Hyperfine (HYPR) AI Portable Brain MRI Cuts ED Scan Wait to 1.28 Hours

Hyperfine (HYPR) said on its 2026-08-06 earnings call that its AI portable brain MRI reached a median emergency-department order-to-scan-start time of 1.28 hours.[1] That reading comes from the scanner and the clinical workflow together, and the company did not break out what its software contributed on its own.

What Hyperfine's AI portable brain MRI is

Hyperfine, Inc. (HYPR) makes Swoop, a portable brain MRI scanner that can be wheeled to the bedside. Its field strength is 64 millitesla, far below the large MRI systems hospitals normally run. Revenue comes mainly from selling complete systems and from service subscriptions that include maintenance and software. Data acquired at that field strength is noisy, so the company's own software, Optive AI, denoises, reconstructs and post-processes it into brain images a physician can read. The next-generation system that carries it is now called Model 2. The users are physicians in emergency departments, intensive care units and neurology offices, who use the images to judge whether a patient has a stroke, a hemorrhage or another intracranial condition. The software sits in the core of the business: what Swoop sells is the image the software produces.

How the disclosure moved from image quality to clinical readings

The application is considerably older than its current name. On the 2023-11-10 call it was still only an AI denoising feature of Swoop, newly extended into the diffusion-weighted sequence used to identify stroke. The software had been cleared but not launched, and only a few customers had it in early access. All management could say was that crisper images might help clinicians diagnose and monitor patients.[2]

By 2025-08-13 it had been productized on its own. It had the name Optive AI, it was cleared as standalone software, and it went both into the next-generation system and back onto scanners already sold. Clinicians reported image quality approaching that of a conventional 1.5 tesla MRI.[3]

On 2025-11-13 the subject shifted from image quality to the buyer's ledger, as hospitals began evaluating the scanner with return-on-investment math.[4]

On 2026-08-06 the next-generation system was renamed Model 2, and the company put forward checkable readings on emergency-department workflow time and diagnostic concordance for the first time.[1]

What the latest study readings show

The evidence in the most recent disclosure is workflow data that can be checked. PRIME reported that in the emergency department the median time from a physician's order to the start of the scan was 7.76 hours with conventional MRI and 1.28 hours with the portable device.[1] What gets compressed is the stretch of time a patient spends queueing for a large MRI; it says nothing about how much pathology a scan can detect.

NEURO-PMR answers a different question. Independent blinded reads agreed with conventional MRI on identifying pathology 92% of the time, rising to 98% once clinical history was included, and patients were four times more likely to choose portable MRI.[1] These figures come from set-piece controlled studies and do not imply the same level across every condition in routine clinical practice.

How the company is selling those readings

Hyperfine already uses these readings as sales material. In the office setting, placements since launch have passed a dozen, and office staff can operate the scanner without an MR technologist.[1] After receiving CE and UKCA marks, the company sold its first two Model 2 systems in Europe and the UK.[1]

The hospital side runs on ledger logic. A one to one-and-a-half year payback estimate converts shorter waits and higher diagnostic concordance into a purchase rationale, and it is markedly shorter than the three to four years typical for capital equipment.[4] The financial line this chain runs into is revenue, specifically device system sales revenue, where added placements and newly opened markets land first.

What is still not separated out

The unit sales, revenue and gross margin the company discloses are the result of the whole scanner plus the entire commercial effort, and management has never carved out the part belonging to this AI software. The 1.28 hours likewise comes from the portable device plus the clinical workflow, with hardware signal-to-noise improvements, scan sequence development and on-site scheduling all mixed in.[3] The hospital payback period also remains a buyer's estimate rather than a verified realized return.[4]

What can be confirmed as an operating change is that this software pushed images from a low-field scanner to the point of being clinically readable, and that the main evidence the company puts on the table when it sells a machine has shifted from clearer images to emergency-department time and diagnostic concordance. What it is worth on its own will stay unanswerable until the company discloses the unit count and average selling price of systems carrying the new-generation software within its system sales, or gives a software-related recurring revenue measure.

Application assessment

• Swoop AI Brain Imaging | Business position: Core business | Deployment stage: Limited production | Scope: Multiple businesses or regions | Value type: Revenue growth

Sources

[1] Drillr · Hyperfine, Inc. (HYPR) · 2026-08-06 · Earnings call

Original: Prime showed that portable MRI reduced the median order to scan start time in the ED from 7.76 hours for conventional MRI to 1.28 hours with portable MRI.

[2] Drillr · Hyperfine, Inc. (HYPR) · 2023-11-10 · Earnings call

Original: This expands our AID noise and capabilities by incorporating advanced image post-processing into the DWI sequence. The system's other sequences, T1, T2, and FLAIR previously benefited from this AI feature.

[3] Drillr · Hyperfine, Inc. (HYPR) · 2025-08-13 · Earnings call

Original: Optive AI software was also cleared as a standalone software, and we have now begun rolling it out to our installed base of Swoop systems bringing significantly improved image quality to our users.

[4] Drillr · Hyperfine, Inc. (HYPR) · 2025-11-13 · Earnings call

Original: Hospitals' return on investment assessments are showing one to one and a half year breakeven timeline versus three to four years typical for capital equipment.