PET Imaging Spatial Resolution & FOV Simulator

An interactive calculator for PET (Positron Emission Tomography) spatial resolution, transaxial FOV, axial FOV, and detector solid angle. Vary ring diameter, crystal size, isotope (¹⁸F-FDG, ¹¹C, ⁸²Rb, ⁶⁸Ga), injected dose, and scan time to see how the three resolution components (crystal pitch, positron range, photon acolinearity) combine and how Long Axial FOV PET boosts sensitivity.

Parameters

Scanner configuration

Preset axial FOV and sensitivity

Ring diameter D

Detector ring inner diameter. Transaxial FOV ≈ 0.7·D

Crystals per ring

Number of crystal elements per ring

Crystal size

Crystal pitch d. Intrinsic resolution ≈ d/2

Radioisotope

Half-life and positron range auto-set

Injected dose

MBq

Adult ¹⁸F-FDG standard: 185–370 MBq

Scan duration

min

Acquisition time per bed position

Results

—

Spatial resolution (mm)

—

Transaxial FOV (mm)

—

Axial FOV (mm)

—

Angular resolution (deg)

—

Detector solid angle (sr)

—

Effective dose (mSv)

—

PET ring · positron decay · 511 keV photon pair

The positron drifts ~range mm before annihilating with an electron, emitting two 511 keV photons 180° ± 0.25° apart. When a crystal pair detects them in coincidence, a Line of Response (LOR) is formed.

Resolution contribution by isotope

Scanner comparison — axial FOV and relative sensitivity

Theory & Key Formulas

$$R_{\text{tot}} = \sqrt{R_{\text{int}}^{2} + R_{\text{range}}^{2} + R_{\text{acol}}^{2}}, \quad R_{\text{int}} = d/2, \quad R_{\text{acol}} = 0.0022\,D$$

Synthesized FWHM resolution. d: crystal pitch, D: ring diameter, R_range: positron range (¹⁸F 0.6 / ¹¹C 1.1 / ⁶⁸Ga 2.9 / ⁸²Rb 5.9 mm).

$$\text{FOV}_{\text{trans}} \approx 0.7\,D, \quad \Omega \propto (\text{FOV}_{\text{trans}}/D)^{2}, \quad N(t) = N_0\,(1/2)^{t/T_{1/2}}$$

Transaxial FOV, detector solid angle, and half-life decay. Long Axial FOV PET (axial ≈ 1 m) gives a 4–5× larger effective solid angle.

$$\text{SNR} \propto \sqrt{N_{\text{total}}}, \quad E_{\text{eff}}[\text{mSv}] \approx 0.019 \cdot A_0[\text{MBq}]$$

SNR follows the √N law of counts. ¹⁸F-FDG effective dose coefficient is about 0.019 mSv/MBq (adult, ICRP-103).

PET Imaging Spatial Resolution & FOV — 18F-FDG

🙋

PET images always look kind of grainy compared to CT. Couldn't we just make the crystals smaller to get a sharper image?

🎓

Good question. PET resolution is actually a quadrature sum of three blurring terms, and crystals alone can't beat the floor. First, the crystal pitch d gives an intrinsic resolution of about d/2. Second is positron range — the positron drifts 0.5 to 6 mm inside the body before annihilating with an electron. Third is acolinearity of the two 511 keV photons: they're not exactly 180° apart, they wobble by ±0.25°, and the larger the ring diameter D the more that wobble blurs the image. Because these add in quadrature, shrinking the crystal from 3 mm to 1.5 mm hardly improves the combined resolution.

🙋

So when I switch isotopes, that "positron range" jumps a lot. ⁸²Rb went above 6 mm. Why even use it if it's so much worse than ¹⁸F or ¹¹C?

🎓

Right, ⁸²Rb has a positron range of about 5.9 mm, so its spatial resolution is worse. But ⁸²Rb has a 1.3 minute half-life and comes from a Sr-82/Rb-82 generator that sits next to the scanner — no cyclotron needed. That's why myocardial perfusion PET often uses ⁸²Rb even in community hospitals. ⁶⁸Ga (range 2.9 mm, half-life 68 min) is also generator-produced and has become the standard for PSMA (prostate cancer) and DOTATATE (neuroendocrine tumor) imaging. ¹¹C has excellent range (1.1 mm) but its 20 min half-life forces in-house cyclotron production. So you trade resolution for logistics and biology.

