Miso Radar · Version 1

Clearance is not adoption.

An open map of healthcare and life-science AI across six APAC markets — and of the regulatory, reimbursement and workflow conditions that decide whether an approved product is ever actually used.

Why this exists

Most APAC health-AI analysis stops at the funding round.

Capital and regulatory clearance are the two things that get reported. Neither tells you whether a hospital changed what it does on a Tuesday morning.

This radar tracks the third thing: the conditions between approval and routine use. Who pays, under which code, inside whose workflow, against what local evidence requirement. That gap is where most cross-border health-AI programmes quietly fail.

The market lens

Six markets. Six different bottlenecks.

There is no APAC market. There are six regulatory regimes, six payment logics and six distribution realities that happen to share a time zone band.

China

Regulatory gate
NMPA treats AI diagnostic software as Class III. In-country clinical data is expected for higher-risk devices, so a foreign clearance does not travel.
Who pays
Provincial pricing and hospital-level procurement decide adoption. National reimbursement for AI software is not a single pathway.
What still blocks it
Tier-3 hospitals concentrate both the data and the buying power. Distribution usually means partnering with an incumbent imaging or IT vendor rather than selling direct.

South Korea

Regulatory gate
MFDS applies detailed algorithm validation and post-market performance monitoring requirements to AI devices.
Who pays
The state is actively pulling products through. In August 2026 the Ministry of Health and Welfare selected six consortiums to accelerate AI medical-device commercialisation in brain disease, cancer and cardiovascular care.
What still blocks it
The domestic market is small relative to R&D cost, so Korean vendors export early. That makes them unusually willing partners for cross-border programmes.

Japan

Regulatory gate
PMDA generally requests local clinical data at application, and will require all-case post-marketing surveillance where local data is limited. Approvals favour locked algorithms over continuously-learning ones.
Who pays
Innovative devices can attract premium reimbursement under the National Health Insurance system, but the premium must be argued.
What still blocks it
Local clinical evidence and a domestic partner are effectively prerequisites. This is the slowest gate in the region and the most often underestimated.

Singapore

Regulatory gate
Smaller domestic market, used primarily as a regulatory and commercial base for the region rather than as the end market.
Who pays
Substantial state co-investment. Startup SG Equity exceeds S$1bn in co-investment capital, and Flagship Pioneering committed up to S$100m jointly with A*STAR for a regional biotech and healthtech hub.
What still blocks it
Genuine adoption volume has to come from elsewhere in APAC. Singapore success is not evidence of regional adoption.

India

Regulatory gate
Lighter regulatory gate than Japan or China, which lets products reach real clinical volume faster.
Who pays
Public-health and screening programmes, plus NGO and donor funding, are meaningful buyers alongside private hospitals.
What still blocks it
Price points are far below OECD levels, so unit economics must be designed for volume from the outset. Products built for US pricing do not port.

Australia

Regulatory gate
TGA clearance plus strong alignment with UK and EU pathways, which makes Australia a useful launch market for products targeting Western reimbursement.
Who pays
Reimbursement is the differentiator. Harrison.ai's Annalise product line reached US Medicare payment of roughly US$240 per scan, which is the kind of proof point that unlocks hospital budgets.
What still blocks it
Domestic volume is small. The market's value is as an evidence-generation and credibility base, not as scale.

The companies

Who is actually building.

