GE HealthCare Technologies Inc. (Nasdaq: GEHC): Deep-Dive Research Report

Calendar fiscal year; quarterly reporting. Q3 2026 results came out on October 28, 2026, and are the most recent period covered here.

1. What the Company Does

GE HealthCare makes the machines hospitals use to see inside the body: CT, MRI, PET and SPECT scanners, X-ray, mammography and ultrasound. It also makes the drugs injected into patients so those machines can see more clearly: iodinated contrast for CT, MRI contrast, and short-lived radioactive tracers for PET. A third, smaller business makes the bedside equipment that watches over sick patients: monitors, anesthesia machines, ventilators and maternal-infant care systems. Around all of this sits a large service business that keeps the installed fleet running, plus a growing software layer that stores, routes and interprets the images.

How it got here. The business started as an X-ray maker. Its roots go back to the Victor Electric Company, founded in 1893 in a basement, which moved into X-ray machines after Röntgen's discovery. In 1920, GE HealthCare entered the imaging technology business with the acquisition of Victor Electric Company. Within GE the unit grew into CT and MRI. In the 1970s, CT scanners were introduced, followed by MRI machines in the 1980s, and GE became a major manufacturer of both. Three later deals shaped today's company:

• in 1998, GE Medical Systems bought Marquette Electronics, the origin of today's monitoring business.
• In 2004, GE acquired Amersham plc for $9.5 billion and formed GE Healthcare. This turned a hardware company into one that also sells consumable diagnostic drugs.
• In 2020 GE did not spin the unit off and opted instead to simply sell off its biopharma manufacturing business to Danaher in a $21.4 billion deal, which was relaunched as Cytiva.

What was left became the standalone company. On January 3, 2023, GE completed the spin-off through a distribution of approximately 80.1% of the Company's outstanding common stock to holders of GE's common stock, which resulted in the issuance of approximately 454 million shares.

Since the spin, the company has been working to become less of a box-seller. The 2024 annual report frames this as evolution from an imaging and critical care equipment company to a healthcare solutions provider. In practice that has meant three things: buying out its Japanese radiopharmacy joint venture, buying a cloud imaging-software company, and agreeing to buy a U.S. PET-tracer distribution network (Sections 3 and 7). Today it employs approximately 54,000 people globally.

The problem it solves. A hospital needs to diagnose disease quickly, accurately and at a cost it can afford, using equipment that will run every day for a decade or more. GE HealthCare sells three things to meet that need:

• The scanner, a 10+ year capital asset.
• The agent, a per-procedure consumable bought again and again.
• The service and software that keep both productive.

Few rivals sell all three at scale (Section 5).

Why it is hard to make. A modern scanner combines high-voltage physics (X-ray tubes, superconducting magnets), detector semiconductors, real-time reconstruction computing and regulated clinical software. Each modality needs its own regulatory clearance. The drug side adds a different problem: chemistry and logistics. Iodinated contrast starts as iodine mined in Chile and ends as a sterile injectable. PET tracers decay within hours of being made, so they must be produced close to patients and delivered on time. The detail is in Section 3.

A concrete example. Take a patient with chest pain in a U.S. health system. The traditional workup is a SPECT nuclear stress test, which has accuracy limits. With GE's cardiac PET tracer Flyrcado, the hospital can run a PET myocardial perfusion scan instead. The drug was approved as the first FDA-approved F 18 PET myocardial perfusion imaging (MPI) agent for coronary artery disease (CAD) detection, offering higher diagnostic accuracy compared to the more commonly used SPECT agent. Because of its roughly two-hour half-life, the radiotracer can be produced off-site and distributed widely to hospitals and imaging centers, removing the need for on-site manufacturing.

In that one workflow, the hospital may scan the patient on a GE PET/CT, use GE iodinated contrast if a coronary CT follows, store the images on GE-owned enterprise imaging software, and keep the scanner under a GE service contract. GE earns money at every step, and the recurring pieces (drug doses and service) continue long after the capital sale.

2. Business Segments

In April 2026 the company folded its four segments into three. It combined its Imaging and AVS businesses into a new operating and reportable segment, Advanced Imaging Solutions ("AIS"), leaving three reportable segments: AIS, PCS, and PDx. The FY2025 mix below has been recast to the new structure.

Mix derived from recast FY2025 segment revenues: AIS 14.60B, PDx 2.90B, PCS 3.09B.

Advanced Imaging Solutions (~71%)

What it does. AIS sells scanners and everything attached to them. AIS includes an AI-enabled portfolio of MR, CT, Molecular Imaging, Women's Health, X-ray, Ultrasound, Image-guided solutions, and enterprise imaging. Customers are hospital radiology, cardiology and surgery departments, outpatient imaging centres and, for ultrasound, obstetric and primary-care practices. The segment covers three core areas: Radiology, Specialized Ultrasound, and Procedural Guidance, and it now includes Solutions for Enterprise Imaging, a portfolio of cloud-enabled, AI-powered software that helps healthcare organizations manage and integrate medical imaging and related clinical data.

Core capability. GE knows how to design and build the hardest physics in medicine at volume: magnets, X-ray tubes, detector stacks. It also has a global field-service network to keep that equipment running. A 2022 spin-off filing described one of the industry's largest global installed bases of medical imaging equipment with approximately 400,000 systems. That installed base feeds upgrades, service and replacement sales.

Why it exists in this form. The merger put GE HealthCare's two largest segments together. The logic is that diagnostic imaging, interventional guidance and follow-up imaging share customers and technology, so one commercial organisation can sell across them. Phil Rackliffe was named CEO of AIS; prior to leading AVS, Phil was President and CEO of Image Guided Therapies. The previous Imaging head, Roland Rott, will leave GEHC to pursue new opportunities.

