A staged deep-tech venture that begins as a simple, safe, affordable handheld light — and evolves, wavelength by wavelength, into a radiation-free optical breast imaging ecosystem.
Concept, invention & design — Amir Kayvanpour
Most breast-imaging ventures try to build the final clinical machine first — burning years and tens of millions before their first sale. LUMÉA inverts that. We ship a safe, useful consumer light in Year 1, then climb a deliberate 10-stage ladder toward clinical-grade optical imaging, funding each rung with revenue, data, IP, and evidence from the one below it.
A handheld multi-wavelength illuminator that lets a person directly see vascular contrast in their own tissue. No camera, no AI, no diagnosis — a wellness, education and self-familiarity device positioned honestly.
Each stage adds one capability — sensing, capture, spectra, function, calibration, 3D, AI, cloud — moving from home, to GP and women's-health clinics, to breast clinics, imaging centres and hospitals.
Skin-tone-adaptive illumination, breast-specific optical contact geometry, and a longitudinal optical dataset that compounds — the training fuel and IP that a late entrant cannot buy.
Dense-breast screening gaps, radiation fatigue, femtech capital, cheap NIR optics, and on-device AI have finally converged.
Optical + embedded + regulatory + clinical fluency, executed as a staged commercial machine rather than a science project.
Lower cost, faster MVPs, early revenue, lighter early regulation, compounding data & IP, and investor proof at every gate.
Seed capital to ship Stage 1–2, file foundational IP, and reach the clinical-grade inflection at Stage 6.
Mammography saves lives and remains the only screening method proven to reduce breast-cancer mortality — a fact we will never dispute. But it is an ionizing, compressive, infrastructure-heavy exam that performs worst exactly where risk is highest: the dense breast. That is the gap safer, cheaper, more frequent optical tools can help fill — as complements, not replacements.
Mammography relies on low-dose X-rays. Dose is small per exam, but it limits how young and how often people are comfortable screening, and rules out casual or high-frequency monitoring.
Breast compression is a widely cited deterrent to attendance. Discomfort and anxiety measurably suppress screening participation.
In extremely dense breasts, mammographic sensitivity can fall from ~90%+ in fatty breasts to roughly 40% in the densest tissue — while those same women carry a 2–3× higher cancer risk.78
Mammography needs fixed capital equipment, shielding, and trained radiographers — scarce in rural, remote and low-resource settings, and in much of the Global South.
Screening is a snapshot every 1–2 years. There is no practical, safe tool for the frequent, longitudinal self-monitoring many patients want between exams.
Many people want to engage with their own breast health but have no safe, non-diagnostic tool between clinical visits — a vacuum currently filled by unvalidated products.
Reported mammographic sensitivity by BI-RADS density. Ranges vary by study and design; comparative studies against a supplemental modality report the lowest figures in the densest breasts.8
Breast cancer is now the most commonly diagnosed cancer worldwide, at roughly 2.3 million new cases each year.1 The International Agency for Research on Cancer projects global cases rising ~38% and deaths ~68% by 2050.5
About 1 in 8 women in the US will be diagnosed in their lifetime;9 roughly 48 million mammograms are performed annually in the US alone.10
Optical imaging exploits a simple truth: blood absorbs red and near-infrared light differently from surrounding tissue, so vasculature and perfusion leave an optical signature. Diffuse optical tomography and NIR spectroscopy already use this in research and adjunctive clinical settings.11 Our vision is to industrialise that physics along a commercial ladder.
Safe multi-wavelength light the eye reads directly. Consumer wellness, education, self-familiarity.
Sensors optimise contrast; cameras capture multi-spectral surface maps for tracking and education.
Calibrated optics estimate functional contrast; clinic-grade hardware enters supervised workflows.
Multi-angle reconstruction, AI decision support, and longitudinal cloud monitoring across a fleet.
The endpoint is ambitious but disciplined: a validated, radiation-free optical platform that could serve as a complementary screening and monitoring modality — and, only after major clinical evidence and regulatory clearance, a potential future alternative in specific populations such as dense-breast or high-frequency monitoring. We treat that final claim as earned, never assumed.
