Neurotech & BCI
A real, FDA-cleared portable-MRI razor finally finding its blade — Q1 device units +67% and runway into 2028 de-risk the going-concern story, but at ~$22M revenue and a ~$140M cap it is still a sub-scale, founder-controlled call option on point-of-care neuro-imaging, not yet a business.
Research
The verdict
A real, FDA-cleared portable-MRI razor finally finding its blade — Q1 device units +67% and runway into 2028 de-risk the going-concern story, but at ~$22M revenue and a ~$140M cap it is still a sub-scale, founder-controlled call option on point-of-care neuro-imaging, not yet a business.
Hyperfine, Inc. (Nasdaq: HYPR; Guilford, Connecticut) sells the Swoop Portable MR Imaging System — the first and still the only FDA-cleared portable, ultra-low-field MRI for brain imaging at the point of care. The pitch is a category inversion: conventional 1.5–3 Tesla MRI is a fixed, shielded, multi-million-dollar room that the patient must be wheeled to; Swoop is a wheeled, ~iPad-operated, ULF scanner that comes to the patient — the ICU bedside, the ED, a neurology office, a stroke unit. Hyperfine claims the unit is ~20× cheaper to build, ~10× lighter, and uses ~35× less power than a 1.5T scanner.
How it makes money. Two lines, both grounded in the Q1 2026 income statement:
The model is explicitly a razor-and-blades aspiration: land the device, then compound on recurring scan-driven service and software (the eleventh-generation Optive AI software, cleared Dec 2025, is the blade). Today it is still mostly razor — device is ~83% of revenue.
Customers. Hospitals, health systems, ICUs/EDs, neurology clinics, and research/academic institutions. Concentration is real but improving: in Q1 2026, three customers each >10% of revenue ($521K, $508K, $461K); in Q1 2025 it took five customers to cross that bar. No single customer is a majority — the risk is thin total volume, not one whale.
Suppliers / contract structure. Single exclusive contract manufacturer Benchmark Electronics, Inc. under a 2018 Manufacture & Supply Agreement; the Swoop magnet is a custom component from a single-source supplier in Europe. Payment terms are ordinary capital-equipment + service contracts; no take-or-pay, deferred revenue only $2.3M total — i.e. little locked-in backlog.
Upstream → Hyperfine → end customer, named where the filing names them:
Chokepoints: the European single-source magnet and the single contract manufacturer are the two points where a disruption stops the whole company. Hyperfine itself flags both as material risks. The AI software, by contrast, is fully owned and is where the durable value sits.
What is genuinely defensible:
Bargaining power — weak, honestly. Hyperfine sits between a single contract manufacturer / single-source magnet supplier (upstream power over it) and large hospital buyers who can simply keep using the high-field MRI they already own (downstream power over it). Its leverage is the unique category, not scale.
The competitive frame. The 10-K names the incumbents directly: GE HealthCare, Siemens Healthineers, Philips, Hologic, Fuji, Toshiba, Canon, Hitachi — all with vastly greater resources. In the portable/low-field niche the nearer peers are Promaxo (0.066T single-sided, prostate/urology, private, ~$99M raised) and Synaptive Medical (Evry 0.5T head/neck). Hyperfine's edge over those two: it owns the bedside-portable brain use case and the FDA + AI stack; their edge: different anatomy (prostate, intra-op) and, for the incumbents, the entire installed base of diagnostic-grade high-field machines.
No reportable operating segments — Hyperfine runs as a single segment. Disaggregation is by product type and geography:
By product (revenue, $000):
| Line | Q1 2026 | Q1 2025 | YoY |
|---|---|---|---|
| Device (point-in-time) | 3,257 | 1,522 | +114% |
| Service (over-time) | 646 | 615 | +5% |
| Total | 3,903 | 2,137 | +82.6% |
FY2025 total sales $13.563M vs FY2024 $12.890M (+5.2%).
By geography:
The trend that matters — units re-accelerated after a down year. FY2025 sold 38 systems, −20.8% vs FY2024's 48 (the next-gen launch mid-2025 created an air-pocket as buyers waited). Then Q1 2026 sold 10 commercial systems vs 6 in Q1 2025 (+67%). So the segment story is: FY2025 = price-driven (ASP up, units down), Q1 2026 = the unit count turning back up with the new product in market. That inflection is the whole bull case.
The strongest print in the company's short history:
Flag vs its own history: the +83% revenue and unit re-acceleration are a genuine break from FY2025's flatline; the tell that the market wants is raised guidance, which did not come.
