast.¹
• Freeze the reconstruction. Lock the reconstruction protocol to whatever the normal database used — iterations, subsets, filter, matrix, and any resolution modeling — and change it only through a deliberate revalidation.¹⁵
• Guard against motion. Immobilize the head, keep the CT and PET acquisitions close in time, and review for CT-to-PET misregistration that can fabricate asymmetry. Reacquire when motion is evident.¹
• Use semiquantitative analysis as a second reader. Bring SPM, 3D-SSP, or Z-score maps to every complex case, but interpret them against the clinical question rather than treating a red voxel cluster as a diagnosis.²⁵
• Match the normal database to the scanner. A database built on a different scanner, reconstruction, or uptake protocol imports systematic error; where feasible, use vendor or harmonized databases appropriate to your system, and revalidate after major upgrades.¹⁵
• Watch the partial volume trap. In atrophic brains, correlate apparent hypometabolism with structural imaging so that atrophy is not over-read as metabolic loss.¹

Regulatory Considerations

Brain FDG-PET sits at the intersection of radiopharmaceutical medical-use rules and imaging accreditation, and both apply. Fluorine-18 FDG is a radioactive drug whose medical use is authorized under the federal or Agreement State framework that governs byproduct material, and the facility's radioactive material license, authorized user status, and radiation safety program must cover its receipt, handling, administration, and waste.¹

On the imaging-quality side, the program's PET/CT scanner is subject to accreditation requirements — for example ACR PET accreditation and the corresponding physicist performance evaluation — and to hospital accreditation expectations such as those of the Joint Commission. These require a qualified medical physicist's periodic performance evaluation of the scanner, documented QC, and dose-optimization processes. The scientific procedure guidance that shapes acquisition and interpretation includes the current EANM procedure guidelines for brain FDG-PET (version 3, which replaced the 2009 edition), the EANM–EAN clinical recommendations, and the ACR–ASNR practice parameter and ACR Appropriateness Criteria in the United States.¹²³⁴

Of the states DRPS serves, radioactive-material medical use is administered by the NRC in Washington DC and Delaware (both non-Agreement, direct-NRC jurisdictions), while Florida, Maryland, Virginia, California, Nevada, Pennsylvania, New York, and New Jersey administer their own Agreement State programs. A facility must confirm which authority licenses its medical use of byproduct material and align its radiation safety program accordingly, then layer accreditation and physics QC on top. DRPS provides this coordination through medical physics consulting and accreditation support.

Frequently Asked Questions (FAQs)

What does brain FDG-PET actually measure?

Brain FDG-PET measures regional cerebral glucose metabolism. Fluorine-18 fluorodeoxyglucose is taken up in proportion to glucose use and trapped after phosphorylation, so the image reflects synaptic activity. Areas of reduced uptake (hypometabolism) indicate regional neuronal or synaptic dysfunction, which is the basis for both dementia pattern reading and epilepsy localization.

How is patient preparation different for a brain FDG-PET than for an oncology scan?

Brain studies emphasize a controlled resting state. The patient fasts, blood glucose is confirmed to be reasonably controlled, and the FDG is injected in a dim, quiet room with minimal talking, reading, or movement during the uptake period. This reduces sensory and motor cortex activation that could otherwise mimic or mask a disease pattern.

What metabolic pattern suggests Alzheimer disease on FDG-PET?

The classic Alzheimer pattern is bilateral, often asymmetric hypometabolism of the temporoparietal association cortex together with the posterior cingulate and precuneus. The primary sensorimotor cortex, primary visual cortex, basal ganglia, thalamus, and cerebellum are relatively spared, which helps distinguish the pattern from other dementias.

Why is FDG-PET useful in epilepsy?

In focal epilepsy, the epileptogenic zone often shows reduced glucose metabolism between seizures (interictal hypometabolism). FDG-PET can help localize or lateralize the seizure focus, especially in temporal lobe epilepsy and in MRI-negative cases, and is used alongside EEG, MRI, and other tests during presurgical evaluation.

Do I need quantitative software, or is visual reading enough?

Visual reading by an experienced interpreter remains central, but semiquantitative tools such as statistical parametric mapping, 3D stereotactic surface projections, and Z-score comparison to an age-matched normal database improve sensitivity and reproducibility. Guideline panels support using these tools to assist, not replace, visual interpretation.

How much radiation dose does a brain FDG-PET involve?

Using published dose coefficients, F-18 FDG delivers an effective dose on the order of 0.019 mSv per MBq administered, so a typical adult brain study corresponds to roughly 2 to 5 mSv from the radiopharmaceutical, plus any dose from a CT used for attenuation correction. Activity is minimized consistent with adequate image quality.

What quality control most affects a brain FDG-PET result?

Consistency is the key. A stable scanner calibration, a fixed reconstruction protocol, a standardized uptake time and uptake environment, and a normal database acquired on a comparable system all matter, because the reader is comparing a subtle metabolic pattern against expectation. Drift in any of these can shift apparent metabolism and change interpretation.