Fit in the group. AIS is the base everything else builds on. Scanners create demand for PDx agents, and the installed base creates service revenue.

Main rivals: Siemens Healthineers, Philips, Canon Medical, United Imaging.

Pharmaceutical Diagnostics (~14%)

What it does. PDx sells the drugs injected during scans:

• Iodinated contrast for CT and X-ray.
• Gadolinium-based contrast for MRI.
• Radiopharmaceuticals for PET and SPECT, including Flyrcado for the heart, Vizamyl for amyloid in the brain and Cerianna for breast cancer.

Customers are hospital pharmacies, radiology departments and radiopharmacies. Buying is per procedure and repeats constantly.

Core capability. PDx combines industrial-scale chemistry (turning iodine into high-purity active ingredients) with sterile fill-finish and time-critical radiopharmaceutical logistics. Management describes the contrast franchise as one where we win through our focus on supply chain resiliency and delivering the diversity of SKUs our customers require.

Why separate. Its economics, regulation (drug approval, not device clearance) and acquisition history (Amersham, then Nihon Medi-Physics) are entirely different from hardware. On March 31, 2025, the Company acquired the remaining 50% interest in Nihon Medi-Physics Co., Ltd., giving it a wholly owned Japanese radiopharmacy business.

Fit in the group. PDx is the group's most profitable segment by margin and its fastest grower. The PDx CEO said radiopharmaceuticals are "the fastest-growing segment of our business today". CEO Arduini put the pitch this way: PDX pairs resilient, recurring revenue today with a pipeline and footprint to extend that growth.

Main rivals: Bayer and Bracco in contrast media; Lantheus in radiopharmaceuticals.

Patient Care Solutions (~15%)

What it does. PCS develops comprehensive health monitoring equipment, anesthesia delivery systems, ventilators, diagnostic ECGs, and maternal-infant care systems. Customers are ICUs, operating rooms, emergency departments and labour wards. Buying decisions sit with clinical engineering and nursing leadership rather than radiology.

Core capability. PCS has proprietary monitoring parameters and an installed base of networked monitors. The 2024 annual report described it as a provider of medical devices, proprietary parameters and consumables, services.

Why separate. It has a different buyer, a different technology base (embedded electronics rather than imaging physics) and a different competitor set.

Fit in the group. PCS is the turnaround case, and possibly a divestiture. In mid-2026 management said while we are focused on returning the business to growth and profitability, we are reviewing strategic options to maximize its long-term value. The review covers "continued ownership, a sale and other value-enhancing transactions." The operational problems are covered in Section 3 and the review in Sections 7 and 8.

Main rivals: Philips and Mindray.

Geographic mix

The company sells in over 160 countries. FY2025 revenue split roughly as follows: North America: $9.5B (46%), EMEA: $5.4B (26%), China: $2.3B (11%), Rest of World: $3.4B (17%).

• United States and Canada is the largest and most profitable region. It anchors installed-base service and is the launch market for new radiopharmaceuticals, because U.S. reimbursement and PET infrastructure are most developed there.
• EMEA has been a steady contributor; EMEA revenues were $5,425 million, growing 7.4% in 2025. European demand is driven by public-hospital tenders and replacement cycles.
• China is the problem region. It has a long GE history and large local manufacturing (Section 3), but revenue fell 15% in 2024 and a further 4.6% in 2025, when China region revenues were $2,251 million, decreasing 4.6% or $108 million with declines in Imaging, AVS, and PCS revenues partially offset by growth in PDx revenues. In Q3 2026 China grew again, though management said a majority of this performance was driven by foreign exchange impacts. The structural reasons (tenders, volume-based procurement, local champions) are in Sections 6 and 8.
• Rest of World (Japan, India, Latin America, Middle East outside EMEA reporting) is a diversified middle tier. The Nihon Medi-Physics buyout made Japan a meaningful radiopharmaceutical market for PDx.

3. Products and Operations

AIS: scanners, ultrasound, guidance and software

CT. The Revolution family is the workhorse. The important new product is Photonova Spectra, GE's first photon-counting CT. Unlike conventional CT systems, which first convert X-ray photons into visible light before measuring them, photon-counting CT directly counts individual photons and measures their energy. GE's version uses a detector material no rival uses: Photonova is the first photon-counting CT technology to use "deep silicon," a material that GE HealthCare acquired from Prismatic (bought in 2020).

Specifications that matter:

• 8-bin energy resolution and wide detector coverage with a 0.23-second rotation speed.
• The ability to distinguish between iodine, calcium, and fat.
• It is purposefully designed for easy install and to fit into existing GE HealthCare CT ready rooms with minimal changes, which lowers the hospital's switching cost into the new technology.
• It incorporates NVIDIA accelerated computing technology to handle the data load.

Regulatory status: FDA clearance came in March 2026, followed by Japanese approval and then CE Mark, which GE announced on August 31, 2026.

MR. The SIGNA family. GE's deep-learning reconstruction, AIR Recon DL, was described at its 2022 investor day as able to scan in half the time of a conventional scan while also improving image quality. Software upgrades to the existing MR fleet are a meaningful revenue line in their own right.

Molecular imaging. Omni Legend PET/CT and the StarGuide SPECT/CT line. In Q2 2025 GE secured our largest-ever order of Omni Legend PET systems in the U.S. The next-generation StarGuide GX, aimed at theranostics, is now 510(k) pending with the U.S. FDA. Molecular imaging hardware and PDx tracers reinforce each other: tracers need scanners, and new tracers drive scanner orders.

Ultrasound. This ranges from point-of-care devices to premium cardiac systems. Two recent launches stand out:

• Vivid Pioneer, GE HealthCare's most advanced AI-powered cardiovascular ultrasound system, and Allia Moveo, a compact, cable-free interventional system designed for space-constrained environments.

