Knowledge Centre
category guide

How to Choose a CT Scanner

April 30, 2026· 11 min read· AI-generated

How to Choose a CT Scanner

A procurement-literate guide to navigating slice count, photon-counting technology, total cost of ownership, and site readiness — before you sign the capital contract.

What this is and who buys it

Computed tomography scanners are rotating X-ray systems that combine a high-voltage tube, a detector array, and reconstruction software to produce cross-sectional anatomical images in seconds. They are indispensable across trauma, stroke, oncology staging, pulmonary assessment, and coronary artery evaluation — and the range of buyers reflects that breadth. A rural critical-access hospital replacing a 16-slice workhorse, a regional community hospital upgrading to a 128-slice system for higher throughput, and an academic cardiac center evaluating photon-counting CT (PCCT) are all legitimate buyers, but they are effectively shopping in different markets with different decision criteria.

The capital landscape shifted meaningfully in 2024–2026. The Siemens Naeotom Alpha.Prime received FDA 510(k) clearance as the first single-source PCCT scanner [S7], and GE HealthCare's Photonova Spectra — using deep-silicon detector technology — followed with its own 510(k) clearance [S8]. Philips separately received clearance for the Spectral CT Verida, combining always-on spectral imaging with AI reconstruction [S9]. These are not incremental upgrades; they represent a platform change in how CT detectors count and differentiate photon energy. For capital planning purposes, this inflection point means buyers in 2025–2027 face a genuine "bridge or leap" decision that did not exist three years ago.

The buyer universe matters for a different reason too: CT is one of the few imaging modalities where a $90,000 refurbished unit and a $5,000,000 photon-counting dual-source system are both commercially available and clinically defensible in their respective contexts. Understanding which tier you actually need is the most consequential decision this guide addresses.

Key decision factors

Slice and detector row count is the single largest driver of both capability and price. Entry-level 16-slice systems (including refurbished platforms from major OEMs) occupy the $90,000–$130,000 range for used hardware, while 64- and 128-slice mid-range systems start around $400,000 new [S12]. The jump to 256-slice and beyond correlates with temporal resolution improvements meaningful for cardiac work — but only if your scan volume and case mix justify it. Matching detector rows to throughput targets is the discipline most buyers skip.

Single-source versus dual-source architecture is the next structural fork. Single-source scanners — one tube and one detector array — handle the overwhelming majority of clinical workflows at lower capital and service cost. Dual-source configurations (two tube/detector pairs offset at approximately 90 degrees) provide half-rotation temporal resolution around 66–75 ms, which is what enables reliable coronary CTA in patients with high or irregular heart rates, and advanced spectral imaging techniques. Dual-source is justified for tertiary cardiac programs; it adds cost and service complexity everywhere else.

X-ray tube technology and consumable cost deserves more attention than it typically gets in RFPs. Older and lower-cost systems often use ball-bearing tube designs; higher-end platforms since roughly 2005 have used liquid-metal bearing (LMB) tubes, which run quieter and last longer but carry higher replacement costs and more limited aftermarket availability [S14]. Tube lifespan ranges from approximately 100,000–200,000 scan seconds on entry-level systems to 300,000–500,000+ on high-end 128- to 640-slice platforms [S15]. At mid-volume clinical use, that translates to roughly 4–7 years before a tube swap — a cost that needs to be modeled into your TCO from day one, not discovered at year five.

Dose reduction and AI reconstruction software are increasingly both a patient-safety imperative and a reimbursement-adjacent concern. Automatic exposure control, tube current modulation, and iterative reconstruction are now table stakes on mid-tier and above systems. Deep-learning reconstruction — including FDA-cleared platforms like GE's True Definition DL [S10] — reduces image noise at lower dose and is a real differentiator when comparing otherwise similar hardware configurations. If a quote includes hardware but lists these software modules as optional line items, price them in before comparing bids.

Bore size and table capacity determine clinical scope more concretely than any marketing term. Bores range from approximately 50 cm on standard systems to 82 cm on wide-bore platforms; table weight capacity runs from 450 lbs to 677 lbs. Both parameters matter for bariatric patients, trauma workflows, and any interventional or radiation-therapy simulation application. Specifying these upfront avoids the costly discovery that a new scanner cannot accommodate a meaningful share of your patient population.

