The surgical AI market

Commercial launches and partnerships matter when they change access, infrastructure, evidence generation, or competitive direction. Announcements remain claims until independently supported.

In this edition

• 01 — Raman AI can read tissue chemistry—but first it must survive another instrument
• 02 — Intuitive’s Q1 numbers show where robotic surgery is becoming infrastructure
• 03 — Intuitive is taking Southern European distribution in-house

01 — Raman AI can read tissue chemistry—but first it must survive another instrument

Raman spectroscopy shines a laser on tissue and measures the scattered light. The resulting spectrum acts like a biochemical fingerprint: potentially useful for separating tumour from normal tissue, identifying molecular subtypes or supporting a rapid intraoperative diagnosis without staining. Machine learning is attractive because those fingerprints are high-dimensional, noisy and difficult to interpret by eye.

This review follows the entire chain—from acquisition and fluorescence correction to classification, explanation, multimodal pathology and clinical deployment. For surgeons, its most important message is that the “AI model” begins long before the classifier.

The instrument can become the hidden diagnosis

Laser wavelength, power, integration time, calibration, sample preparation and spectrometer design all change a spectrum. Preprocessing choices then alter it again: smoothing can erase a weak biological peak, while insufficient correction can leave batch or instrument artefacts that a model learns as shortcuts. Two teams can use nominally similar algorithms yet feed them materially different signals.

The review highlights a particularly dangerous form of leakage. A single patient or tissue specimen may contribute hundreds of spectra. If those spectra are randomly divided between training and testing, the model can recognise the source rather than the disease and report an inflated result. The minimum credible split is therefore at patient or biological-source level. A surgical tool meant to travel between theatres also needs external testing across instruments, operators and acquisition protocols.

Intraoperative promise is real, but uneven

The authors survey Raman and stimulated Raman histology studies in neuro-oncology, colorectal disease and other cancers that have produced rapid tissue classification or molecular information within operative time constraints. Those examples show that the modality can fit a surgical question: “What is this tissue now?” But the review also finds small single-centre cohorts, inconsistent metadata, limited code and data sharing, platform-specific bias and too little prospective external validation.

A clinically useful pipeline must report the whole clock—not only model inference. Acquisition, quality control, preprocessing, prediction and display all consume time. It must also preserve uncertainty, connect highlighted spectral bands to plausible biology, and show that performance holds after a different instrument or site changes the signal. Sometimes a simpler model may be preferable if it is easier to validate and reproduce.

This paper is a narrative review, not a new diagnostic trial or formal meta-analysis. Its value is architectural: it shows surgeons where to interrogate a Raman-AI claim. Ask who or what was split between datasets, whether preprocessing was learned only on training data, how many independent patients and centres were tested, and whether the reported turnaround includes the entire workflow. A high accuracy on spectra is not yet a reliable answer at the resection margin.

Open the original source

02 — Intuitive’s Q1 numbers show where robotic surgery is becoming infrastructure

An earnings release is not clinical evidence. It can still reveal how quickly a surgical technology is becoming infrastructure: how many procedures run through it, how fast the installed base grows and how hospitals choose to finance access. Intuitive’s first-quarter 2026 release is unusually clear on those signals.

The company reports that combined da Vinci and Ion procedures grew about 17% versus the first quarter of 2025. Da Vinci procedures rose about 16%; Ion bronchoscopy procedures rose about 39%. Intuitive placed 431 da Vinci systems during the quarter, including 232 da Vinci 5 systems. A year earlier it placed 367 systems, including 147 da Vinci 5 units. The global da Vinci installed base reached 11,395 systems, up 12% year over year.

Revenue rose 23% to $2.77 billion, and instruments and accessories alone generated $1.69 billion. That recurring revenue is tied to procedure volume: every new system creates a long relationship involving instruments, service, training and data infrastructure. The shift in purchasing model is also visible. Of the 431 placements, 243 were operating leases and 118 of those were usage-based leases. Hospitals are not always buying a robot outright; some are tying access and vendor revenue more closely to utilization.

For surgeons, scale can have real consequences. A larger installed base can expand training exposure, make cross-site workflows more familiar and generate more operational data. It can also deepen dependence on one platform’s instruments, service schedules, analytics and credentialing ecosystem. Usage-linked contracts may lower the initial capital barrier while changing incentives around case volume and programme growth.

Every number above is company-reported. Procedure growth does not reveal case complexity, conversion, complications, equity of access or whether an operation was better than an alternative. A “placement” does not show how intensively the system is used, and installed-base growth does not account for every retirement or local capacity constraint. The full-year procedure-growth forecast of 13.5% to 15.5% is forward-looking, not an observed result.

The correct surgical reading is therefore neither celebration nor dismissal. These numbers show that robotic and endoluminal platforms are spreading quickly and that da Vinci 5 is taking a growing share of new placements. When a platform reaches this scale, procurement becomes a clinical-governance decision. Hospitals should pair utilization targets with outcomes, training capacity, access and total cost; surgeons should know how lease incentives interact with case selection. Market momentum tells us where the operating room is going. It does not tell us whether every journey is justified.

Open the original source

03 — Intuitive is taking Southern European distribution in-house

Intuitive’s March 2026 announcement is easy to file as corporate expansion. For surgical teams in Southern Europe, it is more concrete: the manufacturer has acquired the da Vinci and Ion distribution businesses previously operated by ab medica, Abex, Excelencia Robótica and their affiliates.

The company now runs direct operations in Italy, Spain, Portugal, Malta, San Marino and associated territories. About 250 employees are joining Intuitive, and the release reports a combined installed base of more than 470 da Vinci systems in Italy, Spain and Portugal at the end of 2025. Ion had recently launched in Italy and Spain.

The relationship around the robot changes

Distribution is not only sales. It includes installation, service response, instrument supply, training coordination, software updates and the route by which a hospital escalates a problem. Moving these functions inside the manufacturer could standardize support across countries and connect local teams more directly to product specialists. That is Intuitive’s stated rationale, and it may help new technology or training programmes travel faster.

But direct control also consolidates the commercial and educational relationship. Hospitals that already depend on one platform for instruments and maintenance may now negotiate with the manufacturer rather than a regional distributor. Local knowledge, staff retention and continuity during the transition become practical safety issues. So do contract terms, spare-parts logistics, uptime guarantees, language-specific training and governance of the procedural data moving through digital services.

An access claim still needs an access measure

Intuitive says direct operations will support more agile service and broaden minimally invasive care. The announcement offers no prices, service-level results, training-capacity data or patient-access analysis. The installed-base count says where machines exist, not whether geography, staffing and referral pathways allow patients to use them. It also provides no clinical outcomes from the acquisition—nor should a corporate completion notice be expected to.

Surgeons and hospital leaders can turn this market event into an accountable transition. Track service response and cancelled cases before and after the change. Protect current training pathways while new ones are introduced. Clarify who owns and can export operational data. Compare instrument and maintenance costs over the full contract, and preserve a route for independent evaluation of new platform claims.

This acquisition will not change an operation overnight. It changes the infrastructure around hundreds of robotic systems. If that produces more reliable support and equitable programme development, the benefit should become measurable. Until then, “expanded access” remains a company objective rather than an observed clinical result.

Open the original source