• In Q3 2026, LOGIQ Focus for general imaging.

Procedural guidance and surgery. Interventional X-ray (the Allia family) and intra-operative ultrasound from the almost $1.5 billion BK Medical acquisition (2021).

Enterprise imaging software. The Company closed the acquisition of Intelerad, for a purchase price of $2.3 billion in Q1 2026, a cloud PACS and enterprise imaging vendor. GE also bought icometrix (AI brain-MRI analysis) in 2025. This is the most "SaaS-like" corner of the company.

PDx: contrast media and radiopharmaceuticals

Iodinated contrast (Omnipaque, Visipaque) is the backbone. The value chain runs from iodine mine to injection:

1. Raw iodine. Secured through a multi-year deal with Sociedad Quimica y Minera de Chile S.A. (SQM).

2. Active ingredient. Made in Lindesnes, Norway, where a 100-acre site has been operating since 1974 and employs 450 people, who use multiple chemical pharmaceutical processes to convert iodine into finished API.

3. Fill-finish. Sterile filling into bottles and polymer packs at Cork (Ireland), Shanghai and other sites. The network along with other sites in Shanghai, China, and Oslo, Norway, supplied over 100 million contrast doses across the globe in 2024.

4. Capacity. GE is investing $138 million (USD) to construct a new contrast media manufacturing site in Cork, expected to add about 25 million additional does of contrast media annually by the end of 2027.

Owning every stage matters. During the 2022 COVID lockdown in Shanghai, the global shortage that followed showed hospitals the value of a supplier with redundant capacity.

MR contrast includes Clariscan and Omniscan. A next-generation manganese-based MR agent is in a combined Phase 2/3 trial (Section 7).

Radiopharmaceuticals

• Flyrcado (flurpiridaz F-18) is the flagship. GE HealthCare acquired exclusive global commercialization rights for flurpiridaz F-18 from Lantheus in 2017 and has led the funding and development of the product through to approval.

• Vizamyl (amyloid PET for Alzheimer's) and Cerianna (breast cancer PET) round out the line.

How PET tracers reach patients. F-18 tracers are made in cyclotrons and must reach the patient within hours. Today GE relies mainly on contract cyclotron networks. On the Q4 2025 call, management said contract manufacturing organization (CMO) partners have been operating at roughly 95% on-time delivery. The planned SOFIE acquisition would bring part of that "final mile" in-house: SOFIE maintains 15 contract manufacturing sites across the country, in addition to a theranostic-focused research facility, with 21 cyclotrons that help deliver hundreds of thousands of doses annually.

PCS: monitoring and life support

Products: bedside and central monitoring, anesthesia delivery, ventilators, maternal-infant care (incubators, warmers, fetal monitors) and diagnostic ECG. The 2025-26 problem was operational, not demand. Q2 2026 revenue fell because of operational fulfillment challenges and supply chain disruptions, and segment profit turned negative from volume shortfalls and component shortages.

The Q3 2026 fix came from the company's lean system. Per the CEO, the team identified root causes of prior supply challenges and implemented standard work and daily management to drive ongoing improvements. This drove backlog conversion and stronger operational performance.

Manufacturing footprint and operating system

We have 14 manufacturing facilities located in the United States and 30 located outside of the United States, including in China, Japan, Mexico, Norway, India, Israel, Germany, Brazil, Austria, Denmark, France, Ireland, the Netherlands, Sweden, Finland, and South Korea. This multi-country footprint was the main lever against U.S. tariffs in 2025. Management cut tariff exposure by multi-sourcing through factories, rationalizing SKUs, and selective price increases.

Plants are run under "Heartbeat," the company's lean management system, which management credits for the PCS recovery and for order-win rates (Section 9).

Recent plant and supply events:

• Q1 2026: a PDx supplier recall. It was roughly $0.05 of EPS impact and occurred late in the quarter, leading to both a write-off and sales shortfall.

• Q1 2026: input-cost shock. An approximate $100 million increase in the price of memory chips, which are critical components utilized in many of our products, plus a similar amount for oil and freight. GE has since entered multi-year strategic agreements with technology partners to secure supply amid growing demand from AI and data centers.

• 2026: PCS component shortages, now largely resolved (above).

4. Customers

Who buys

• Hospitals and health systems: academic medical centres, integrated delivery networks, community hospitals.
• Outpatient imaging centres and cardiology practices.
• Radiopharmacies and nuclear medicine departments (for PDx).
• Governments and public-hospital tender authorities in Europe and China.

Rough distribution of buying by channel:

How capital purchases happen. A modality chair builds the clinical case, the hospital's capital committee ranks it against other projects, and a GPO contract usually sets the commercial frame. GE's revenue is recorded net of group purchasing organization fees, rebates, or credits. The cycle is long. On the Q3 2026 call management noted that the equipment funnel remains as strong as ever, with decisions for installations often planned 12 to 24 months in advance. Imaging competes well for capital: management said imaging remains a top priority for hospital CFOs and they have not seen evidence of procedure pullbacks in scan or contrast usage data.

Why customers choose GE

1. Installed base and room compatibility. New systems drop into existing rooms and workflows (Photonova fits existing CT rooms, Section 3).
2. Service coverage. At spin-off the company cited a global team of over 10,000 sales professionals, 8,500 field service engineers, which matters to a hospital whose CT going down means cancelled patients.
3. Security of supply in consumables. The contrast-media supply discipline described in Section 3.
4. Breadth. One vendor can cover the CT, the contrast, the PET tracer, the reporting software and the service.

