Medical devices & diagnostics · Point of care ultrasound
Lumius
A Duke PhD spinout building a low cost 3D ultrasound attachment meant to turn a standard 2D probe into what its founders call “a 3D camera for the body,” starting with central line placement. The company has a peer reviewed technical paper behind it, a Y Combinator badge, and, as of this writing, no product on the market and no FDA filing.
Three PhDs and a shared frustration
Tri Vu, Luca Menozzi and Chenhang Li met at Duke, where all three earned doctorates in ultrasound and wearable sensing labs. “Luca Menozzi, Chenhang Li, and I bonded over ultrasound during our PhDs at Duke,” one of the founders wrote in the company’s Y Combinator launch post. “For years, it has bothered us that ultrasound looks largely the same as it did decades ago.”
Vu and Menozzi both trained in the Photoacoustic Imaging Lab run by Junjie Yao at Duke’s Pratt School of Engineering. Menozzi is the first author, and Vu a co-author, on a January 2025 Nature Communications paper describing three-dimensional diffractive acoustic tomography, or 3D-DAT: a way to pull near-isotropic 3D images out of an ordinary, off-the-shelf linear array transducer by adding a slit that diffracts the returning sound wave, then reconstructing the volume on a GPU. The paper reports a synthetic aperture of more than 50,000 elements from hardware that normally has about 1,000, and a reconstruction speed 50 times faster than prior methods, demonstrated on frogs, mice and tumor models rather than people. Li, trained in mechanical engineering and materials science at Duke and Northwestern, works on wearable ultrasound elastography and leads the mixed reality visualization side of the product, per an investor account of the founding team.
The problem: 2D ultrasound and a coin flip
Ultrasound is used an estimated 500 million times a year, and most of it is still two dimensional, according to the company. A clinician moves a flat probe over the body and has to mentally assemble a 3D picture of the anatomy underneath while guiding a needle or catheter. For central venous catheterization, the company cites a roughly 50 percent first attempt failure rate, especially among newer clinicians, and hospitals spend heavily every year training staff on 2D probes. True 3D ultrasound already exists, mostly for cardiac and fetal imaging, but it uses dense matrix array transducers that cost far more than standard 2D probes. Lumius is pitching its diffraction based approach as a way to get 3D imaging out of hardware clinics already own.
From lab bench to Y Combinator
Duke’s Christensen Center for Innovation says the company grew directly out of a Duke research project, supported by the Center’s PhD Research Accelerator and a Duke Investor Showcase. “What began as a Duke research project has evolved into a promising medical imaging startup, Lumius Imaging, with the potential to expand access to fast, intelligent, and affordable 3D ultrasound technology,” the center wrote in June 2026, adding that the team was “currently raising their seed round” at the time.
Lumius went through Y Combinator’s Spring 2026 batch as a four person team based in Durham. Tracxn and Dealroom both list Lumius as a seed stage company with about $500K in disclosed funding, without naming a lead investor or a closing date, and no SEC Form D notice for a company under this name turned up in searches for this profile.
What the pitch adds up to now
An investor writeup published by 007 Venture Partners in June 2026, announcing its own investment, is the most detailed public account of where the company stands. It describes a working hardware prototype and an AI assisted, mixed reality layer, built within “roughly seven weeks of joining Y Combinator,” that overlays segmented 3D anatomy onto a patient through a headset. It also lays out a roadmap: an FDA pre-submission meeting targeted for Q3 2026, an internal research-lab device by Q4 2026, a comparison study against 2D ultrasound in early 2027, and, if that goes well, 510(k) clearance and a clinical launch target of Q1 2028. None of those milestones have been independently confirmed as complete.