Showing 19 of 19 entries

Healthcare and life-science AI companies operating in APAC markets, with the public signal recorded for each and links to the supporting sources.
CompanyMarketSegmentWhat it doesPublic signalSources
Airdoc 鹰瞳科技ChinaDiagnosticsRetinal image analysis for chronic-disease and cardiovascular risk screening.Publicly listed, with Class III NMPA device licensing underpinning commercialisation.12
DeepWise 深睿医疗ChinaMedical imagingMulti-modality diagnostic imaging AI.Named among the Chinese imaging-AI vendors running large-scale hospital rollouts with real-world validation.12
HelixonChinaDrug discoveryAI protein and antibody design.Grouped with XtalPi and CSPC among Chinese AI-biotech firms striking large deals with global pharma.1
Infervision 推想医疗ChinaMedical imagingImaging AI across chest CT, stroke and oncology workflows.Among the most established Chinese imaging-AI vendors; partners with domestic hospitals and global pharma for real-world validation.12
Insilico MedicineChinaDrug discoveryGenerative AI for target discovery and molecule design, with a robotics lab in Suzhou.Hong Kong headquartered. IPO on HKEX in December 2025 raising approximately US$293m.12
Shukun Technology 数坤科技ChinaMedical imagingCardiovascular, neuro and oncology imaging analysis.Reached IPO on the back of Class III licensing and hospital rollout.12
Wision AI 希诺医疗ChinaClinical decision supportReal-time endoscopy AI for polyp and lesion detection.Listed among China's AI medical-diagnostics companies; procedural AI is a distinct adoption pattern from radiology.1
XtalPi 晶泰科技ChinaDrug discoveryAI plus robotic-laboratory platform for small-molecule discovery and materials science.HKEX-listed, roughly US$785m raised. Signed a 2025 pipeline collaboration with DoveTree reported at up to US$5.99bn in total potential value.Deal value is potential, milestone-contingent consideration — not revenue.12
Coreline SoftSouth KoreaMedical imagingLung, cardiac and chest CT quantification, with a screening-programme focus.Nine radiology FDA clearances, placing it in the FDA AI top 20. Acquired VUNO's LungCT line in early 2026.123
JLKSouth KoreaMedical imagingStroke and brain-imaging AI.Secured four US FDA clearances within a single year. Also selected in the August 2026 MOHW consortium programme.12
LunitSouth KoreaMedical imagingOncology imaging AI for mammography and chest radiography, plus tissue biomarker analysis.Reported reaching roughly US$60m revenue. Selected in the August 2026 Ministry of Health and Welfare consortium programme.123
VUNOSouth KoreaClinical decision supportDeepCARS deterioration prediction and cardiac monitoring.In early 2026 transferred its VUNO Med-LungCT line to Coreline Soft for ₩3.4bn to concentrate on DeepCARS and cardiac monitoring.A rare public example of APAC medical-AI portfolio consolidation rather than growth-at-all-costs.12
EcoNaviStaJapanClinical decision supportSleep and vital-sign monitoring for care settings.Acquired in 2025 in a deal reported at US$113.7m — one of the region's largest digital-health exits that year.1
UbieJapanPatient navigationPhysician-supervised symptom assessment guiding patients to conditions and nearby care.Over US$125m raised. Backers include Google, Japan Post Capital, NTT DOCOMO and Seven-Eleven Japan. Partnership with Mayo Clinic; expanding into the US.The corporate-strategic investor base is characteristic of Japan and shapes distribution more than the capital does.123
HolmuskSingaporeReal-world evidenceReal-world evidence platform for behavioural health and chronic disease.US$21.5m Series A led by Optum Ventures and Health Catalyst Capital, with Heritas Capital participating; expanding US operations.Evidence infrastructure rather than a clinical device — a different and often overlooked adoption path.1
MiRXESSingaporeDiagnosticsRNA-based early cancer detection, notably gastric cancer screening.IPO in 2025 raising approximately US$139m — the largest APAC digital-health exit of that year.1
Qure.aiIndiaMedical imagingX-ray, CT and ultrasound interpretation, with strong deployment in TB and lung screening.Founded 2016 in Mumbai; roughly US$125m raised including a US$65m Series D. Named in TIME's Most Influential Companies 2025.Describes itself as the world's most deployed healthcare AI. Deployment breadth here is genuinely public-health scale, not pilot scale.12
Annalise.aiAustraliaMedical imagingComprehensive chest X-ray and CT brain triage covering a wide finding set.Cleared in over 40 countries and used in more than 1,000 healthcare facilities including the NHS, supporting care for over 6 million patients annually. US Medicare approved roughly US$240 per scan.The single clearest APAC example of the full chain completing: clearance, then reimbursement, then volume.12
Harrison.aiAustraliaMedical imagingParent company of Annalise.ai (radiology) and Franklin.ai (pathology).Roughly US$252.6m raised in total including a US$112m Series C. Australia's National Reconstruction Fund took a A$32m stake.123

The capital layer

Selective, and consolidating.