Switching costs. These differ by product:

• Imaging hardware. Switching vendors means retraining technologists, rebuilding protocols, re-integrating with PACS and often reconfiguring rooms. Most hospitals stay with their incumbent unless there is a clear technology gap or price gap. Service contracts tie the customer to the installed fleet through its 10+ year life.
• Contrast media. Switching costs are lower (drugs are interchangeable at the molecule level), but hospitals that lived through the 2022 shortage weight reliability heavily.
• Radiopharmaceuticals. Switching is shaped by clinical protocols and reimbursement. A cardiology department that converts from SPECT to PET with Flyrcado is moving to a new protocol, not just a new supplier.
• Monitoring. A hospital standardises one monitoring network across its wards, so switching means replacing hundreds of devices at once. That makes PCS sticky but also exposed when fleet renewals come round.

Concentration. No single customer dominates. The company's risk sits with customer types (U.S. hospital budgets, Chinese tender policy), not individual accounts.

Contract structures and revenue visibility. Equipment deals are multi-element. Contracts for the sale of products and services often include multiple distinct performance obligations, usually involving an upfront deliverable of equipment and future performance obligations such as installation, training, or the future delivery of products or services. Product backlog converts fast and service contracts run long: the company said product-related remaining performance obligations are expected to be recognized substantially within two years, while service-related obligations are expected to be recognized substantially within five years. Overall the mix is about two-thirds product and one-third service. With PDx consumables added, a large share of revenue is recurring.

5. Competitive Landscape and Moat

Structure

Large-modality imaging (CT, MR, PET, interventional X-ray) is an oligopoly. Three Western incumbents (GE HealthCare, Siemens Healthineers, Philips) and Canon Medical have long held most of the global installed base. A Chinese challenger, United Imaging, has scaled quickly from a protected home market. A 2021 assessment called medical imaging increasingly commoditized, but one where GE also benefits from intangible assets, scale, and switching costs. That is a fair description: price competition is real at the mid-tier, while the top end turns on technology leadership (photon-counting CT, high-field MR, total-body PET).

Contrast media is a different oligopoly of a few global producers. Radiopharmaceuticals are fragmented and growing quickly. Monitoring is a three-way global contest with a strong Chinese entrant.

Competitors

Market cap sources: Siemens Healthineers EUR 47.3 billion as of October 27, 2026; Philips EUR 23.3 billion market capitalization on October 27, 2026; United Imaging CN¥93.07B as of Oct 28, 2026; Mindray CNY 202.7 billion on October 16, 2026; Bayer EUR 44.5 billion as of October 26, 2026; Lantheus USD 6.5 billion as of October 5, 2026.

Where GE wins and loses

Against Siemens Healthineers. Siemens took the lead in photon-counting CT. GE's clearance positions it to compete with Siemens Healthineers' Naeotom Alpha, the first photon-counting CT machine to reach the U.S. market. GE's answer is a different detector (deep silicon) and drop-in room compatibility. Whether that wins share is unproven; Citi noted the FDA nod positions GE HealthCare to show whether its claims about the advantages of its technology translate into market share.

In ultrasound the field has just tilted towards GE. Siemens Healthineers has ended sales of its general imaging ultrasound systems to focus solely on procedural imaging and image-guided therapy. As of October 12, the company no longer offers its Acuson Sequoia, Redwood, Juniper, and Maple systems. That opens Siemens' general-imaging installed base to GE, Philips, Canon and others. Arduini called general ultrasound, imaging ultrasound

a super important market for us on the Q3 call.

Against Philips. Philips has long been described as GE's fiercest competitor in clinical care (monitoring). GE's PCS weakness in 2025-26 was self-inflicted (fulfillment), which suggests share may have leaked to Philips and Mindray during that period. In imaging GE has broader molecular imaging and a contrast/tracer franchise Philips lacks.

Against United Imaging in China. This is where GE loses most clearly. A January 2026 Mergermarket report said GE's China imaging business faces severe competition in China, especially from the listed local rival Shanghai United Imaging, and that public hospitals would face challenges in procuring US brands amid geopolitical tensions. The same report, citing anonymous sources, said GE had begun pre-marketing a carve-out sale of the China imaging business. The company has not confirmed this. When asked about media reports in Q3 2025, management declined to comment and said its mid-single-digit goals hold if, you know, China remains roughly flat stable market.

Against Bayer and Bracco in contrast. GE competes on supply security and SKU breadth (Section 3). Management said in Q2 2026 that current market demand is close to outpacing total market supply. In a supply-tight market, the producer with spare capacity wins volume and pricing.

Against Lantheus in radiopharmaceuticals. Lantheus is both rival and former partner. GE's edge is pairing tracers with its own PET/SPECT scanners and, after SOFIE, controlling U.S. distribution. Even after the deal, SOFIE helps produce Lantheus' Pylarify (piflufolastat F-18) for prostate cancer imaging and Neuraceq (florabetaben F-18) for beta amyloid plaques, so GE will be a supplier to a direct competitor.

Barriers to entry

High, but not absolute.

• Imaging. Entry needs decades of physics and detector IP, regulatory clearances in every major market, a global service network and installed-base references. United Imaging shows a well-funded, state-backed entrant can break in. It did so from a protected home market, and it took more than a decade.
• Contrast. Requires integrated API chemistry, iodine supply, sterile manufacturing and drug approvals.
• Radiopharmaceuticals. Barriers are clinical trials plus cyclotron logistics. That is why distribution assets like SOFIE are strategic.

