A-Portfolio

The company in one breath

Hyperfine, ticker HYPR on the Nasdaq Global Market, sells Swoop, an artificial-intelligence-powered brain MRI scanner that rolls into hospitals and neurology offices instead of making sick patients travel to a giant machine.

The pitch is easy. Swoop needs no shielded room, helium, or specialized MRI technician, and U.S. scans can use existing reimbursement code CPT 70551. That means a clinic can add brain imaging without building a bunker. Good would look like hospitals buying several units, offices scanning enough patients to earn an attractive return, and service or software revenue following each machine.

Swoop does not replace a full-strength MRI. It is an ultra-low-field machine for situations where a conventional scan is impractical or too slow. Good would look like it becoming the standard first look in emergency rooms, intensive-care units, and neurology offices, then sending only harder cases to the big machine.

Why now

Preliminary third-quarter revenue reached about $6.0 million, up 53% from the prior quarter and roughly 72% from one year earlier. Revenue is customer money before expenses, and this was Hyperfine's best quarter yet. Good for a tiny medical-device company would look like growth above 30% without cash burn rising, so this clears the growth bar.

The product also has evidence beyond a glossy booth at a medical conference. In the 100-patient PRIME study, Swoop cut the median wait from an imaging order to the start of a scan from 7.76 hours to 1.28 hours. That means patients reached brain imaging about 6.5 hours sooner. Good would look like a larger, multi-hospital study repeating that result and proving faster imaging improves care or lowers hospital costs.

The commercial footprint is starting to look real. Hyperfine reports more than 225 installed systems and about 250,000 images captured worldwide. That means the product has left the laboratory. Good would look like installed systems climbing at least 25% a year, usage rising at existing sites, and recurring service revenue growing faster than the machine count.

The stock closed at $0.9064 on October 12. That gives the company a market value of about $96.1 million, which is what the stock market says all common shares are worth. Good entry quality would mean paying a modest price for fast growth without inheriting an immediate financing emergency. Hyperfine has the modest price. The financing question still has teeth.

The numbers, translated

• Nine-month revenue was about $13.8 million, up 66%. That means customers spent two-thirds more than they did in the same stretch last year. Good would look like full-year revenue landing at or above the $21 million midpoint of management's forecast, which requires about $7.2 million in the fourth quarter.
• Second-quarter gross margin was 50.7%. Gross margin is what remained from each sales dollar after building and servicing the product, before payroll, research, and selling costs. Hyperfine kept about 51 cents. Good for a scaling medical-device company would look above 55 cents, then moving toward 60 cents as manufacturing volume rises.
• Preliminary third-quarter cash burn was about $5.6 million, down 29% from the prior quarter. Cash burn is the money that left the bank to run the business. Good would look below $5 million a quarter while revenue keeps growing, then revenue overtaking burn for good.
• Cash was about $37.9 million on September 30. That is nearly seven quarters of runway at the latest quarterly burn rate before interest, working-capital swings, or new investment change the math. Good would look like the cash lasting until operations fund themselves, not merely reaching the next stock sale.
• Hyperfine had about $15 million of debt and another $25 million available if it meets loan conditions. Debt buys time but gives lenders a claim ahead of shareholders. Good would look like never drawing the extra amount because sales growth and lower burn close the gap.
• About 106.0 million shares were outstanding, up roughly 24% in one year. More shares mean each old share owns less of the company. Good would look like annual share growth below 5%, unless every new share clearly creates more value than it takes.
• The operating business is valued near $73.2 million after adding debt and subtracting cash. Investors call that enterprise value, the rough price of the business itself. It equals about 3.5 times the midpoint of expected 2026 revenue. Good would look below 3 times for a cash-burning hardware company, or a higher multiple backed by repeated growth above 40% and a clear path to profit.
• The stock sits below Nasdaq's usual $1 minimum bid price. That means the listing can become a problem if the price stays under a dollar long enough. Good would look like the business lifting the stock above $1 naturally, without a reverse split that merely changes the number printed on the screen.

