QA phantoms

The phantoms/qa/ library encodes the standard nuclear-medicine quality-assurance phantoms, so you can assess an OpenTOPAS-SPECT system the way you would a real gamma camera: tomographic uniformity, spatial resolution, and contrast recovery. Every dimension comes from a manufacturer datasheet or an international standard, not an assumption (provenance in research/knowledge/spect-qa-phantom-geometry/).

What ships

Phantom	Encodes	Test it supports
Jaszczak Deluxe	water cylinder + cold rods + cold spheres	uniformity, resolution, contrast, in one acquisition
NEMA IEC body	hot spheres in a warm body with a cold lung	contrast recovery coefficients
NEMA NU-1 sources	thin line/point sources at 7.5 cm spacing	system spatial resolution (FWHM)

Running one

The phantoms are IncludeFile fragments; the shipped example decks image them end-on through a Siemens Symbia LEHR head (Tc-99m). Run from the example's own folder:

cd examples/qa
/path/to/OpenTOPAS-build/topas jaszczak_projection.txt

Each deck writes a planar projection (EnergyDeposit per crystal pixel, CSV). Apply the 140.5 keV photopeak window in post-processing, and rotate the detector for a tomographic series (see examples/phantom/spect_acquisition.txt).

The Jaszczak phantom

The six cold-rod sectors (finest red to coarsest purple) and six cold spheres, with the water cylinder hidden.

A single water-filled acrylic cylinder (216 mm inner diameter, 186 mm tall, 3.2 mm wall) with three inserts, so one acquisition covers three tests:

• Cold rods — six 60 deg sectors of PMMA rods, one diameter per sector (Deluxe: 4.8 to 12.7 mm), in a hex lattice with center-to-center pitch equal to twice the rod diameter (from Jaszczak's patent). The finest sector still resolved sets the tomographic spatial resolution.
• Cold spheres — six PMMA spheres (9.5 to 31.8 mm) for contrast versus size.
• Uniform section — the plain water gives integral and differential uniformity.

Because the rods and spheres are non-active PMMA sitting in a uniformly active water background, they read as cold voids. Other Jaszczak variants (Ultra Deluxe, Standard, Benchmark) differ only in the rod and sphere diameters and are one flag away:

python3 tools/make_qa_phantom.py jaszczak --model ultradeluxe

The NEMA IEC body phantom

The six fillable spheres (10 to 37 mm) around the central lung insert, body wall hidden.

The IEC 61675-1 / NEMA NU 2-2018 body: an elliptical PMMA cavity (interior 290 mm lateral x 221 mm AP x 193 mm, ~9.7 L) holding six fillable spheres (inner diameters 10, 13, 17, 22, 28, 37 mm, centers 70 mm from the lid) and a cold low-density lung insert (45 mm foam, 0.3 g/cm3). The spheres are hot at a selectable sphere-to-background concentration ratio (default 4:1); measure how much of each sphere's true concentration you recover as a function of its size.

python3 tools/make_qa_phantom.py nema-iec --ratio 8   # 8:1 spheres

The small spheres carry few counts by design (their activity is concentration x volume, and a 10 mm sphere is a tiny fraction of a 9.7 L background) — that is exactly why they are the hard ones to recover. Scale all histories up together to gain statistics without changing the ratio.

NEMA NU-1 resolution sources

Three thin line sources (capillary bore under 1 mm) at 0 and +/-75 mm, the NEMA 7.5 cm spacing. Imaged end-on they appear as point spread functions; the fitted FWHM of each is the system spatial resolution. Add the surrounding 20 cm water cylinder for the with-scatter measurement:

python3 tools/make_qa_phantom.py resolution --in-water

Caveats

• Emission defaults to Tc-99m (140.5 keV). For another isotope, edit So/*/VolumetricEnergy and use a collimator preset matched to its energy.
• Sphere walls (~1 mm PMMA) are not modeled — a small, documented approximation.
• The IEC body cross-section is modeled as an ellipse; the exact IEC contour is a slightly non-elliptical refinement.