No transcripts on disk (transcripts/ empty) — this lens is ``. Across the last several quarters management's framing has shifted from "building clinical evidence / awaiting next-gen clearance" (2024–H1 2025) to "accelerating commercial engine / scan-volume milestones / runway secured" (Q1 2026). The recurring new vocabulary is "sites of care," "scan volume," "stroke program," "cash runway into 2028". What they've stopped leaning on: the going-concern hedge — the March-2026 Horizon debt facility let them pivot the narrative from survival to growth. The June-18 standalone scan-volume press release (CHRISTUS >500 scans/9 months) is management deliberately seeding the utilization story between prints. Tone: cautiously confident, still pre-profit-honest.
Hyperfine is a sub-scale, pre-profit single-product device maker; classic EV/EBIT and P/E are n/a (negative earnings). The honest comp set is portable/low-field MRI peers (mostly private) and, for scale context, the imaging incumbents.
| Company | Ticker | Mkt cap | EV/Sales | P/E | Note |
|---|---|---|---|---|---|
| Hyperfine | HYPR | ~$140M | ~6–7× FY26E rev ($20–22M) | n/a (loss-making) | Only FDA-cleared portable ULF brain MRI |
| Promaxo | private | n/a — private | n/a | n/a | 0.066T prostate; ~$99M raised over 13 rounds |
| Synaptive Medical | private | n/a — private | n/a | n/a | Evry 0.5T head/neck |
| GE HealthCare | GEHC | ~mega-cap | n/a | n/a | Imaging incumbent |
| Siemens Healthineers | SHL (DE) | ~mega-cap | n/a | n/a | Imaging incumbent |
| Philips | PHG | ~large-cap | n/a | n/a | Imaging incumbent |
| Canon Medical | (Canon) | ~large-cap | n/a | n/a | Imaging incumbent |
Provenance discipline: I am not fabricating incumbent multiples — they are n/a here. The only defensible valuation anchor is HYPR's own ~6–7× forward sales, which is a story multiple, justified only if the +55% growth compounds for years. Portable-MRI TAM context: ~$4.4B 2025, ~5.9% CAGR — but that figure bundles incumbent mobile-MRI trailers; Hyperfine's true ULF-bedside sub-segment is far smaller and earlier.
HYPR is a sub-$2 micro-cap; it moves on liquidity, dilution, and clinical/regulatory headlines more than fundamentals. Pattern over the last ~2 years:
CEO — Maria Sainz (President & CEO since Oct 24, 2022; offer letter Oct 4 2022 per 10-K exhibit 10.3 ). This is the single most underrated asset in the story. ~30-year medtech operator with a serial-exit record: CEO of Concentric Medical (sold to Stryker, 2011), AEGEA Medical (sold to CooperSurgical, 2021), Cardiokinetix; a decade at Guidant; board seats at Spectranetics (sold to Philips, ~$2B, 2017) and Shockwave Medical (acquired by J&J, ~$13B, 2024). She is a commercialization-and-sell-the-company specialist parachuted into a founder-science company — a strong tell about the likely terminal outcome (strategic acquisition, not independent compounding).
CFO — Brett Hale (CFO / Chief Administrative Officer, Treasurer, Corporate Secretary). Co-signs the financing architecture (ATM + registered directs + the Horizon term loan).
Founder / Chair-adjacent — Jonathan Rothberg, Ph.D. Founded Hyperfine 2014 inside his 4Catalyzer incubator; board member since the 2021 SPAC close. One of the most decorated entrepreneurs in the space: National Medal of Technology (high-speed DNA sequencing — 454 Life Sciences, Ion Torrent; also founded Butterfly Network, the portable-ultrasound analog). This is a serial deep-tech founder with a pattern of building category-creating devices — and a pattern of long, dilutive, slow-to-profit commercial ramps (Butterfly is the cautionary mirror).
Capital allocation: consistent with stage — they have funded a ~$330M+ cumulative deficit with equity, then finally added non-dilutive-ish debt (Horizon, March 2026) once revenue could service interest-only payments. The discipline tell is the FY2025 opex cut (−10.7% to $43.8M) and the Q1 R&D roll-off — management is steering toward the $26–28M burn / runway-into-2028 plan, not spending wildly.
Red flags (governance): dual-class control. Rothberg beneficially owns all 15,055,288 Class B shares at 20 votes each = ~78.9% voting power despite a minority economic stake, making Hyperfine a "controlled company" exempt from several Nasdaq governance requirements. He continues to receive option grants (Form 4: 157,700 options @ $1.49). Public minority holders have economics without control. Archetype: founder-controlled, professional-CEO-operated — the right combination for this stage, but the voting asymmetry is a permanent governance discount.