Structural shifts

1. Photon-counting CT is the biggest technology transition in CT in two decades. Siemens has a head start; GE's entry makes it a two-horse race at the top end.
2. Portfolio pruning. Siemens' ultrasound exit and GE's own PCS review show the incumbents focusing on fewer, deeper franchises. Reports also say Siemens plans to deconsolidate Siemens Healthineers, which would change its ownership and possibly its strategy.
3. China localisation via volume-based procurement and "buy domestic" policies. The EU's response (a ban on Chinese firms bidding on EU tenders over €5m for a five-year period) adds a protectionist layer in GE's favour in Europe.
4. Radiopharmaceutical boom. New PET tracers and theranostics are turning tracer distribution into strategic real estate.

Moat: Narrow - switching costs, intangible assets (regulatory approvals, detector IP), scale in service and supply

The installed base, service network and room-level integration (Section 4) make incumbent replacement the default, and the integrated contrast supply chain is hard to copy. The advantage has not held everywhere, though. In China, a determined, policy-backed competitor took share through 2024-25, and Siemens beat GE to market in photon-counting CT. Neither looks like a "wide" moat that survived a well-funded attack. The single thing most likely to erode it is a technology-generation loss in a top modality (photon-counting CT) combined with continued localisation in China. Together these would wear away both the premium segment and a large emerging market.

6. Industry

What drives demand

• Ageing populations and chronic disease (cardiovascular, oncology, neurology) increase scan volumes. Most procedures need both a scanner and, often, a contrast agent or tracer.
• Replacement cycles. Hospitals replace CT and MR every 8-12 years, and new technology (AI reconstruction, photon counting) can pull replacement forward.
• Precision medicine. Targeted therapies need imaging to select and monitor patients. The radiopharmaceutical pipeline industry-wide includes approximately 20 PET radiotracers and more than 30 radiotherapeutics currently under development.
• Contrast demand. GE expects iodinated contrast demand will double over the next 10 years.

Size and growth. Estimates vary by scope. IMARC puts global diagnostic imaging equipment at USD 44.4 Billion in 2025, to reach USD 70.4 Billion, growing at a CAGR of 5.09% during 2026-2034. DelveInsight puts it at USD 51.5 Billion in 2025 at ~5% growth, with North America held the largest share of the diagnostic imaging equipment market in 2025, accounting for approximately 41% of the global market. Contrast media, radiopharmaceuticals, monitoring and imaging IT add markets of similar order. The picture is a mid-single-digit industry with faster pockets (radiopharma, imaging software).

Where GE sits in the supply chain. GE is an OEM and drug manufacturer. Upstream it relies on:

• Iodine from Chile.

• Semiconductors and memory chips. Here medical imaging now competes with AI data-centre buildouts. Arduini said data-centre growth creates a "vortex effect on certain components,"

• Helium and rare metals for MR magnets and X-ray tubes.

• Contract cyclotron operators for PET tracers.

Downstream it sells directly or through distributors to providers.

Regulation

• Devices need FDA 510(k)/PMA, EU MDR CE marking and China NMPA registration.
• Contrast and tracers are drugs that need full approvals. New PET tracers depend on reimbursement codes and pass-through payments.
• Trade policy is now a first-order industry variable:

  • IEEPA tariffs struck down. The February 20, 2026 ruling struck down tariffs imposed under IEEPA, but Section 232 and Section 301 duties rest on separate authority and remain in force.

  • Section 232 investigation still open. A probe into medical equipment opened in September 2025, and as of 2026-11-01 the BIS Section 232 page still lists this investigation under ongoing investigations, with its initiation date, and displays no report, findings or action for it. Its scope explicitly includes X-ray and imaging equipment.

  • China. Procurement is shaped by volume-based purchasing, and the 2023-onward anti-corruption campaign contributed to delayed orders and sales in our China business.

Import substitution. In China, policy explicitly favours domestic devices. The EU found that China's policies favoured domestic medical devices over imported ones by design. In the U.S., the Section 232 probe is the mirror image: a possible push to reshore device manufacturing, which would hurt every imaging OEM with non-U.S. plants, GE included.

Cyclicality. Capital equipment is moderately cyclical. It follows hospital budgets, reimbursement and financing conditions, and it lags the economy by several quarters because of long purchasing cycles. Service and consumables are much steadier: scan volumes rarely fall, and contrast or tracer use follows procedures, not budgets. The 2020 pandemic is the clearest stress test. Elective imaging collapsed for a quarter, then rebounded, and capital orders followed with a lag.

Tailwinds: ageing demographics, AI-enabled productivity (faster scans offset radiologist shortages), radiopharmaceutical innovation, the photon-counting replacement cycle, and Siemens leaving general ultrasound.

Headwinds: tariff uncertainty, memory-chip and freight inflation, China localisation, and U.S. hospital margin pressure from possible changes in public reimbursement.

7. Growth Triggers

Ranked by potential impact on the business. All are drawn from the six results calls between July 2025 and October 2026.

• Flyrcado ramp to a medium-term revenue target of $500M+ by 2028. Repeated in Q4 2025, Q2 2026 and Q3 2026. Management reiterated confidence in its prior target of $500 million or more in revenue by year-end 2028 and described a supply-driven, "go slow to go fast" launch approach. (Q4 2025 call, Feb 4 2026). The company delivered 545 Flyrcado doses for the week ending July 24 (~40% increase vs. April levels) (Q2 2026 call, Jul 29 2026). Arduini cited roughly 680 doses for the week ending October 23, broad-based adoption, and growing funnels (Q3 2026 call, Oct 28 2026). More than 80% of the long-term opportunity is expected to come from SPECT-to-PET conversion.