What I like

The category lead is real. Hyperfine says Swoop is the first FDA-cleared portable ultra-low-field brain MRI, and the company holds roughly 200 issued patents. That means competitors cannot copy the exact recipe without legal and engineering work. Good would look like those patents protecting better image reconstruction, easier workflows, and new clinical uses while Hyperfine builds customer habits.

The machine solves a pain that administrators understand. PRIME found a wait of 1.28 hours with portable MRI versus 7.76 hours with conventional MRI. Nobody needs an MBA to understand six hours of emergency-room gridlock. Good would look like buyers documenting shorter stays, fewer patient transfers, and a payback period under two years.

The office market could widen the runway. More than 3,000 office scans occurred during the first year of that launch, and patients in the NEURO PMR study preferred portable MRI by four to one. That means convenience has clinical pull, not just investor-deck charm. Good would look like busy neurology practices ordering repeat systems and reimbursement covering the economics without heroic assumptions.

Management has started matching spending to reality. Revenue rose sharply while quarterly cash burn fell. That is operating leverage, which means each new sales dollar does more work instead of hiring an entourage. Good would look like full-year revenue above $21 million, gross margin above 52%, and annual burn below $26 million.

What worries me

The company still burns more cash than it brings in. Third-quarter revenue was $6.0 million and cash burn was $5.6 million, before counting the cost already embedded in making those sales. Good would look like quarterly gross profit covering a growing share of operating expenses until burn falls below $3 million. Today, growth is improving the math, not finishing it.

Dilution has been brutal. The share count rose about 24% in one year, and the company still has an at-the-market program that can sell shares into public trading. Good would look like no new equity through the November report and share growth below 5% next year. A brilliant scanner can still be a lousy stock if management keeps slicing the ownership pie thinner.

The product is an adjunct, not a replacement for conventional MRI. Lower magnetic strength can mean lower detail, and the FDA label limits Swoop to brain imaging where a full diagnostic examination is not practical. Good would look like independent studies proving it reliably answers defined clinical questions, not marketing trying to turn a pocketknife into a whole kitchen.

Competition can arrive with a bigger sales force. GE HealthCare (GEHC), Philips (PHG), and other imaging giants already know every hospital purchasing department on earth. Good would look like Hyperfine locking in workflows, evidence, patents, and multi-unit health-system deals before an incumbent decides this niche is worth owning.

Public discussion is thin. Thirty-day research found only two company-specific Reddit posts, each with one upvote, and no useful YouTube, Hacker News, GitHub, or prediction-market evidence. Good would look like radiologists and hospital buyers debating image quality, economics, and repeat use. Right now, the crowd is not early. It is mostly absent.

What would make this a buy

My position size is zero. This is research and commentary, never a buy signal.

I want the November 9 report to confirm the preliminary quarter. Revenue needs to stay near $6.0 million, gross margin needs to hold above 52%, and quarterly cash burn needs to stay at or below $5.6 million. Good would mean the jump came from broad demand, not one order that moved between quarters.

Then I want fourth-quarter revenue of at least $7.2 million. That would put full-year revenue at the $21 million guidance midpoint and prove the record quarter was not the finish line. I also want the contrast-MRI submission filed by year-end, because approval could widen what offices and surgical teams can do with Swoop.

Price matters. At $0.85, the operating business would cost about $67.2 million, or roughly 3.2 times midpoint revenue after cash and debt. That is not dirt cheap, but it gives more room for a delayed order or slower adoption. Above $1.00, I need confirmed fourth-quarter growth, lower burn, and no fresh equity sale. Hope does not get a premium multiple on my tab.

What makes me walk away

I walk if fourth-quarter revenue falls below $5.5 million. That would mean the record quarter likely borrowed demand from the future rather than starting a durable climb.

I walk if gross margin drops below 48%. That would mean pricing, manufacturing, or product mix moved the wrong way just as scale should help.

I walk if cash falls below $25 million before quarterly burn drops under $4 million, or if the share count rises another 15% before the company reaches positive cash flow. That combination would make existing shareholders the financing plan.

