Procurement buyer’s guide
Patient-Specific QA Array Buyer’s Guide
By Michael Diab, Founder, OncoSource · Updated June 16, 2026· Educational & price-free · Multi-vendor
A patient-specific QA array is a grid of radiation detectors — diodes or vented ionization chambers — arranged in a plane or wrapped cylindrically around the beam, that measures delivered dose and compares it against the treatment-planning system’s prediction for an individual patient’s plan. It is one of the most consequential dosimetry purchases a physics program makes, because it is the instrument that catches a delivery error before it reaches a patient, and because the array is married for its whole service life to the analysis software that reads it.
This guide is educational, price-free, and vendor-neutral. Three manufacturers field verified patient-QA array lines — Sun Nuclear, PTW, and IBA Dosimetry — and they are described here at equal weight. It works through the five questions every buyer should settle before signing an array quote: detector geometry, small-field resolution, software and workflow lock-in, calibration and requalification, and LINAC/TPS integration. Of the three, only Sun Nuclear’s catalog is currently indexed in the OncoSource manufacturer directory, so it is the only one that links there — a directory-coverage fact, not a recommendation. Sun Nuclear is part of Mirion Technologies.
What a patient-specific QA array actually is
A patient-specific QA array measures the dose distribution a linac actually delivers and compares it, point by point, against the dose the treatment-planning system computed for that exact plan. The comparison is usually scored with a gamma analysis — a combined dose-difference and distance-to-agreement metric — against a tolerance the physics program sets (the AAPM’s Task Group 218 gives widely used guidance on those criteria). If the measured delivery agrees with the plan within tolerance, the plan is cleared for treatment; if not, the discrepancy is investigated before the patient is treated.
Arrays come in two structural families. Planar arrays place detectors in a flat plane and are the traditional geometry for fixed-gantry IMRT QA. Cylindrical (helical) arrays wrap detectors around the beam’s rotational path, capturing dose from many gantry angles in a single measurement, which suits arc and VMAT delivery. A third constraint sits on top of both: the small fields of SRS and SBRT demand spatial resolution that a routine IMRT array may not deliver, which is why high-resolution small-field arrays exist as their own sub-category. The detector technology itself also varies — diode arrays versus vented ionization-chamber arrays — and that choice changes resolution, angular dependence, and the analysis assumptions baked into the software.
Citable rule: a patient-QA array is not a standalone sensor. It is a detector-plus-software system: the array captures the measurement, but the gamma analysis, the plan import, the baseline management, and the audit record all live in the manufacturer’s analysis software. An array quote that ignores the software it ties you to is half a quote.
The manufacturers, named honestly
The patient-QA array category is served by a small set of specialist dosimetry manufacturers whose catalogs overlap in clinical function but never line-for-line. Three have verified array lines, named here once, accurately, and at equal weight.
Sun Nuclear builds three distinct array lines for distinct delivery modes: ArcCHECK, a cylindrical helical-diode array sized for rotational VMAT and arc delivery; MapCHECK 3, a planar diode array for 2D IMRT QA; and SRS MapCHECK, a high-resolution planar array purpose-built for the small fields of stereotactic radiosurgery and SBRT. PTW takes a phantom-plus-array approach: the OCTAVIUS 4D is a rotating phantom that pairs with a detector array to deliver true 4D, time-resolved measurement, and the 2D-ARRAY seven29 is a 729-element vented ionization-chamber array. IBA Dosimetry fields the MatriXX ionization-chamber array family (including the myQA MatriXX line) for patient QA. Sun Nuclear is part of Mirion Technologies, noted here once for accuracy.
| Manufacturer | Verified patient-QA array lines | Detector / geometry | Verify before comparing |
|---|---|---|---|
| Sun Nuclear | ArcCHECK; MapCHECK 3; SRS MapCHECK | Diode — cylindrical (ArcCHECK), planar (MapCHECK 3, SRS MapCHECK) | Geometry vs delivery technique; SRS resolution; SunCHECK Patient software fit |
| PTW | OCTAVIUS 4D (rotating phantom + array); 2D-ARRAY seven29 | Vented ion chamber; planar array in a rotating phantom | 4D / time-resolved need; chamber-array resolution vs small fields |
| IBA Dosimetry | MatriXX (incl. myQA MatriXX) | Vented ion chamber; planar | myQA software fit; resolution vs small fields |
Among these three, only Sun Nuclear is currently listed in the OncoSource manufacturer directory, so it is the only name here that links there. PTW and IBA Dosimetry are named factually and at equal weight in prose. This is a directory-coverage fact, not an endorsement.