• New-product wave, with Photonova Spectra photon-counting CT at the front. Repeated in Q4 2025, Q2 2026 and Q3 2026. Management expects new products along with Flyrcado, to help drive 1-2 points of additional sales growth over the medium term, with the biggest contributions expected in 2027 (Q4 2025 call, Feb 4 2026). On Photonova, we're seeing strong interest in Photonova Spectra, our photon-counting CT platform, and we expect CE marking in the second half of 2026, which will expand our opportunity in Europe (Q2 2026 call, Jul 29 2026). Arduini said top-line momentum should continue into 2027, with higher-margin new products contributing more (Q3 2026 call, Oct 28 2026). He also said the Q4 order book looks better than Q3:

"Our funnel is as strong as it's ever been at this point in time."

• Peter Arduini, Q3 2026 call, Oct 28 2026 (source)

• Outcome of the PCS strategic review. Q2 and Q3 2026. The review is evaluating "continued ownership, a sale and other value-enhancing transactions." (Q2 2026 call, Jul 29 2026). Incoming PCS CEO Neal Sandy has been heavily involved in actually leading much of the work around the strategic assessment. (Q3 2026 call, Oct 28 2026). He takes over on November 16, 2026. A sale would change the group's mix, growth rate and capital deployment.

• SOFIE Biosciences acquisition: owning the U.S. "final mile" of PET supply. Announced in October 2026; the transaction is expected to close in the first half of 2027, subject to customary closing conditions, including regulatory approvals. Management said SOFIE's revenue will add approximately 1% of total company revenue on an annualized basis in the full year of ownership and that SOFIE is expected to grow in the low double digits. The deal includes Sofie's FAPI-74, an F-18-labeled fibroblast activation protein inhibitor (FAPI) PET agent in phase III development. (Q3 2026 call, Oct 28 2026)

• Intelerad cloud enterprise imaging. Q4 2025 and Q1 2026. Management expects Intelerad to add about $270 million of revenue in the first full year of ownership, growing low double digits, with adjusted EBITDA in excess of 30%. (Q4 2025 call, Feb 4 2026). After closing, We're excited about the opportunity to grow our AI, cloud and software capabilities, leveraging our Intelerad platform. (Q1 2026 call, Apr 29 2026)

• Manganese-based MR contrast agent. Q1 and Q3 2026. First patient was dosed in the Phase 2/3 LUMINA trial under FDA Fast Track designation (Q1 2026 results, Apr 29 2026). Management said it has the potential to address a nearly $1.2 billion market, and we're making good progress on our combined phase II and phase III international trial. (Q3 2026 call, Oct 28 2026). A gadolinium-free MR agent would be a new franchise.

• StarGuide GX SPECT/CT and theranostics. Q3 2026. GE aims to advance a new era in theranostics with StarGuide™ GX 4D SPECT/CT, now 510(k) pending. Arduini framed SOFIE as positioning GE to pursue the next generation of precision care, including PSMA, FAPI and theranostics (Q3 2026 call, Oct 28 2026).

8. Key Risks

8.1 Section 232 device tariffs. Mechanism: GE imports scanners, components and finished drugs into the U.S. from its 30 non-U.S. plants (Section 3). In 2025, IEEPA tariffs cost the company about $0.45 of adjusted EPS even after mitigation. Those tariffs have since been struck down and partly refunded. The Section 232 probe (Section 6) could impose new, longer-lasting duties on imaging equipment specifically. Management called it the biggest tariff swing factor and the company is planning for higher rates, preparing pricing, cost productivity, and component moves to offset them. Calibration: moderate-to-high probability of some action; moderate impact, because GE showed in 2025 it could offset more than half. The risk is a duty level beyond what pricing and sourcing can absorb in a year.

8.2 Component inflation driven by AI-hardware demand. Mechanism: memory chips, freight and metals added roughly $250M of 2026 cost pressure (Section 3). GE has secured supply through multi-year agreements, but not price:

"Getting locked in pricing agreements on some of these things isn't probably realistic"

• Peter Arduini, Q3 2026 call (source)

If AI data-centre demand keeps tightening DRAM and NAND supply, GE pays more every year for parts it cannot redesign quickly, and hospital contracts priced 12-24 months in advance (Section 4) cannot pass it through fast. Calibration: high probability, moderate drag on profitability; low risk to the franchise.

8.3 China: structural share loss and possible exit. Mechanism: China was a double-digit share of revenue at spin and has fallen two years running (Section 2), squeezed by volume-based procurement, the anti-corruption campaign and United Imaging (Section 5). Management assumed a China decline for 2026 and described Q3's growth as mostly currency. A reported carve-out of the China imaging business, if it happens, would change the group's scale and manufacturing footprint (Shanghai also makes contrast media, Section 3). Calibration: high probability of continued pressure. The tail risk is a disorderly exit or geopolitical escalation that hits both sales and China-based production.

8.4 PCS: execution, leadership churn and transaction risk. Mechanism: PCS swung from growth to negative segment profit in Q2 2026 on fulfillment failures (Section 3), recovered in Q3, and now changes CEO mid-review. Three things could go wrong. A monitoring-fleet renewal cycle could be lost to Philips or Mindray while execution is unstable. A sale could come at a depressed value. Separation could strand shared costs and break cross-selling with AIS. Management noted Q4 PCS growth will be constrained by the $45 million revenue comparison from a prior year product release. Calibration: moderate probability, moderate impact. PCS is about one-seventh of revenue but has been a large share of recent negative surprises.

8.5 Radiopharmaceutical supply chain fragility and Flyrcado execution. Mechanism: The highest-growth part of the story depends on tracers that decay in about two hours, made mostly by third-party CMOs (Section 3). Q1 2026 showed what a single supplier failure can do. Flyrcado's ramp also depends on cardiology practices buying PET capacity, reimbursement and workflow change. Clinicians have noted the Flyrcado protocol is longer than that of traditional Rubidium-82 PET imaging, primarily due to the recommended 30-60-minute delay between the rest and stress injections. SOFIE (Section 7) helps but adds integration risk, and SOFIE supplies competitors. Calibration: moderate probability of delay against the 2028 target; moderate impact on the growth narrative more than on the overall business.