The graveyard of MedTech is full of companies that raised huge rounds to build the final device, then died in the gap between prototype and pivotal trial. Staged commercialisation converts that single fatal leap into ten survivable steps — each one cash-generative, evidence-generating, and independently valuable.
| Dimension | Moonshot: build the clinical device first | LUMÉA: staged ladder |
|---|---|---|
| Time to first revenue | 5–10 years (post-clearance) | 6–18 months (Stage 1 consumer) |
| Capital at risk before proof | Tens of millions, single bet | Small tranches, gated by results |
| Regulatory burden early | Full diagnostic device from day one | Wellness / accessory first, escalating deliberately |
| Customer feedback | Arrives after the design is frozen | Continuous, from Stage 1 users |
| Dataset for AI | Collected late, expensively | Compounds from the first shipped unit |
| IP creation | Concentrated, late | Layered filings across every stage |
| Investor proof points | Binary — works or doesn't | A new de-risking milestone each round |
| Failure mode | Terminal | Recoverable — pause, sell the current tier, iterate |
Our roadmap is literally a spectrum. Each stage lights up one more wavelength and one more capability — starting in the visible red (~630 nm) and climbing into the near-infrared, where blood absorption and tissue penetration open the door to functional imaging. Note: NIR is invisible; we render it here as violet→cyan for the eye.
Tap a node above or a dot to jump to any stage · the player auto-advances every 3.2s
Each card is a self-contained business — form factor, technology, target user, economics, risk, IP and the exact capability it hands to the next stage. Costs and timelines are planning estimates for a lean team and should be validated against firm quotes.
The commercial timeline, mapped to the technical ladder. Each band pairs a product release with the revenue, evidence, IP, funding, team and regulatory milestone that unlocks the next round.
LUMÉA spans a large, fast-growing consumer/femtech opportunity (early stages) and the clinical breast-imaging market (later stages). We separate verified third-party figures from our own labelled assumptions everywhere.
Independent analysts converge on roughly US$4.9–5.9B in 2024, growing to ~US$9–13.6B by 2030–2034 at a ~6.5–8.9% CAGR.34 Ionizing modalities still hold the majority (~66–74%); non-ionizing (incl. optical) is the smaller, faster-growing slice.2
⚠ = LUMÉA planning assumption (illustrative, to be validated). ✓ = anchored to a cited third-party source.
| Stage | Primary market | Buyer | Price band | Model | Comparable / adjacent |
|---|---|---|---|---|---|
| 1–2 | Consumer wellness / femtech / women's-health devices ⚠ assumption | Direct consumer | $120–$350 | D2C hardware | Home health & self-exam aids; consumer NIR/red-light devices |
| 3 | Connected consumer / prosumer | Consumer + educators | $250–$600 | Hardware + app freemium | Connected health wearables; vein-visualisation tools |
| 4–5 | Women's-health & GP clinics, wellness clinics ⚠ | Clinic owner | $2k–$12k | Hardware + SaaS | Vein finders (AccuVein-class); point-of-care optical |
| 6–7 | Breast clinics, imaging centres ✓ imaging mkt | Radiology / procurement | $25k–$120k | Capital + service + SaaS | Automated breast US; SoftVue UST12; DOT research systems11 |
| 8–9 | Hospital & enterprise (AI + cloud) ✓ | Health system / IT | per-seat / per-scan SaaS | Hardware + recurring SaaS | AI breast-imaging platforms; cloud radiology |
| 10 | Screening ecosystem ✓ | Health systems, programs | platform + reagentless recurring | Platform + data services | Full breast-imaging suites (GE, Siemens, Hologic, Philips)3 |
Trust, clarity of non-diagnostic claim, skin-tone performance, avoiding false reassurance, education burden.
Evidence bar, reimbursement, radiologist workflow fit, procurement cycles, regulatory clearance, liability.
E-commerce & educators early; women's-health & GP networks mid; medical-device distributors, imaging groups and health-system tenders late.