What is proven, and what is still claimed
| Evidence | What the record shows | Source type |
|---|---|---|
| Underlying science | “Three-dimensional diffractive acoustic tomography” (3D-DAT), Menozzi, Vu et al., Nature Communications, published online Jan 29, 2025 (PMID 39880853). Demonstrates 3D photoacoustic and ultrasound imaging from a standard linear array transducer in phantoms and small animals; no human clinical data. | Public record |
| Federal funding behind the science | The Nature Communications paper acknowledges several NIH grants (including RF1 NS115581 and R01 EB028143) and an NSF CAREER award, all to Duke professor Junjie Yao, plus a separate NSF Graduate Research Fellowship to co-founder Luca Menozzi. These are academic grants to a Duke lab and a graduate student, not awards to the Lumius corporate entity. | Public record |
| FDA clearance | No 510(k), De Novo or PMA record found for Lumius. The company’s Y Combinator launch post and a partner LinkedIn post both describe the device as under development and not FDA cleared. | Not found |
| Federal contracts or SBIR awards to the company | Lumius Imaging, Inc. is a registered federal contractor (CAGE 114D5) with no contract or award history on record. No NIH or NSF award to the company was found. A $370,000 ARPA-H subcontract is claimed in one investor account; it does not appear in federal award databases searched for this profile. | Not found |
| Funding round | Tracxn and Dealroom both report a Seed round totaling about $500,000, with Y Combinator, FundersClub, Boost VC, Robinhood Ventures and Vento Ventures named as investors in a June 2026 investor blog post. No SEC Form D for a “Lumius” or “Lumius Imaging” entity was found. | Differs across sources |
| Cost claim | The company’s homepage says the device “costs 10 times less” than a matrix array 3D probe; its solution page says “50 times less.” A backing investor separately cites a $5,000 cost of goods against a $30,000 target price, “roughly 13x” cheaper than high end incumbents. The three figures do not agree. | Company materials disagree |
| Patents | An investor account describes a “foundational Duke University patent estate” plus a Lumius provisional patent covering the mixed reality interface, with Menozzi, Vu and Li named as inventors. No independent patent filing was located to confirm this for this profile. | Company-stated |
| Clinical trials | No registered study for Lumius or its device found on ClinicalTrials.gov. | Not found |
| Commercial pull | The same investor account cites “$5.7M in signed LOIs” and a pipeline including “credible names in surgical robotics and medical devices,” without naming any of them. | Company-stated |
Read plainly: the physics behind Lumius is published and peer reviewed, in animals, not people, and it belongs to a Duke lab as much as to the company. Everything about the product, the funding, the patents and the customer pipeline currently rests on the company’s own materials and on one investor’s account of why it wrote a check. None of it has been tested against an FDA filing, a Form D, or a named customer yet.
What to watch
- Whether the FDA pre-submission meeting an investor placed in Q3 2026 happens, and what the agency says about pathway (510(k) versus De Novo).
- A first SEC Form D or amended filing that would show the size, timing and investor list of the seed round independently of the company and its backers.
- Human data. Everything published so far is bench, phantom and small-animal work; the company’s own roadmap puts a head-to-head study against 2D ultrasound in early 2027.
- Named hospital or clinical customers behind the reported letters of intent.
In their words
“Luca Menozzi, Chenhang Li, and I bonded over ultrasound during our PhDs at Duke. For years, it has bothered us that ultrasound looks largely the same as it did decades ago.”
Lumius co-founder, Y Combinator launch post, 2026 · Company-stated
“We're starting with vascular access, particularly central line placement. This is a procedure where ~50% of first attempts fail, especially among new clinicians.”
Lumius co-founder, Y Combinator launch post, 2026 · Company-stated
“What began as a Duke research project has evolved into a promising medical imaging startup, Lumius Imaging, with the potential to expand access to fast, intelligent, and affordable 3D ultrasound technology.”
Christensen Center for Innovation, Duke Pratt School of Engineering, June 2026 · Independent
“There is zero evidence of a working product, prototype performance data, clinical partnerships, or any traction whatsoever.”
The YC Tier List, unofficial batch review, 2026 · Independent
“The science isn't a pitch deck claim. It's published.”
007 Venture Partners, investor blog post announcing its investment, June 2026 · Investor-stated
“Lumius is currently under development and not FDA cleared.”
Y Combinator company update, LinkedIn, May 2026 · Accelerator-stated
Related companies
Sources
- Public recordThree-dimensional diffractive acoustic tomography, Menozzi, Vu, Canning et al.
- Public recordThree-dimensional diffractive acoustic tomography, PMID 39880853
- Public recordLUMIUS IMAGING, INC federal contractor registration, CAGE 114D5
- IndependentDuke Alumni Startup Lumius Imaging Accepted Into Y Combinator
- IndependentTri Vu, faculty and PhD bio page
- IndependentLumius, The YC Tier List batch review
- IndependentLumius company profile
- IndependentLumius company profile and funding summary
- AcceleratorLumius company page
- AcceleratorY Combinator company update on Lumius
- CompanyLaunch YC: Lumius, The 3D Camera for the Body
- InvestorLumius: Fast, smart, accessible 3D ultrasound that anyone can use
- CompanyLumius homepage and solution page
Profile researched and written by Healthcare Discovery. Last updated September 29, 2026.