8.6 Losing the photon-counting CT generation. Mechanism: Siemens reached market first (Section 5). If deep-silicon detectors underperform in clinical preference, cost or reliability, GE could lose top-end CT share, its highest-value modality, for a decade. Calibration: low-to-moderate probability; high impact if it happens, because CT leadership supports service and contrast pull-through.

8.7 Management turnover during a reorganisation. Mechanism: Within six months GE replaced its CFO (Jay Saccaro left for an expanded role covering finance and operations at Zoetis), merged two segments and replaced their leaders, rebuilt the commercial organisation, and announced a PCS CEO change. New CFO William Grogan joined from Xylem; per the company's appointment release, he served there as executive vice president and chief financial officer since 2023 and earlier at IDEX as senior vice president and chief financial officer from 2017 to 2023. Turnover raises the odds of guidance slips while new leaders reset expectations. Calibration: moderate probability, low-to-moderate impact.

9. Walk the Talk

Results calls used:

1. Q2 2025, July 30, 2025
2. Q3 2025, October 29, 2025
3. Q4 2025, February 4, 2026
4. Q1 2026, April 29, 2026
5. Q2 2026, July 29, 2026
6. Q3 2026, October 28, 2026 (within 90 days of this report)

The Q1 2026 miss. In February, management set FY2026 profit guidance assuming a benign cost environment. Twelve weeks later it cut that guidance. The CFO's explanation for the quarter was narrow:

"Were it not for this issue, we would have achieved the quarter."

• Jay Saccaro, CFO, Q1 2026 call (source)

The supplier recall itself was small. The larger cut came from inflation management had not anticipated in February: about $250M of memory, oil, freight and metals cost (Section 8.2). The pattern is reasonable top-line forecasting combined with underestimating cost-side volatility. The CFO left three months later. That departure was for an outside role and was not tied to the miss in any disclosure.

The medium-term promise. On the Q4 2025 call, Arduini was emphatic:

"We're committed to those midterm targets, top and bottom, full stop."

• Peter Arduini, Q4 2025 call (source)

He repeated the "full stop" phrasing in Q3 2026. The top-line half looks credible: trailing-year organic growth is now about 5%, and orders rose a record 11% in Q2 2026. The margin half needs both the new-product mix and continued cost relief to land, and has not yet been tested against a full year of the new cost base.

Unusually accurate calls. The product-milestone calendar (Intelerad close, Photonova FDA clearance and CE mark, the Flyrcado dose ramp) has been met on time or early. Orders and backlog guidance across 2025-26 was consistently borne out.

Assessment. On revenue, product launches and deal closings, this management does what it says, and is slightly conservative on the top line. On profit it has been optimistic at the start of the year and then forced to adjust when external costs moved (tariffs in 2025, inflation in 2026). In both years it clawed back most of the cut by the third quarter. The result is a team that delivers on what it controls and over-promises on what it does not.

10. Ownership, Governance and Shareholder Friendliness

Part 1: Ownership and control. GE HealthCare is widely held, with no controlling shareholder, family, parent or state. As of March 31, 2026 the largest holders included Blackrock Inc. 35.98M 7.91%, Dodge & Cox Inc. 31.31M 6.88%, Vanguard Capital Management LLC 28.86M 6.35%, Capital Research Global Investors 21.11M 4.64%, State Street Corporation 20.54M 4.52%, plus further Vanguard entities. More recent data shows Dodge & Cox increasing its stake, with Dodge & Cox 10.48% as of July 31, 2026. Insiders own very little: 0.35% % of Shares Held by All Insider. The balance sheet shows one class of common stock, so there are no dual-class shares or golden share. For a minority shareholder, voting power is proportional to economic ownership, and control rests with the board answering to large institutions. With an active value manager approaching 10%, those institutions have some leverage over portfolio decisions such as PCS and China.

Part 2: Governance. Governance is largely clean, with a few points worth knowing:

• Board. Two directors left at the 2026 meeting, and the size of the Board will be 8 directors.

• Chairman's independence. H. Lawrence Culp Jr. was formerly classed as non-independent under Nasdaq rules because of his GE ties. The 2026 proxy states Culp, our Chairman, is now independent. The board still appointed a lead director: Effective as of May 7, 2026, the Company's Board of Directors appointed Rodney F. Hochman as the Board's independent lead director.

• New director. Kevin A. Lobo joined in March 2026; per the company's announcement he has served as the Chief Executive Officer of Stryker Corporation since October 2012 and as chair of the Stryker Board of Directors since July 2014.

• Pay. Pay is tied to measurable metrics. Long-term PSUs pay on Organic revenue (50%) + Cumulative Adjusted EPS (50%) X Relative TSR Modifier vs. Compensation Peer Group (+/-20%). The CEO's 2025 total compensation was $18.6 million in total compensation for 2025, down 5% from $19.5 million the year before, and say-on-pay received 94.4% support in 2025.

• Not found: material related-party transactions with management, insider share pledges, or an auditor qualification in the filings reviewed.

Part 3: Capital returns (FY2023-FY2025, plus 2026 to date)

Dividends. GE HealthCare pays a token dividend:

• FY2023 (post-spin, partial year): three quarterly payments of $0.03, about $0.09.
• FY2024: four payments of $0.03, $0.12.
• FY2025: quarterly dividend of $0.035 per share to stockholders of record for all quarters in 2025, $0.14.
• 2026: the board raised it again, with a quarterly dividend of $0.04 per share to stockholders of record as of October 23, 2026, an increase of 14% from the previous quarter.

The dividend has grown every year, but from a base that makes it symbolic. Cash goes mainly to reinvestment and acquisitions.