Site readiness is regularly underestimated. CT gantries are dense — floor loading must be structurally verified, particularly in multi-story buildings. Three-phase power feeds, HVAC for tube cooling (chillers add square footage and cost), lead shielding, and PACS/RIS integration all precede the first scan. A structural or electrical deficiency discovered during installation is a five-to-six-figure schedule problem. Commission a site survey before the purchase order, not after.

Service network density varies meaningfully by geography and brand. The density of trained field engineers and the availability of OEM and aftermarket parts in your region directly affects uptime. Some platforms benefit from deep aftermarket tube ecosystems; others are tightly OEM-dependent. This is a factor to verify by region, not assume from brand reputation.

What it costs

CT scanner pricing spans nearly two orders of magnitude, which makes headline price comparisons almost meaningless without a full bill of materials. The ranges below reflect hardware purchase price; software licenses, installation, shielding, and first-year service are additive and can represent 20–40% of the hardware cost on complex configurations [S11, S12].

  • Entry ($90,000–$350,000): Refurbished 16- to 64-slice systems or new low-end single-source platforms. Appropriate for rural hospitals, imaging centers, and facilities with sub-5,000 annual scan volumes.
  • Mid ($400,000–$750,000): New 64- to 128-slice single-source systems with modern reconstruction software. The most common tier for community hospitals and high-volume outpatient imaging centers.
  • Premium ($1,000,000–$5,000,000+): High-end single-source starts around $1M–$2M; dual-source from approximately $2M; photon-counting dual-source configurations can reach $5M or more. Justified only in high-volume tertiary, academic, or specialized cardiac/oncology programs.

Common use cases

The clinical and operational context should drive tier selection more than vendor preference. Four concrete scenarios illustrate where different configurations fit:

  • Rural and critical-access hospitals: A 16-slice refurbished system handles head, chest, abdomen, and spine imaging for a facility doing 2,000–4,000 scans annually, with modest service contract costs and a manageable tube replacement cycle.
  • Community hospital and outpatient imaging: A new or certified-refurbished 64-slice system with iterative reconstruction supports 8,000–15,000 annual scans, ACR accreditation, and a broad general radiology workflow without the capital or service overhead of dual-source equipment.
  • Tertiary care and cardiac centers: Dual-source and 256+ slice systems for coronary CTA, structural heart planning, and heavily calcified or high-heart-rate anatomy. The clinical case is strong; the ROI depends entirely on scan volume and cardiology program maturity.
  • Academic and research hospitals: PCCT systems, with their material-decomposition and reduced-artifact capabilities, fit here — alongside research software licenses and longer procurement timelines driven by institutional review processes.

Regulatory and compliance

CT scanners are Class II medical devices regulated under 21 CFR § 892.1750, product code JAK, cleared through the FDA's 510(k) Premarket Notification pathway [S1, S2]. Each AI reconstruction add-on or CAD module typically requires its own separate 510(k) clearance under 21 CFR 892.2050/2060/2070. Buyers should verify that every software module included in a quote — particularly AI-based reconstruction and lung or cardiac CAD tools — carries independent FDA clearance and is listed in the device's current cleared labeling.

For billing purposes, ACR CT Accreditation is the dominant CMS-recognized program for advanced diagnostic imaging services; IAC and The Joint Commission are accepted alternatives [S3]. As of January 1, 2026, ACR requires all phantom exams to be performed within 14 months of the testing package release date [S6]. A qualified medical physicist must conduct initial and annual surveys and provide a signed report — this is non-negotiable and should be budgeted as a recurring annual cost. Reference dose limits under ACR standards include 60 mGy CTDIvol for adult head, 35 mGy for adult abdomen, and 25 mGy for pediatric abdomen (5-year-old) [S4]. Applicable equipment standards include IEC 60601-1 (general electrical safety), IEC 60601-2-44 (CT-specific requirements), and NEMA XR-25/XR-29 for dose check and MITA Smart Dose compliance. State radiation control bureaus require separate registration and periodic inspection independent of federal clearance.

Service, training, and total cost of ownership

Installation lead times of 8–16 weeks are typical for new systems once a purchase order is placed. That clock starts after site-prep is complete — structural work, shielding, power, and HVAC can add weeks or months if discovered late. Vendor-provided technologist applications training runs approximately 3–5 days on-site; radiologist orientation is typically shorter but equally necessary. Budget for refresher training whenever major software upgrades are deployed, as reconstruction and protocol interfaces change substantially between generations.