Funding is necessary and nowhere near sufficient. It is included here because it determines which companies survive long enough to hit the adoption gate.

US$1.2bn

APAC digital-health funding, Q2 2025

AI-powered ventures captured roughly 63% of the capital deployed.

Galen Growth
US$1.6bn

First nine months of 2025, cumulative

Q3 2025 alone recorded US$416.3m in new venture investment — slower growth, more selective deals.

Galen Growth
14 exits / US$434.4m

APAC digital-health exits, 2025

Led by MiRXES (Singapore IPO, US$139m), UniHealth (China IPO, US$178m) and EcoNaviSta (Japan M&A, US$113.7m).

Galen Growth
50 ventures / 8 countries

Galen Growth HealthTech 250, APAC 2026 cohort

South Korea 14, Japan 10, Singapore 9, India 6, Australia 5 — a useful proxy for where regional density actually sits.

Galen Growth

The regulatory backdrop

Three structural facts that shape every entry above.

  1. 01AI devices went from rare to routine

    AI medical devices grew from roughly 1 in 700 FDA clearances in 2019 to 1 in 28 in 2025, with over 1,500 authorised in the US alone and national programmes scaling in Korea, China, Taiwan and Japan. Pure Global

  2. 02Fragmentation is the dominant cost

    Each APAC authority maintains its own processes, templates, timelines and data expectations. There is no regional mutual recognition, so market entry cost scales close to linearly with the number of markets. RegDeskPure Global

  3. 03Regulation assumes frozen algorithms

    Traditional device regulation assumes an algorithm is fixed at launch. Self-learning systems break that assumption, and most APAC regimes have not fully addressed continuously-evolving products. Frontiers in MedicinePure Global

The platform layer

Where the ecosystem is already organised.

Chip and cloud platforms have built the most complete startup-facing programmes in the region. They are a useful map of where density and capital already cluster — and of which markets still have no equivalent structure.

NVIDIA InceptionMore than 5,000 healthcare and life-science startups across the programme, digital health the largest segment at over 2,000 members. US, China, India and the UK together account for just over half of all Inception startups. A Korean alliance adds accelerated computing via SK Telecom with VC support from IMM Investment, Korea Investment Partners and SBVA. NVIDIA · NVIDIA
Microsoft for StartupsAzure infrastructure, technical guidance and go-to-market support, routing healthcare organisations toward vetted startups. The invite-only Pegasus programme pushes venture-backed B2B startups into Microsoft's enterprise customer network — distribution, not just compute. Microsoft · Business Wire
Amazon Web ServicesRuns an AI Startup Pavilion at its Life Sciences Symposium and operates regional accelerators on a cohort model, including a ten-week APJ programme taking up to 40 startups with dedicated business and technical support. Amazon Web Services · Amazon Web Services
Alibaba Cloud AI CatalystThe most significant APAC-native equivalent. Launched at Apsara 2025 with expert guidance, access to the Model Studio generative-AI platform and up to US$120,000 in cloud credits, against a stated US$50bn AI commitment. Alibaba Cloud · Digital Commerce 360
State capital, ChinaIn December 2025 the State Council launched a National Venture Capital Guidance Fund with CNY 100bn of central capital intended to mobilise CNY 1tn, requiring 70%+ to go to seed and early-stage companies. Platform programmes are not the only organising force in this market. Peony

Method

How this was built.

Inclusion. A company appears if it applies AI to a healthcare or life-science problem, operates from an APAC market, and has at least one verifiable public signal — a funding round, a listing, a regulatory clearance, a named deployment or a government programme selection.

Sourcing. Every figure is attributed and dated. Where a source reports a potential or milestone-contingent value, that is stated rather than presented as revenue.

Known limits. This is a v1 sample with deliberately uneven coverage. China and Korea are better represented than Southeast Asia. Private companies without public disclosure are systematically under-counted. Nothing here should be read as a market share estimate, an investment recommendation, or a clinical endorsement.