Buybacks and dilution.

• FY2023-FY2024: no repurchases.

• April 2025: the Board authorized a share repurchase program pursuant to which GE HealthCare may repurchase up to $1.0 billion of its common stock. The program does not have an expiration date.

• FY2025: the Company repurchased 2.8 million shares for total consideration of $200 million.

• First nine months of 2026: we repurchased 4.7million shares for total consideration of approximately $ 300million. That includes 1.4 million shares in March at an average price of $71.35.

• Q3 2026 and the last ~90 days: the company did not repurchase any shares during the three months ended September 30, 2026. The MoatMap feed also records zero buybacks since August 4, 2026. As of September 30, 2026, we had $ 500million available under the repurchase program authorization.

• Cumulative: $500M of the $1B authorization used, about 7.5M shares retired.

Net share count was roughly flat from spin to end-2025, as equity-plan issuance offset early buybacks. Diluted weighted-average shares then fell to 453 million in Q3 2026 from 457 million a year earlier, a decline of about 1%.

Verdict: Neutral. GE HealthCare returns a small amount through a symbolic dividend and opportunistic buybacks, while most free cash flow goes to acquisitions (Nihon Medi-Physics, Intelerad, SOFIE) and R&D.

11. Insider Activities

Window: November 2025 to November 2, 2026. Base data from the MoatMap insider feed (current to November 2, 2026), cross-checked against SEC Form 4 filings and coverage of them. The MoatMap feed recorded only three of the seven open-market purchases in the window. The other four (Culp, Stromberg, Arduini, Jimenez) were identified from Form 4 coverage and are added below. No Form 4 filings dated after mid-September 2026 were found in the cross-check of the most recent two weeks.

The other entries in the window are routine. Most are the September 1-3, 2026 tax withholdings on vested RSUs across the executive team, for example Arduini's Withholding of shares to satisfy tax withholding obligations in connection with the vesting of restricted stock units. MoatMap also lists two May 15, 2026 "Other" entries for Jeannette Bankes and Philip Rackliffe; their nature was not verified and they are not open-market trades.

Buys: reading the signal. This is textbook cluster buying. Seven insiders (the chairman, the CEO, the then-CFO, the general counsel and three independent directors) bought on the open market within about three weeks. The buying started the day after the April 29 profit-guidance cut (Section 9), with the stock near its lows. The prices paid ($59.92-$64.18) sit below the ~$71 at which the company itself had bought back stock in March.

• Chairman Culp's ~US$5.0M purchase was by far the largest. Against the 398,670 shares he reported in the 2026 proxy, it lifted his holding by roughly 20%. A single large open-market purchase by the board chair is a very bullish signal.
• Arduini's ~US$250k is roughly two months of his base salary of $1,340,000. That is meaningful but modest next to his equity awards.
• Lobo's purchase more than tripled his position within ten weeks of joining the board. A sitting medtech CEO who knows the industry well chose to buy.
• Saccaro bought about $200k in May and left in July for a role at another company. The purchase shows conviction at the time; his departure was disclosed as being for a broader role elsewhere.

Sells. There were no open-market sales by any director or officer in the window. All dispositions were tax withholdings on vesting.

Net assessment. Insiders were net buyers, and the buying was broad: management, the chair and independent directors, concentrated in a single window of price weakness after a guidance cut. Nothing has been sold since. The new CFO's sign-on equity is standard and carries no signal. Read: bullish signal.

12. Scenarios

Bull case. Photonova Spectra wins a real share of the photon-counting replacement cycle in the U.S. and Europe, and Siemens' exit from general ultrasound hands GE a run of installed-base conversions. Flyrcado keeps compounding as cardiology practices convert from SPECT to PET, and SOFIE closes on time, giving GE control of U.S. tracer logistics and a late-stage FAPI pipeline. PCS is either fixed and kept, or sold for a fair price with the proceeds put into buybacks and radiopharmaceuticals. Section 232 lands at a level GE can price through, and memory-chip prices ease as data-centre supply catches up. By 2028-29 GE HealthCare looks less like a capital-equipment cyclical and more like an imaging platform with a fast-growing consumable drug and software layer. The record order book and the newer product generation would support mid-single-digit growth near the top of management's medium-term range, with profitability approaching its target band.

Base case. Management delivers roughly what it has guided, as it has on revenue and milestones (Section 9). Orders and the backlog convert at a mid-single-digit pace. Flyrcado and Photonova scale steadily but take until 2027-28 to matter at group level. Intelerad and SOFIE integrate without drama. PCS stabilises under new leadership, and the review ends in either a sale or a slimmer, retained business. Each year brings at least one external cost shock (tariffs, components) that forces a mid-year profit reset, mostly recovered by the fourth quarter. China drifts sideways to down. The company reaches the lower end of its medium-term profitability ambitions later than hoped, with capital going mainly to bolt-on deals and moderate buybacks.

Bear case. Risks 8.1 and 8.2 compound: Section 232 duties on imaging equipment arrive while AI-driven memory inflation persists, squeezing profitability for two years because contracts priced 12-24 months ahead cannot pass costs through. Risk 8.6 plays out as Siemens' photon-counting head start holds, keeping GE's premium-CT share flat. Risk 8.3 deepens, with VBP and local preference pushing GE further out of China, possibly through a forced or discounted carve-out that also disrupts Shanghai contrast production. Risk 8.5 adds a second radiopharmaceutical supply stumble, delaying Flyrcado well past 2028, and risk 8.4 ends with PCS sold cheaply after losing a monitoring-renewal cycle. With a new CFO and new segment heads (8.7), credibility on the medium-term targets erodes, and GE HealthCare is again seen as a low-growth hardware vendor.