We benchmark honestly against every relevant modality. Mammography is the evidence leader and our reference point; our opportunity is safety, comfort, access and frequency — as a complement.
| Modality | Safety | Cost | Access | Comfort | Clinical acceptance | Reg. difficulty | Home use | Data richness |
|---|---|---|---|---|---|---|---|---|
| Mammography / DBT | Ionizing | High | Fixed sites | Low | Gold std | Established | No | Structural |
| Ultrasound / UST | Safe | Med–High | Operator-bound | Good | Adjunct | Cleared | No | Structural |
| MRI | Safe* | Very high | Scarce | Confining | High-risk std | Cleared | No | Rich |
| Thermography | Safe | Low | Moderate | Good | Adjunct only13 | Cleared (adjunct) | Limited | Thermal only |
| DOT / NIR optical | Safe | Low–Med | Portable | Good | Research/adjunct11 | Emerging | Possible | Functional |
| Vein finders | Safe | Low | Portable | Good | Accepted (venous) | Class I/II | Yes | Surface |
| LUMÉA (target) | Safe | Low→Med | Home→clinic | Good | To be earned | Staged I→II→SaMD | Yes (early) | Functional+longit. |
*MRI is radiation-free but uses gadolinium contrast in many protocols. Ratings are directional editorial judgments for positioning, not measured benchmarks.
Whole-breast UST ring; FDA PMA 2021 as dense-breast adjunct; no compression/radiation; ~2–4 min/breast.12 Our north star for the "adjunct, not replacement" pathway.
Safe and cheap, but marketed as a replacement without evidence — repeated FDA warnings since 2011.13 Every claim we make is chosen to avoid this fate.
Ultrasound-guided DOT and photoacoustic systems lead the optical research frontier as adjuncts; translation is the open challenge we industrialise toward.11
Blood's chromophores — oxy- and deoxy-haemoglobin — absorb visible-red and near-infrared light far more than surrounding lipid and water, which is why vasculature reads as darker contrast. Tissue scattering blurs and limits depth, so early products stay honest about what the eye can and can't see.
The "NIR window" where tissue is most transparent. We span amber/red (visible contrast) into NIR (~730–940 nm) for depth and haemoglobin sensitivity.
Haemoglobin absorbs red/NIR more strongly than fat/water, so vessels and vascularised regions appear as darker zones under diffuse illumination.
Diffuse scattering limits surface observation to shallow depth; skin pigmentation and ambient light degrade contrast. These bound every early claim.
Reference targets & per-device optical calibration to make readings comparable across units and time.
Photodiode/CMOS capture → denoise → flat-field → multi-spectral registration → contrast maps.
Vascular segmentation → longitudinal alignment → change detection → clinician-reviewed flags (human-in-the-loop).
Encrypted store, de-identified pipelines, role-based access, audit trails, region-resident data.
Scientific credibility is our licence to operate. We state clearly which claims physics supports today, which are engineering challenges, and which require clinical evidence we have not yet earned.
Our regulatory posture climbs in lock-step with evidence: wellness/education first, medical-device accessory next, then Class I/II devices, then software as a medical device (SaMD) with clinical decision support. This is guidance, not formal regulatory advice — engage qualified consultants per market.
| Stage | Positioning & claim | US (FDA) | EU (MDR) | Australia (TGA) |
|---|---|---|---|---|
| 1–2 | General wellness / education / visualisation — no diagnostic claim | Likely outside device scope if claims stay wellness (FDA general-wellness policy); avoid disease claims | Wellness product; MDR applies if a medical purpose is claimed | Wellness; not therapeutic goods if no medical claim |
| 3 | Connected visualisation / tracking; software carefully scoped | Wellness or low-risk; watch software-function guidance | Borderline — assess intended purpose | Assess against "medical device" definition |
| 4–5 | Optical imaging aid / accessory; possible Class I–II | Likely Class II 510(k) once a medical purpose is claimed | Class IIa/IIb depending on function | Class IIa style, TGA inclusion |
| 6–7 | Clinic-grade imaging device (adjunct) | 510(k) or De Novo; PMA if novel high-risk claims | Class IIb, notified-body review | Class IIb–III conformity |
| 8 | AI clinical decision support (SaMD) | SaMD pathway; PCCP for adaptive models | MDR SaMD (often IIb+); AI Act interplay | TGA SaMD framework |
| 9–10 | Diagnostic imaging / screening platform | PMA-level evidence for screening claims | High class + clinical evaluation | High class + clinical evidence |
ISO 13485 QMS from Stage 4; ISO 14971 risk management throughout; design controls & DHF.