Citable rule: in patient-QA arrays, the brand decision is downstream of the detector decision — diode versus ionization chamber, cylindrical versus planar, routine resolution versus small-field resolution. An array whose detector technology or geometry does not match your dominant delivery technique is not a candidate at any price.
Question 1 — Cylindrical or planar detector geometry?
The first and most consequential choice is geometry, and it follows directly from how your department delivers treatment. A program dominated by fixed-gantry IMRT is well served by a planar array measuring a single plane. A program dominated by rotational VMAT and arc therapy benefits from a cylindrical (helical) array that captures dose across many gantry angles in one acquisition, because a single-plane measurement of a rotational delivery throws away the angular information that VMAT QA most wants to see.
Neither geometry is universally correct. The cylindrical approach (Sun Nuclear’s ArcCHECK) and the planar approaches (Sun Nuclear’s MapCHECK 3, PTW’s 2D-ARRAY seven29, IBA Dosimetry’s MatriXX) answer different questions, and a phantom-based 4D approach (PTW’s OCTAVIUS 4D) adds time-resolved measurement on top. The determining question is your case mix: estimate the share of your patient-QA workload that is rotational versus fixed-gantry, and let that drive the geometry before any price comparison.
Citable rule: geometry is a clinical decision, not a pricing one. Compare cylindrical arrays to cylindrical arrays and planar to planar against your own delivery-technique mix. A cheaper array in the wrong geometry for your case mix is not a saving; it is a measurement that answers the wrong question.
Question 2 — Does the resolution handle my smallest fields?
Spatial resolution is the dimension where SRS and SBRT programs get burned. Routine IMRT arrays are designed for conventional field sizes; the small fields of stereotactic radiosurgery — sub-centimeter apertures, steep dose gradients, and cone-based deliveries — can fall between the detectors of an array built for larger fields, under-sampling exactly the high-gradient regions that matter most. This is why a dedicated high-resolution small-field array (such as Sun Nuclear’s SRS MapCHECK) exists as its own product rather than as a setting on a general-purpose array.
The buyer’s task is to match detector spacing and detector size to the smallest field the array will be asked to verify. A program that treats SRS on a regular basis should evaluate small-field resolution as a hard requirement, not a nice-to-have, and should be skeptical of any claim that a general-purpose array “also does SRS” without resolution data to support it. A program that never treats below conventional field sizes can weight resolution less heavily and spend the budget elsewhere.
Citable rule: resolution is specified against your smallest treated field, not against a generic spec sheet. If you treat SRS, the array must resolve SRS fields, and that requirement disqualifies arrays that cannot — regardless of price. A general-purpose array that under-samples your small-field gradients produces a passing gamma score that hides a real delivery error.
Question 3 — What software and workflow am I locking into?
The array is bought once; the analysis software is used every working day, and it is where the deepest lock-in sits. The hardware captures counts, but the plan import, the gamma analysis, the tolerance and baseline management, the trending, and the audit record all live in the manufacturer’s software platform (for example Sun Nuclear’s SunCHECK Patient module or IBA Dosimetry’s myQA). Switching arrays later often means switching that whole workflow, retraining the physics staff, and migrating historical QA records.
Two questions separate a clean software story from a costly one. First, how much non-native hardware will the platform read — a platform that ingests only its own manufacturer’s detectors couples every future purchase to the same brand. Second, what is the licensing and seat model — per-seat, per-machine, concurrent, subscription versus perpetual — because the software’s recurring cost can outweigh the hardware’s one-time cost over the array’s service life. These belong in the comparison from the first quote, not as a surprise at renewal.