I also walk if the contrast submission slips past the first quarter of 2027 without a precise clinical reason, or if larger studies fail to repeat PRIME's workflow benefit. The thesis needs widening clinical use and repeatable economics. Lose either one and this becomes an impressive machine attached to a weak business.

IPO Watch

• Iambic Therapeutics, private, proposed ticker IAM, expected October 15 at $15 to $17. It uses artificial intelligence and automated laboratories to design cancer drugs. I would write it up when the final valuation leaves room for clinical failure and the lead HER2 program shows convincing tumor responses in people.
• City Therapeutics, private, proposed ticker CTY, expected October 16 at $17 to $19. It develops RNA-interference medicines that silence disease-causing genes. I would write it up when human data shows one program can turn a proven scientific method into a differentiated drug, not another platform speech with a lab coat.

The verdict

Score: 68 out of 100. Bucket: watch. Hyperfine earned points for a genuine category lead, fast revenue growth, improving cash burn, gross margin above 50%, existing reimbursement, strong early workflow data, an experienced medical-device chief executive, and clear catalysts in contrast imaging and international sales.

It lost more for tiny scale, large losses, only $37.9 million of cash, debt, roughly 24% yearly dilution, a sub-dollar share price, single-product risk, giant competitors, and weak independent discussion. A candidate starts at 80. Hyperfine sits twelve points short. The scanner has earned a deeper look. The financing model has not earned trust.

The DBTech read

Enterprise value is the price of the operating business after you add debt and subtract cash. Hyperfine's stock is worth about $96.1 million, but its cash offsets part of that price and its debt adds some back, leaving roughly $73.2 million. Good enterprise value gives you a cleaner way to compare companies with different bank balances. Think of buying a bar, you pay for the building, assume its loan, and keep the cash in the register. The sticker price alone never tells the whole story.

Sources and as-of date

• Price: $0.9064 at 4:00 p.m. Eastern on October 12, 2026, the latest completed regular-session close, with a $0.93 open, $0.93 high, $0.8756 low, and 990,975 shares traded, The Wall Street Journal. Retrieved at 7:07 a.m. Eastern on October 13, 2026. Robinhood (HOOD) showed HYPR near $0.88 with an approximately $96.22 million market value during the October 13 premarket page refresh, Robinhood. The completed $0.9064 close drives every valuation here. Read-only Robinhood MCP quote and option-chain tools were not available in this run, so no contract, fill, bid, ask, or Greek is assumed.
• Preliminary third-quarter revenue, nine-month growth, cash, cash burn, full-year guidance, November 9 reporting date, installed base, patents, clinical evidence, and catalysts: Hyperfine October 8 Form 8-K and exhibits. The third-quarter figures are company estimates and have not completed the normal quarterly review.
• Second-quarter revenue, gross margin, losses, debt, share count, at-the-market stock sales, liquidity, business description, reimbursement, competition, and risks: Hyperfine second-quarter Form 10-Q, filed August 7, 2026.
• PRIME study design and results: Hyperfine PRIME release and NeuroNews International, retrieved October 13, 2026. The study enrolled 100 patients at one center, so broader proof still matters.
• Office scans, patient preference, clinical concordance, product label, and current positioning: Hyperfine office-market update and Swoop reimbursement resources, retrieved October 13, 2026.
• IPO dates and ranges: Iambic Therapeutics amended S-1, City Therapeutics amended S-1, and MetaCap IPO calendar, retrieved October 13, 2026. Expected dates and ranges can move until pricing.
• Community research covered September 13 through October 13, 2026. It found 8 Reddit threads, but only two tiny posts were company-specific. YouTube, Hacker News, GitHub, and prediction markets produced no useful company evidence. Raw research: workspace/newsletters/a-portfolio/research/hyperfine-hypr-swoop-portable-mri-raw-v3.md.
• Calculations use decimal arithmetic, the $0.9064 close, 106.04 million shares, $37.9 million of preliminary cash, $15 million of debt principal, and the $21 million midpoint of 2026 revenue guidance. Research and commentary only, never a buy signal.