IEC 60601-1 / -1-2 (electrical & EMC), IEC 62471 (photobiological/optical safety), IEC 60825 if any laser sources.
IEC 62304 software lifecycle, IEC 62366 usability, and cybersecurity/privacy per FDA & MDR expectations.
Every label reviewed against the claim ladder; explicit "not a diagnostic device / not a mammography substitute" statements early.
Planning ranges for a lean team, in USD. "Prototype" = working proof of concept; "MVP" = sellable first units incl. tooling/compliance; "Launch" = pilot manufacturing + market entry. Validate against firm supplier and consultant quotes before committing.
| Stage · Product | Duration | Prototype | MVP | Launch | Team | Tech risk | Reg. risk | Revenue potential |
|---|---|---|---|---|---|---|---|---|
| 1 · Halo | 6–9 mo | $30–80k | $120–300k | $0.2–0.6M | 3–5 | Low | Low | Volume D2C |
| 2 · Halo Sense | 6–12 mo | $60–150k | $250–500k | $0.4–1.0M | 4–7 | Med | Low | Higher ASP D2C |
| 3 · Halo Link | 9–14 mo | $120–300k | $0.4–0.9M | $0.8–1.8M | 6–10 | Med | Med | HW + app subs |
| 4 · Spectra | 12–18 mo | $0.3–0.7M | $0.8–2.0M | $1.5–4M | 8–14 | Med | Med | Clinic HW + SaaS |
| 5 · Perfusa | 12–18 mo | $0.5–1.2M | $1.5–3.5M | $3–7M | 12–20 | High | Med–Hi | Clinic HW + SaaS |
| 6 · Clara | 18–24 mo | $1–2.5M | $3–7M | $6–15M | 20–35 | High | High | Capital sale + service |
| 7 · Aura 3D | 24–36 mo | $3–8M | $8–20M | $15–40M | 35–60 | Very high | High | High-value capital |
| 8 · Insight (AI) | parallel, 18–30 mo | $0.5–2M | $3–8M | $6–15M | +AI 6–12 | High | SaMD | Recurring SaaS |
| 9 · Continuum (cloud) | parallel, 18–30 mo | $0.4–1.5M | $2–6M | $5–12M | +cloud 5–10 | Med | High | Recurring SaaS |
| 10 · Horizon | ongoing | — | integration | $30M+ (trials) | 60+ | High | Highest | Platform + data |
Early stages are dominated by industrial design, optics & firmware. Mid stages add multi-spectral capture, calibration & SaaS. Late stages are dominated by clinical studies, QMS, regulatory and manufacturing scale-up — where budgets step up by an order of magnitude.
This is why staging matters: consumer revenue and grants can carry Stages 1–4, while dilutive capital is reserved for the capital-intensive clinical climb.
Revenue mix shifts from one-off consumer hardware toward recurring SaaS and clinical service contracts as we climb — improving margins, retention and enterprise value.
D2C e-commerce, marketplaces, distributors. Optional app freemium/subscription for tracking & education. Accessory & replacement revenue.
Hardware sale or lease + annual service & calibration + per-seat SaaS. "Research edition" for institutions; "clinical edition" once cleared.
Per-scan / per-seat AI & cloud SaaS, enterprise licences, and — only with consent, ethics and legal review — de-identified research data partnerships.
Home · Prosumer · Research · Clinical — one optical core, tiered by claims, calibration & support.
Hardware-led early gross margins give way to high-margin recurring software late.
Longitudinal data creates switching cost: a patient's optical history lives in Continuum.
Data monetisation only where consented, de-identified, lawful and non-exploitative.
File early and broad on the foundational optical control, then layer method, design and software patents at each stage. Provisionals buy a 12-month runway to convert with data behind them.
Mix: utility (methods/hardware), design (form factors), trade secrets (calibration & training pipelines), and data rights.