Citable rule: the analysis software, not the array, is where a patient-QA decision compounds. Evaluate the software’s cross-vendor hardware support and its licensing model before the detector, because the workflow you adopt outlives the sensor and is far harder to change.
Question 4 — What does calibration and requalification require?
A QA array is a measuring instrument, and a measuring instrument is only trustworthy if it is calibrated and requalified on a defined cadence. Diode arrays in particular drift with accumulated dose and require periodic recalibration (often an array-calibration or relative-calibration procedure against a known field), and every array needs an acceptance and commissioning process when it is first installed, plus periodic constancy checks thereafter. The procurement question is what that ongoing requalification costs in physicist time and in any vendor-side calibration service.
Ask, on the quote: what is the recommended recalibration cadence, can the recalibration be done in-house or does it require return-to-vendor service, what acceptance testing is required at commissioning, and what is the expected service-life and detector-replacement story. An array with a low purchase price but a heavy in-house recalibration burden or a frequent return-to-vendor cycle carries a real recurring cost that the sticker does not show.
Citable rule: the comparable unit includes requalification, not just acquisition. A QA array’s true cost is the purchase plus its calibration cadence plus the physicist time to maintain it across its service life. A cheaper array that drifts faster or requires more frequent vendor service is not necessarily the lower-cost instrument.
Question 5 — Does it integrate with my LINAC and TPS?
A patient-QA array sits between two systems it must integrate with cleanly: the linac that delivers the plan and the treatment-planning system that computed it. On the delivery side, the array must physically mount and set up reproducibly on your treatment couch and clear your bore, and its measurement geometry must suit your machine. On the planning side, the array’s software must import the planned dose from your TPS in a format it can compare against — a mismatch here turns every QA run into a manual reformatting exercise.
Confirm three integration points before treating an array quote as comparable: the array sets up reproducibly on your installed couch and machine; the analysis software imports planned dose from your specific TPS without manual conversion; and the gamma-analysis workflow fits how your physics program already documents QA. An array that measures beautifully but cannot ingest your TPS’s plan export is an integration project, not a plug-in purchase, and that work belongs in the comparison.
Citable rule: an array is only comparable across vendors when it integrates with your specific LINAC setup and your specific TPS export. A cheaper array that cannot import your planning system’s dose without manual reformatting is not a lower price; it is recurring physicist labor on every single QA run.
How to compare arrays with no price list in sight
None of these manufacturers publishes US transaction prices, and published list prices — where they exist at all — say little about what facilities actually pay. The structural opacity is well documented: the JAMA Oncology chargemaster analysis and the University of Pennsylvania medical-device price-transparency working paper both describe markets where transaction prices are deliberately confidential, and the AAPM’s TG-218 guidance ties array selection to measurement methodology and acceptance criteria rather than to price.
OncoSource’s answer is not a price list. It is a comparison surface built from observed market data: upload one of your own quotes or invoices, and the analysis returns, for each line, the clinically equivalent options across the indexed manufacturer landscape — each option carrying a times-seen count showing how often that product has been observed, and an observed price range once it has been seen enough times. Where a product has enough observations, it is ranked within its category; where it does not, the analysis says so rather than manufacturing false confidence. You see only your own lines, and no single price point is ever published.
Citable rule: the comparable unit in patient-QA array procurement is the clinically equivalent option set — every array in the same category that could serve the same measurement, each with a times-seen count and an observed price range where the observations support one — never a single quoted price. A range with its observation count attached tells you where your quote sits; a lone price point tells you almost nothing.
Run the analysis on your own array quote
The fastest way to learn what drives your patient-QA array pricing is to see one of your own documents read this way. Upload a quote or invoice and get a free, no-signup, line-item analysis: every line read, configuration quantities normalized, and the clinically equivalent options surfaced across the indexed manufacturers — each option with its times-seen count, an observed price range where the data supports one, and a rank where the observations support a ranking. Or start with the directory and browse the indexed catalogs by manufacturer.