Acting as a forensic analyst on the filings:
Regulatory findings (required sub-section):
Built bottom-up from FY2025 actuals + management's FY2026 guide. Hyperfine is loss-making, so the scoreable metric is revenue / runway, not positive EPS (EPS stays negative across the window in every case).
Anchors: FY2025 revenue $13.563M; FY2026 guide $20–22M (mid $21M, ~+55%), GM 50–55%, burn $26–28M, runway into 2028; FY2025 EPS $(0.43); ~98M weighted shares and climbing.
Base case (FY2026 → FY2028):
Bull case: units compound 40%+ for three years, service/software attach inflects (recurring becomes >25% of revenue), an incumbent (GE/Philips/Siemens) or strategic acquires — FY28 revenue $55M+ and a takeout well above the current ~$140M cap.
Bear case: the Q1 unit re-acceleration proves to be a next-gen restocking bump, NIH/research-funding pressure and hospital capex caution flatten US placements, FY26 lands at the low end ($20M) or below, burn forces a dilutive raise at a depressed price, and the stock round-trips toward the $0.66 low.
Per the SKILL, the Brier forecast.ts create step is skipped in the unattended --watchlist loop — logged here as a note only: base-case scoreable claim would be "HYPR FY2026 revenue ≥ $20M, p≈0.80, resolves 2027-03."
Bull case. Hyperfine owns a genuine category — the only FDA-cleared portable ULF brain MRI — at the exact moment its commercial engine is turning over. Q1 2026's +83% revenue and +67% units suggest the next-gen Swoop + Optive AI finally cleared the adoption hurdle, and the scan-volume milestones (CHRISTUS >500 scans/9 months) show the razor-and-blades flywheel starting to spin in stroke and neurology workflows. The AI-reconstruction stack is a compounding data moat that incumbents can't shortcut, the IP runs to 2043, the Horizon debt facility bought runway into 2028 without crushing dilution, and a CEO with a serial-medtech-exit résumé is exactly who you want if the endgame is a strategic sale. At ~$140M, you are paying ~6–7× forward sales for a real, growing, defensible franchise with optionality on becoming the standard of care for bedside neuro-imaging — and a credible takeout target for GE/Philips/Siemens.
Bear case. Three things could permanently impair this: (1) it never reaches scale — at $20–22M revenue against a ~$45M cost base, the math only works if growth compounds for years, and a single soft year (US hospital capex caution + the explicit NIH research-funding freeze hitting academic buyers ) breaks the narrative and forces a dilutive raise; (2) the incumbents wake up — GE/Siemens/Philips have the channel, the installed base, and the balance sheet to build or buy a "good-enough" portable and bundle it, collapsing Hyperfine's category-of-one premium; (3) single points of failure — one European magnet supplier, one contract manufacturer, any disruption is existential. Pre-mortem (18 months out, thesis broke): FY2026 came in at the low end, the unit re-acceleration was a restocking blip, NIH/hospital budgets stayed tight, cash hit the danger zone, and they raised equity at ~$1 — diluting holders ~25% again and re-rating the multiple down. Contrarian view the market is missing: the bull obsession is image quality and TAM; the thing that actually decides this is CMS reimbursement economics per scan — if a Swoop scan reimburses well enough that a stroke/ED program pays back the device in <18 months, adoption is unstoppable; if it doesn't, no amount of AI sharpness matters. That's the real swing variable, and it's underdiscussed.
Dismantling the bull case:
The safest, most surgically scalable path into the brain and the only BCI Apple made a native input — but it bet the company on "good enough" 16-electrode bandwidth, and its own third-gen "whole-brain" pivot is a confession that the moat it is famous for may be the ceiling it has to escape.
A first-mover NGPS tools story whose commercial engine is going backwards (revenue −20% in FY25, −69% in Q1'26 to $258K) while it bets the company on Proteus by end-2026; ~$190M cash funds the bet, but a 69%-Rothberg-controlled sub-$1 microcap with collapsing instrument demand is a binary option, not an investment — WATCHING until Proteus ships and consumable pull-through proves real.
The only FDA-cleared, commercially-shipping cortical BCI — but it is selling a 30-day surgical-monitoring tool, not the chronic implant the $500M valuation is priced on; Medtronic is the real tell, IP overhang from Rapoport's Neuralink past is the real risk.