Key Takeaways

• FDG-PET images synaptic function through glucose metabolism. Hypometabolism marks regional dysfunction, which underlies both dementia pattern reading and epilepsy localization.¹
• Preparation is a physics problem. Fasting, glucose control, a resting uptake environment, and a fixed uptake time protect the validity of a subtle metabolic map.¹
• Patterns carry the differential in dementia. Temporoparietal/posterior cingulate loss suggests Alzheimer disease; frontal/anterior temporal loss suggests frontotemporal dementia; occipital loss with a preserved posterior cingulate suggests dementia with Lewy bodies.¹²
• Interictal FDG-PET localizes seizure foci. It adds the most value in temporal lobe and MRI-negative epilepsy and is strongest when co-registered with MRI and combined with complementary modalities.⁸⁹
• Semiquantitative analysis assists, not replaces, the reader. SPM, 3D-SSP, SUVr, and asymmetry indices improve reproducibility when the reconstruction and normal database are matched.²⁵
• Consistency beats cleverness. Locked reconstruction, controlled motion, and a scanner-matched normal database are what make a borderline read defensible.¹⁵

Conclusion

Brain FDG-PET is deceptively simple to acquire and genuinely demanding to do well. The tracer physics gives a high-contrast map of glucose metabolism, but the clinical answer lives in relative regional pattern — and that pattern is only trustworthy when preparation, uptake environment, reconstruction, and analysis are held constant against a matched normal reference. For dementia, that discipline lets FDG-PET separate overlapping syndromes; for epilepsy, it lets interictal hypometabolism contribute to a surgical plan.

The medical physicist's role is to protect that consistency: to lock the acquisition and reconstruction, to validate the normal database and revalidate it after upgrades, to keep the scanner calibrated and accredited, and to make sure administered activity stays as low as reasonably achievable without starving the count statistics the reader depends on. Done well, brain FDG-PET turns a metabolic map into a decision that changes care.

How DRPS Can Help

Diagnostic Radiation Physics Services helps PET/CT programs build and defend the consistency that neurological FDG-PET demands. That includes scanner performance evaluation and calibration, reconstruction protocol review, normal-database validation after hardware or software changes, dose optimization, and accreditation support for ACR PET and related programs. We also coordinate the radioactive-material and radiation-safety side of a PET program through PET/CT and nuclear medicine physics, medical physics consulting, and radiation safety officer services.

DRPS supports facilities across our service locations, including Florida, Maryland, Virginia, Washington DC, California, Nevada, New York, Pennsylvania, New Jersey, and Delaware.

Related Resources

• Amyloid and tau brain PET
• DaTscan (I-123 ioflupane) SPECT
• Brain perfusion SPECT with HMPAO and ECD
• PET SUV quantification
• PET/CT attenuation correction
• The PET partial volume effect
• PET/CT and nuclear medicine physics services
• Accreditation support

References

1. Guedj E, Varrone A, Boellaard R, et al. EANM procedure guidelines for brain PET imaging using [¹⁸F]FDG, version 3. Eur J Nucl Med Mol Imaging. 2021;49(2):632-651. doi:10.1007/s00259-021-05603-w. doi.org
2. Nobili F, Arbizu J, Bouwman F, et al. European Association of Nuclear Medicine and European Academy of Neurology recommendations for the use of brain ¹⁸F-fluorodeoxyglucose positron emission tomography in neurodegenerative cognitive impairment and dementia: Delphi consensus. Eur J Neurol. 2018;25(10):1201-1217. doi:10.1111/ene.13728. doi.org
3. Frey KA, Lodge MA, Meltzer CC, et al. ACR-ASNR Practice Parameter for Brain PET/CT Imaging Dementia. Clin Nucl Med. 2016;41(2):118-125. doi:10.1097/RLU.0000000000001037. doi.org
4. Moonis G, Subramaniam RM, Trofimova A, et al. ACR Appropriateness Criteria® Dementia. J Am Coll Radiol. 2020;17(5S):S100-S112. doi:10.1016/j.jacr.2020.01.040. doi.org
5. Gallivanone F, Della Rosa PA, Castiglioni I. Statistical Voxel-Based Methods and [¹⁸F]FDG PET Brain Imaging: Frontiers for the Diagnosis of AD. Curr Alzheimer Res. 2016;13(6):682-694. doi:10.2174/1567205013666151116142039. doi.org
6. Boccardi M, Festari C, Altomare D, et al. Assessing FDG-PET diagnostic accuracy studies to develop recommendations for clinical use in dementia. Eur J Nucl Med Mol Imaging. 2018;45(9):1470-1486. doi:10.1007/s00259-018-4024-1. doi.org
7. Walker Z, Gandolfo F, Orini S, et al. Clinical utility of FDG PET in Parkinson's disease and atypical parkinsonism associated with dementia. Eur J Nucl Med Mol Imaging. 2018;45(9):1534-1545. doi:10.1007/s00259-018-4031-2. doi.org
8. Guo K, Wang J, Cui B, et al. [¹⁸F]FDG PET/MRI and magnetoencephalography may improve presurgical localization of temporal lobe epilepsy. Eur Radiol. 2022;32(5):3024-3034. doi:10.1007/s00330-021-08336-4. doi.org
9. Tang Y, Xiao L, Deng C, et al. [¹⁸F]FDG PET metabolic patterns in mesial temporal lobe epilepsy with different pathological types. Eur Radiol. 2024;34(2):887-898. doi:10.1007/s00330-023-10089-1. doi.org
10. Raji CA, Torosyan N, Silverman DHS. Optimizing Use of Neuroimaging Tools in Evaluation of Prodromal Alzheimer's Disease and Related Disorders. J Alzheimers Dis. 2020;77(3):935-947. doi:10.3233/JAD-200487. doi.org
11. International Commission on Radiological Protection. ICRP Publication 128: Radiation Dose to Patients from Radiopharmaceuticals — A Compendium of Current Information Related to Frequently Used Substances. Annals of the ICRP. 2015;44(2S). icrp.org