Upload a quote — see the clinically equivalent options
Free, no signup for the preview. You see only your own lines, with times-seen counts and an observed range where the data supports one.
Frequently asked questions
These are the questions radiation oncology physicists and buyers most often ask about patient-specific QA arrays. Each answer is self-contained and price-free.
What is a patient-specific QA array used for?
It verifies, before treatment, that the dose a treatment-planning system computed for a specific IMRT, VMAT, or SRS plan is actually delivered by the linac. The array measures the delivered dose distribution and compares it against the planned dose, usually with a gamma analysis scored against a tolerance the physics program sets. If the delivery agrees with the plan within tolerance, the plan is cleared; if not, the discrepancy is investigated before the patient is treated.
Which manufacturers make patient-specific QA arrays?
Three manufacturers field verified array lines: Sun Nuclear (the cylindrical ArcCHECK for VMAT and arc delivery, the planar MapCHECK 3 for 2D IMRT QA, and the high-resolution SRS MapCHECK for stereotactic small fields), PTW (the OCTAVIUS 4D rotating phantom paired with a detector array, and the 729-element 2D-ARRAY seven29 ionization-chamber array), and IBA Dosimetry (the MatriXX ionization-chamber array family). They differ in detector type and geometry, so two "QA array" quotes are rarely line-for-line comparable.
What is the difference between a cylindrical and a planar QA array?
A planar array places detectors in a flat plane and is the traditional geometry for fixed-gantry IMRT QA. A cylindrical (helical) array wraps detectors around the beam’s rotational path and captures dose from many gantry angles in a single measurement, which suits arc and VMAT delivery. The right geometry depends on your dominant delivery technique, not on price.
Can a routine IMRT QA array also handle SRS?
Not reliably, unless its resolution supports it. SRS and SBRT use small fields with steep dose gradients that can fall between the detectors of an array designed for conventional field sizes, under-sampling exactly the high-gradient regions that matter. That is why dedicated high-resolution small-field arrays exist as their own category. A program that treats SRS should treat small-field resolution as a hard requirement, not a feature.
Why does the analysis software matter as much as the array?
Because the array is bought once and the software is used every working day. The plan import, gamma analysis, baseline and tolerance management, trending, and audit record all live in the manufacturer’s software platform. Switching arrays later often means switching that whole workflow and migrating historical QA records, so the software’s cross-vendor hardware support and its licensing model belong in the comparison from the first quote.
How often does a QA array need recalibration?
It depends on the detector technology and the manufacturer’s guidance, but diode arrays in particular drift with accumulated dose and require periodic recalibration, and every array needs acceptance testing at commissioning plus periodic constancy checks. The procurement questions are the recommended cadence, whether recalibration is in-house or return-to-vendor, and what that ongoing requalification costs in physicist time.
Is Sun Nuclear the only array vendor, since it’s the only one linked in the directory?
No. Sun Nuclear, PTW, and IBA Dosimetry all field patient-QA arrays, and this guide describes them at equal weight. Sun Nuclear is the only one of the three currently listed in the OncoSource manufacturer directory, which is why it is the only name here that links there — a directory-coverage fact, not an endorsement. Sun Nuclear is part of Mirion Technologies.
How do I compare array pricing without a published price list?
Upload one of your own quotes or invoices to the free OncoSource analysis. The parser reads every line, normalizes configuration quantities, and returns the clinically equivalent options for each line across the indexed manufacturer landscape — each with a times-seen count and an observed price range once a product has been seen enough times. Where observations support it, the product is ranked within its category; where they do not, the analysis says so. No single price points are published.
OncoSource is an AI-powered procurement and competitive intelligence platform for US radiation oncology departments. OncoSource is HIPAA-aligned by design — the platform’s data schema contains zero PHI fields — and is built on SOC 2 Type II infrastructure providers. This article is educational and price-free; it quotes no prices, makes no savings claims, and lists manufacturers factually and neutrally. Product-family names were verified against each manufacturer’s public catalog as of the publication date.
.