Stockholm's Eyedentity raises €1.3M to screen for deadly eye cancer inside routine optician appointments

Eyedentity (Stockholm) has closed an oversubscribed first investment round at €1.3 million (SEK 15 million) led by Norrsken Launcher, with participation from Karolinska Institutet Innovation and a group of angel investors. The company uses AI to screen for uveal melanoma — the most common primary intraocular cancer in adults — from fundus images already taken by opticians during routine eye examinations.

The company was founded in 2025 by Gustav Stålhammar, an ocular oncologist at St. Erik Eye Hospital and Professor at Karolinska Institutet, and Mats Holmström, a machine learning engineer who previously built ML models for automated cancer care at RaySearch Laboratories. The pairing is deliberate: a specialist who has seen the diagnostic gap in clinical practice, and an engineer who has solved similar classification problems in adjacent oncology.

The optician as the triage layer

Uveal melanoma presents a specific diagnostic problem. The cancer is rare — but the benign choroidal naevi it closely resembles are common, appearing in roughly 6% of the adult population. When a suspicious lesion is found during a routine eye examination, the current referral pathway sends the patient to specialist eye care. That pathway is under pressure: if screening widened, the referral volume would overwhelm the specialists who decide whether a lesion is malignant.

Eyedentity's solution is to insert an AI triage step before the referral decision. The platform receives a fundus image from the optician, analyses it, and returns a decision-support output: refer or not. The company reports the system reduces false-positive referrals by 90% — which means the same specialist capacity processes a much higher-quality queue of genuinely suspicious cases, while the vast majority of benign lesions are cleared without a hospital visit.

Critically, no new equipment is involved. Opticians already capture fundus images during routine appointments. The platform simply receives those existing images. As Holmström has put it, the goal is "zero clicks for the optician" — the AI sits in the background of the existing workflow.

Norrsken's thesis and the Synologen pilot

Anna Fredrixon, Partner at Norrsken Launcher, describes the investment in terms of system load: "This is a clear example of technology that can save lives but, unlike other AI companies, does not send more patients to an already overburdened healthcare system." That framing reflects Norrsken Launcher's specific investment criteria — it backs impact companies that relieve rather than strain public infrastructure.

Eyedentity is currently running a live pilot with Synologen, a network of independent opticians across Sweden. The pilot is the first commercial validation beyond controlled trials: whether the model's performance holds in a real clinical workflow, across the variability of equipment and image quality found in community opticians, is what the next 12 months test.

The capital also funds an expanded diagnostic scope. Beyond uveal melanoma, the company plans to apply its platform to diabetic retinopathy and glaucoma — two conditions with much larger patient populations and well-established screening infrastructure into which Eyedentity's model could slot. The longer-term ambition is a multi-condition AI diagnostics platform for opticians, eye clinics, and hospitals. The uveal melanoma product is the first, not the only, application.

This is Eyedentity's first external capital. The company is based in and around the Karolinska Institutet Innovation Incubator, which provided early business development support and — in Holmström's account — the investor introduction that led to the Norrsken relationship. The round was oversubscribed.