Corrections. If an entry is wrong or out of date, tell us.

Sources

Everything this radar is built on.

  1. At 8K+ Members, NVIDIA Inception Displays Global AI Startup EcosystemNVIDIA
  2. NVIDIA and Microsoft Boost AI Innovation for Healthcare StartupsNVIDIA
  3. Korea Joins AI Industrial Revolution: NVIDIA CEO Unveils Partnership at APEC SummitNVIDIA
  4. How AI and Microsoft for Startups are transforming healthcare and life sciencesMicrosoft
  5. Autonomize AI joins the Microsoft for Startups Pegasus ProgramBusiness Wire
  6. AWS Life Sciences Symposium — AI Startup PavilionAmazon Web Services
  7. AWS Startups — Future of Healthcare & Life SciencesAmazon Web Services
  8. Alibaba Cloud AI Catalyst ProgramAlibaba Cloud
  9. Alibaba Cloud commits $50 billion to AIDigital Commerce 360
  10. Startup accelerators in China — National Venture Capital Guidance FundPeony
  11. APAC Digital Health Funding in Q2 2025: AI-Driven MomentumGalen Growth
  12. APAC Digital Health Funding Q3 2025: Slower Growth, Stronger SignalsGalen Growth
  13. HealthTech 250 APAC CohortGalen Growth
  14. AI as a Medical Device: Global Regulation & Market AccessPure Global
  15. Medical Device Registration in APAC: The Complete Reference GuideRegDesk
  16. A decade of review in global regulation and research of AI medical devices (2015–2025)Frontiers in Medicine
  17. How Does Japan Handle Reimbursement For AI-Driven Medical Devices?Pacific Bridge Medical
  18. Regulatory Transparency in AI-based Radiology Software: PMDA-approved SaMD ProductsmedRxiv preprint
  19. AI as a Medical Device: Global Regulation & Market AccessPure Global
  20. Korea Accelerates Medical AI Commercialization, Backing Lunit, JLKSeoul Economic Daily
  21. AI Imaging Expansion in National Health Screenings Lifts Lunit, Coreline HopesSeoul Economic Daily
  22. Korean medical imaging AI continues global saturation in 2025MobiHealthNews
  23. Vuno invests $2M in Korean rival and more briefsMobiHealthNews
  24. Clearance isn't enough: Korean medical AI firms rethink overseas playbooksKorea Biomedical Review
  25. Korea Digital Healthcare: Medical AI Enters Its Revenue EraSeoulz
  26. China's AI drug discovery startups cash in on billion-dollar dealsRest of World
  27. XtalPiWikipedia
  28. AI Drug Discovery Companies: The 2026 Power ListVision Life Sciences
  29. Top startups in AI in Medical Diagnostics in ChinaTracxn
  30. China's AI Healthcare Market (Part II): Opportunities and Strategies for Foreign InvestorsMade-in-China Insights
  31. Ubie Secures Over $125 Million in Total FundingThe SaaS News
  32. Health Tech Startup Ubie Raises Funding from Japan Post CapitalBRIDGE
  33. Japanese medical AI startup Ubie snaps up $26M in Series C fundingMobiHealthNews
  34. Health Tech Company Holmusk Secures $21.5M Series ACrunchbase News
  35. Qure.ai — Company Profile & FundingCrunchbase
  36. Qure.ai among TIME's Most Influential Companies in the World for 2025Radiology Business
  37. Australian health tech startup Harrison.ai scores $112M Series CYahoo Finance
  38. National Reconstruction Fund takes $32 million stake in Harrison.aiStartup Daily
  39. Harrison.ai: The Sydney Medical AI Company Behind AnnaliseSmallBizAI
  40. Sydney-based medtech startup Harrison.ai gets $129M AUD led by Horizons VenturesTechCrunch
  41. Top VCs in Singapore and Essential Resources for FoundersVisible.vc

Next revision

Where should version 2 go deeper?

Planned additions: Southeast Asian coverage, procedural and surgical AI, and a deployment-stage field separating pilots from routine clinical use.

Suggest a market or segment