Annual service contract costs range from approximately $52,500 to $147,500 depending on scanner tier and contract scope, with monthly figures running $500 to $7,000 [S13]. Tube replacement is the largest unplanned cost event: a high-end MRC 600/800 tube assembly replacement takes 4–12 hours for the swap plus up to 7 additional hours for purging and recalibration, with used MRC 600 tubes averaging $40,000–$65,000 on the aftermarket [S14]. Aftermarket tube manufacturers including Varex, Dunlee, and Chronos produce tubes matching OEM specifications for many platforms, keeping replacement costs competitive — but availability varies by model, so verify this before signing. The AHRA benchmark for CT replacement planning is 8 years; in practice, mechanically sound 16-slice systems have run 15–20 years [S16]. Clinical obsolescence — driven by dose standards, AI capability gaps, and reimbursement policy — almost always triggers replacement before mechanical failure does.

Red flags to watch for

A headline price without a full bill of materials is the most common source of procurement regret in CT. Two systems with similar hardware specifications can diverge by hundreds of thousands of dollars once cardiac software suites ($35,000–$100,000), lung applications ($15,000–$35,000), installation, shielding, and first-year service are itemized separately. Insist on line-item breakdowns before comparing bids.

For refurbished equipment, any reputable seller should provide a tube usage report showing scan-seconds consumed versus expected life, warranty status, and estimated replacement cost. A seller who cannot or will not produce this document is a meaningful warning. Pair that with a review of the system's service and error logs — patterns of thermal shutdowns, arc errors, or repeated cooling-system calls can signal underlying problems that cosmetic refurbishment does not resolve.

The absence of a defined uptime SLA is a third flag. If a service contract does not specify response time for engineer dispatch, parts availability (same-day shipping matters), and a defined remedy for SLA misses, the buyer absorbs all downtime risk. Finally, verify that the contract does not prohibit third-party ISO access — locked OEM-only service restricts your options and typically increases service costs substantially at years 4–7.

Questions to ask vendors

  1. Provide a full bill of materials with hardware, all software licenses (cardiac, lung, AI reconstruction, dual-energy), installation, applications training, warranty term, and uptime guarantee as separate line items. What is the post-install cost to add each optional software module?
  2. What is the current tube's installed scan-second count, expected remaining life, and OEM versus aftermarket replacement cost? Are Varex-, Dunlee-, or Chronos-compatible tubes available for this model?
  3. What is the contractually guaranteed uptime percentage, on-site engineer response time, and remedy (credits, loaner unit) if the SLA is missed?
  4. Provide the FDA 510(k) clearance numbers for this platform and any open recalls, field corrections, or cybersecurity advisories issued in the past 24 months.
  5. Which dose-reduction technologies (iterative reconstruction, deep-learning reconstruction, organ-based tube current modulation) are standard versus optional, and does the system meet NEMA XR-25 and XR-29 requirements out of the box?
  6. What is the projected total cost of ownership over 7 and 10 years, including tube replacements, software upgrade fees, annual service escalators, and the vendor's committed end-of-life support date for this platform?

Alternatives

The refurbished market deserves serious evaluation for facilities in the entry and mid tiers. As-is units can sell for as little as 25% of original sticker price; certified-refurbished systems with documented tube data and warranty coverage typically run 50–75% of new [S12]. A certified-refurbished 64-slice system is often the highest-value option for community hospitals where scan volume does not justify new mid-tier capital.

On financing, equipment finance agreements and $1-buyout leases suit long-lived, stable assets where you intend to own the equipment for its full service life; fair-market-value leases preserve capital and make technology refresh easier in a period (like now) when detector technology is genuinely evolving. Terms of 3–7 years and interest rates in the 5–6% range are typical depending on credit and market conditions. For the service contract structure, a common and defensible pattern is OEM warranty coverage in years 1–2 — when the OEM holds unique expertise in newly deployed hardware — transitioning to an ISO or multi-vendor service arrangement in years 3–7, when aftermarket knowledge has matured and cost savings become substantial. For facilities doing fewer than approximately 3,000 scans per year, contracted mobile CT routes may defer the capital decision entirely until volume justifies a fixed installation. On PCCT specifically: clinical reimbursement pathways are still maturing, and many experienced buyers are bridging on high-end energy-integrating dual-source systems and re-evaluating PCCT economics in 2027–2028.

Sources

Sources

Browse vendors in

MedSource publishes neutral guidance. We do not accept payment from vendors to influence the content of articles. AI-generated articles are reviewed for factual accuracy but cited sources should be the primary reference for procurement decisions.