🙋

I keep coming back to this Long Axial FOV preset. The detector solid angle and axial FOV change dramatically. What's the big deal?

🎓

It's the biggest PET hardware breakthrough in 30 years. Conventional PET has 15–25 cm axial FOV, so a whole-body scan stitches 5–7 bed positions together. Long Axial FOV scanners like uEXPLORER (UC Davis/United Imaging) or Siemens Quadra have ≥1 m axial FOV, capturing from head to mid-thigh simultaneously. The detector solid angle goes up by 30–40×, which means you can drop the injected dose by 10× or the scan time by 20× while keeping the same image quality. That enables pediatric PET with minimal dose, full-body dynamic imaging for drug pharmacokinetics, and ultra-fast oncology scans. It's a real game-changer.

🙋

How do you balance dose and scan time? Cranking both up gives better SNR, but also more radiation, right?

🎓

That's the heart of PET protocol design. Image SNR scales as √N where N is total counts, and N = dose × solid angle × scan time × decay-remaining. So to double SNR you need 4× total counts — quadruple the dose, quadruple the time, or quadruple the sensitivity. Standard adult ¹⁸F-FDG is 185–370 MBq with 15–30 min scan. When Long Axial FOV PET multiplies sensitivity by 30, you can reduce dose to 1/10 and time to 1/3 and still get the same SNR. Effective dose drops 10× too. This tool reports effective dose using the 0.019 mSv/MBq coefficient — 370 MBq ≈ 7 mSv, comparable to a chest CT.

🙋

One more question. PET resolution is 2–3 mm, but CT is 0.5 mm. So why use PET at all?

🎓

PET doesn't image anatomy — it images metabolism and molecular function. CT and MRI capture structure at high resolution, but PET sees glucose uptake (FDG), amino acid synthesis (C-methionine), or receptor binding (PSMA, DOTATATE). A tumor may look identical to surrounding tissue on CT, yet light up brightly on PET because of its altered metabolism. That's why modern clinical practice uses hybrid PET/CT or PET/MRI — PET for function, CT/MRI for anatomy. A 2–3 mm resolution is more than enough for molecular imaging; if you need finer anatomical detail, you read the CT side of the same scan.

Frequently asked questions

The intrinsic PET resolution is governed by the detecting crystal pair. Even for an ideal point source, the triangular response between crystals of pitch d has an FWHM of about d/2. Two physical blurring terms are added in quadrature: positron range (0.6 mm for ¹⁸F, 5.9 mm for ⁸²Rb) and acolinearity of the 511 keV photon pair (0.0022·D mm for ring diameter D). With D=900 mm, ¹⁸F-FDG, and a 3 mm crystal, the synthesized resolution is √(1.5²+0.6²+1.98²) ≈ 2.6 mm — far from the crystal half-pitch of 1.5 mm. Making crystals smaller helps only up to this physical floor.

Standard PET has 15–25 cm axial FOV, so whole-body scans require 5–7 bed positions. Long Axial FOV PET scanners (uEXPLORER, Siemens Quadra) feature about 1 m axial FOV, imaging the entire torso in one bed position. The detector solid angle increases by 30–40×, allowing the same image quality with 1/10 of the injected dose or 1/20 of the scan time. This simulator's "Long Axial FOV" preset sets axial FOV to 1060 mm and you can directly see the difference in solid angle and total counts.

PET image SNR scales as the square root of total counts (the √N law). Total counts = injected dose × solid angle × scan time × decay-remaining fraction. ¹⁸F-FDG (110 min half-life) retains 88% of the dose after 20 minutes, but ⁸²Rb (1.3 min half-life) drops below 50% in 2 minutes. Halving the dose and doubling the scan time preserves the same total counts while halving patient dose. Long Axial FOV PET dramatically shifts this trade-off in favor of the patient.