A preliminary landscape scan of public patents and applications adjacent to multi-wavelength optical breast & vascular sensing. This is a starting map for freedom-to-operate work — not a formal FTO opinion. A qualified patent attorney must run a full clearance search before any filing or commercial launch.
| Publication / Patent No. | Title | Priority era | What it claims (paraphrased) | Relevance to LUMÉA |
|---|---|---|---|---|
| US 11,304,456 B2 (US2021/0337885) | Smart Bra with Optical Sensors to Detect Abnormal Breast Tissue | 2021 | Wearable bra with NIR emitters & receivers woven in; transmission-mode light through breast tissue; spectral/intensity change flags abnormal tissue.P1 | Close — Stage 4–9 |
| US 2023/0148868 A1 | Wearable Device (Smart Bra) with Compressive Chambers & Optical Sensors | 2023 | Same family; adds compressive chambers to standardise breast geometry against the optical sensor array.P2 | Close — Stage 4–6 |
| US 2023/0389803 A1 | Multilayer Smart Bra or Bra Insert for Optical Detection of Breast Cancer | 2023 | Bra-insert form factor; CW + frequency-domain optical methods targeting haemoglobin, lipid, collagen & water biomarkers.P3 | Close — Stage 4–5 |
| US 11,457,818 "DOB-Scan" | Handheld Probe & System for Imaging Human Tissue | ~2019–2022 | Handheld multi-λ NIR LED pencil-beam probe (690/750/800/850nm) + linear detector array + cross-sectional image reconstruction; validated on phantoms & patients.P4 | Adjacent — Stage 4–6 |
| US 2003/0195417 A1 US 2002/0045833 A1 | Medical Optical Imaging Scanner Using Multiple-Wavelength Simultaneous Data Acquisition for Breast Imaging | 2000–2003 likely lapsed | Tabletop scanner; multi-λ illumination + photodetector arc; optional fluorescent contrast agent (ICG) embodiment.P5 | Distant — Stage 6–7 |
| US 8,706,200 B2 US 8,380,291 B2 "VeinViewer" family | Scanned Laser Vein Contrast Enhancer | ~2005–2014 | Scanned multi-wavelength laser system with independently modulated laser intensities to project a real-time vein-contrast image onto skin.P6 | Analogous risk — Stage 2 |
The two risks that can kill this company are overclaiming and false reassurance. Everything in our positioning, labelling and validation is built to neutralise them.
| Risk | Sev. | Mitigation |
|---|---|---|
| Overclaiming (implying diagnosis) | Critical | Claim ladder; legal/reg review of every claim; wellness-only early; the thermography precedent as a standing warning.13 |
| False reassurance (user skips screening) | Critical | Mandatory in-product messaging: not a screening test; keep your mammograms; any symptom → see a clinician. Reinforced in UX, not buried in a manual. |
| Skin-tone performance gap | High | Adaptive illumination; diverse test panels across full Fitzpatrick range from Stage 1; report performance by skin tone. |
| Technical (contrast/depth) | Med | Scope claims to superficial vascular contrast early; phantom + volunteer validation before escalation. |
| Clinical (no benefit shown) | High | Staged evidence; do not make clinical claims until studies support them; adjunct positioning. |
| Regulatory (misclassification) | Med | Early consultant engagement; pre-submissions; conservative intended-use wording. |
| Optical safety | Med | IEC 62471 limits; hardware-enforced output caps; thermal safeguards. |
| Privacy / data | High | Consent, de-identification, encryption, regional residency, ISO 27001-aligned controls. |
| Consumer misuse | Med | Clear UX guardrails, education, and refusal to present outputs as findings. |
| Liability | Med | Insurance, disclaimers, documented intended use, adverse-event reporting from launch. |
| Manufacturing / supply | Med | Dual-source LEDs/optics; DFM early; pilot line before scale. |
| Market (slow adoption) | Med | Consumer revenue funds patience; multiple beachheads (home, femtech, vein-adjacent). |
| Funding | Med | Revenue + grants reduce dilution; each stage is a fundable milestone; recoverable failure modes. |
Validation runs from bench to bedside. Early rungs are cheap and fast; clinical correlation and pivotal studies are gated behind revenue and prior evidence — never rushed to support a premature claim.