Transaxial FOV is roughly 70% of ring diameter D. Whole-body PET needs D=700–900 mm to cover shoulder and abdomen, giving a transaxial FOV of 500–630 mm. Brain PET only needs to fit a head, so D=300–400 mm is sufficient. The smaller ring reduces acolinearity blur (0.0022·D is smaller), improving spatial resolution. Preclinical small-animal PET pushes this further with D ≤ 200 mm and 1 mm crystals to achieve sub-millimeter resolution.

Real-world applications

Oncology staging (FDG-PET/CT): ¹⁸F-FDG is the most widely used PET tracer, imaging the elevated glucose metabolism of malignant tumors. Lung cancer, lymphoma, colorectal cancer, and head-and-neck cancer staging, treatment response assessment, and recurrence detection rely on FDG-PET as the standard modality, with millions of scans performed globally each year. This simulator's defaults (370 MBq, 20 min) reflect a typical adult protocol. On Long Axial FOV PET, the same diagnostic quality has been reported at 100 MBq.

Myocardial perfusion PET (⁸²Rb, ¹³N-NH₃): PET myocardial perfusion uniquely allows absolute quantification of blood flow at rest and under stress. ⁸²Rb's 1.3 min half-life and Sr-82/Rb-82 generator-on-site supply enable repeat imaging without a cyclotron, making it accessible to community hospitals. Spatial resolution is 6–7 mm — coarse but sufficient for whole-myocardium flow quantification — and it is rapidly displacing SPECT in cardiac imaging.

Neuroendocrine tumor and prostate cancer (⁶⁸Ga-DOTATATE / ⁶⁸Ga-PSMA): ⁶⁸Ga is generator-produced with a 68 min half-life. DOTATATE targets somatostatin receptors on neuroendocrine tumors, and PSMA targets prostate-specific membrane antigen on prostate cancer cells. These tracers reveal disease even at very low PSA values (0.2–1.0 ng/mL) of biochemical recurrence, directly guiding radiation therapy or surgical planning. Spatial resolution of 3–4 mm is required, so they are paired with high-resolution scanners using 3 mm or smaller crystals.

Drug discovery and translational research (microPET): Preclinical small-animal PET for mice and rats uses ring diameters under 200 mm and 1–2 mm crystals to reach sub-millimeter resolution. It enables non-invasive, longitudinal measurement of drug pharmacokinetics, receptor occupancy, and tumor metabolism — a standard tool in early drug development. The "Preclinical mouse PET" preset in this tool shows how shrinking the ring and crystals dramatically improves resolution.

Common pitfalls

The biggest pitfall is confusing reconstruction voxel size with spatial resolution. Reconstructions typically use 256×256 or 400×400 matrices, giving voxels of 1–2 mm. But the actual spatial resolution computed by this tool is the 2–6 mm FWHM, and structures below this scale cannot be resolved no matter how fine the voxels are. Small lesions suffer partial volume effects, with SUV values underestimated by up to 50% when lesion diameter is less than twice the resolution. Always check the lesion-size-to-resolution ratio when interpreting SUV.

Second, "more dose always means better image" is not true. Total counts do increase, but at higher count rates detector dead time and random coincidences degrade effective counts and SNR. ¹⁸F-FDG above 555 MBq pushes most scanners past their count rate efficiency peak. Long Axial FOV PET, with its much larger solid angle, reaches this regime even at modest dose — which is why low-dose protocols are explicitly recommended. Dose should be set by the scanner's count rate characteristics and patient dose balance, not maximized.

Finally, ignoring Time-of-Flight (TOF) when judging resolution and SNR. Modern PET scanners measure photon arrival time differences to 200–400 ps, localizing the emission point along each LOR. This boosts effective SNR by √(D/Δx) ≈ 2–3×, equivalent to a 4× dose reduction. This tool's formulas do not include the TOF gain — its values represent the theoretical floor without TOF. The clinical sweet spot today is Long Axial FOV PET combined with TOF reconstruction, which together define state-of-the-art PET.

PET Imaging Spatial Resolution & FOV Simulator

PET Imaging Spatial Resolution & FOV — 18F-FDG

Frequently asked questions

Real-world applications

Common pitfalls

How to Use

Worked Example

Practical Notes