Cheap/non-dilutive capital carries the risky-but-inexpensive early science; equity is reserved for scaling proven tiers. Each round is unlocked by a specific, demonstrable proof point.
| Phase | Source | Funds | Investors want to see first |
|---|---|---|---|
| Pre-seed | Bootstrap, friends & family, angels | Stage 1 prototype + provisional IP | Working illuminator, credible team, honest positioning |
| Grants | Gov / university / femtech & med-research programs | Optical validation, usability, skin-tone data | Scientific plausibility, IP, safety data |
| Seed | Deep-tech / MedTech VC + accelerator | Stage 1–2 launch, Stage 3–4 build | Real revenue, retention, repeatable optical method |
| Series A | MedTech VC | Clinic-grade Stage 5–6, pilot studies | Clinical traction, calibrated performance, reg. path |
| Series B+ | Growth + strategic (imaging OEMs) | 3D / AI / cloud, pivotal trials | Clinical evidence, SaaS ARR, defensible data moat |
| Non-dilutive / partnership | Clinical-trial & strategic funding | Pivotal studies toward screening claim | Multi-centre data, regulatory momentum |
A small founding core ships Stages 1–3; specialist depth (optics, AI, clinical, regulatory) is added exactly when the stage demands it.
Systems owner; optics-to-clinic translation.
LED drivers, sensing, firmware, power & safety.
Ergonomics, contact geometry, manufacturability.
GTM, fundraising, partnerships.
Multi-spectral capture, calibration, contrast.
Capture app, cloud, dashboards.
Segmentation, longitudinal change, validation.
Breast radiologist & surgeon; regulatory consultant; quality manager.
E-commerce & marketplaces; women's-health communities & educators; responsible influencer/education content; content-led SEO. No fear marketing — ever.
Women's-health & GP clinics, wellness clinics; research edition into universities to seed evidence and KOLs; regional distributors.
Breast clinics & imaging centres via medical-device distributors; hospital tenders; clinical pilots; strategic partnership or channel with an imaging OEM.
The FDA has issued repeated warnings that thermography — safe and radiation-free — must not be sold as a mammography substitute, because doing so risks patients missing early, treatable cancer.13
We treat that as the defining constraint of this company. A radiation-free breast device earns trust by under-claiming and over-proving — the opposite of the products that drew those warnings.
Every consented scan makes the next model better — but only under a governance regime that would survive scrutiny from a regulator, an ethics board and a journalist.
Explicit, revocable, purpose-specific consent; no data use without it.
Strip identifiers; privacy-by-design pipelines; regional residency.
Deliberately balanced across skin tones & body types; performance reported by subgroup.
AI flags; clinicians decide. Labels from qualified reviewers; validated before deployment.
A simple, safe, affordable handheld device that helps people see vascular contrast directly.
Tomorrow the same optical foundation becomes adaptive. Then image-based. Then quantitative. Then clinical. Then AI-supported. Over ten years, one honest light-based product evolves — wavelength by wavelength — into a radiation-free breast-imaging ecosystem.
Launch Stage 1–2, file foundational IP, run optical + usability validation across the full skin-tone range, stand up a lean commercial engine.
Demonstrate repeatable optical method + real revenue + retention — the proof stack that unlocks Series A and the clinical-grade Stage 6 climb.
Every stage sells. Every stage de-risks the next. A moonshot outcome with a downside that pays its own way — the rare MedTech bet that can't quietly die.
Data room, live Stage 1 prototype demo, and a term-sheet conversation for the seed round that ships light into the world.
Market and clinical figures are attributed below. Analyst market estimates vary by methodology; we cite a convergent range. Consumer/femtech and vein-finder sizing, all per-stage costs, timelines, prices and TAM/SAM/SOM splits are LUMÉA planning assumptions for illustration and must be independently validated before any financing decision. Nothing here is medical, legal, regulatory or investment advice.
Patent landscape compiled from public patent-office and Google Patents listings for orientation only; not a legal opinion. Status (granted / pending / lapsed) and claim scope should be independently verified.
Amir Hossein Kayvanpour · امیر حسین کیوانپور
LUMÉA — concept, staged product roadmap, and this document's design and construction — originated by Amir Kayvanpour. Prepared as an illustrative investor-facing proposal.
LUMÉA is an illustrative venture concept prepared for discussion. Company and product names are placeholders. Renders are schematic concepts, not engineering drawings. © concept document — confidential.