Healthcare DX and AI adoption are moving from behind the scenes into actual procurement, benefiting companies with differentiated data and product capabilities over the long term

Overview

This report focuses on the latest progress in DX and AI adoption in the medical technology and healthcare services industry. J.P. Morgan believes that healthcare institutions had previously been constrained in AI applications by issues such as accuracy, safety, ethics, and operational fit, but that recent advances in generative AI, the increase in approved AI medical devices, government subsidies, and hospital efficiency pressures are jointly accelerating practical adoption. The core conclusion of the report is that future market share gains will not depend merely on whether AI functionality exists, but on whether AI and digital solutions are accepted by hospitals, reduce operational burdens, and create difficult-to-replicate differentiation.

Core views

First, Nihon Kohden's DHS and patient monitoring-related solutions still have small revenue scale in the short term, but products such as alarm fatigue mitigation, ward monitoring, and Patient Flow Management could increase its installed base share in North American patient monitoring systems if they establish a deployment track record in Japan and gradually gain recognition from U.S. hospitals. Second, Fukuda Denshi's AI ECG products show that reducing hospital capex for AI usage can itself become a differentiated advantage. Third, JMDC and M3 will benefit over the long term from healthcare data and platform data, but in FY2026 they will face increased spending on AI engineers, data consultants, and infrastructure in the short term. Fourth, companies such as Shimadzu, Sysmex, Terumo, and Asahi Intecc are also advancing AI applications in areas including analytical instruments, IVD, vascular imaging, factory efficiency, and regulatory documentation, though the pace of contribution and certainty of commercialization differ.

Analysis framework

The report uses a method combining industry case synthesis and company mapping: it first summarizes hospital and clinic AI application cases presented at the Eucalia Friendship Meeting, and then assesses the potential impact of AI functions on operational efficiency, capex, competitive advantage, and market share across areas such as medical devices, diagnostic imaging, IVD testing, life sciences, healthcare big data, and pharma marketing support.

Methodology notes

• Industry ResearchDX and AI Adoption Impact Framework

  Assessing the impact of AI adoption across four dimensions: hospital demand, product differentiation, cost burden, and market share transmission

  Rather than simply evaluating whether AI functionality exists, the report examines whether it addresses real hospital pain points, reduces labor hours or capex, relies on proprietary data to form barriers, and ultimately translates into installed equipment share or service profitability.

• Competitive AnalysisBarriers from Proprietary Data and Deployment Cases

  Competitive advantage in healthcare AI comes from data credibility, hospital deployment records, and fit with workflows

  The advantages of companies such as JMDC, M3, and Shimadzu are attributed to highly credible healthcare data, proprietary platform data, or long-accumulated analytical data; Nihon Kohden's potential advantage, by contrast, depends on the recognition and deployment record of DHS in hospital settings.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Nihon Kohden (6849)

  One of the core beneficiary names, benefiting from digital upgrades in DHS, ITS, and patient monitoring businesses.

  Strengths

  Products such as alarm solutions, ward monitoring, and Patient Flow Management can improve hospital operational efficiency; new DHS products in the U.S. in FY2024 have already contributed to monitor sales growth.

  Weaknesses

  Sales scale of DHS standalone products remains small, with limited current contribution.

  Comparison

  Compared with Philips and GE HealthCare's high share in the U.S. patient monitoring market, Nihon Kohden needs greater DHS recognition to expand its installed base.

  Risks

  Slower-than-expected recovery in Japan and overseas market conditions, weaker hospital capex, inability of DHS to form large projects, or limited gains in U.S. share.

• Fukuda Denshi

  A differentiated AI ECG case, not under coverage.

  Strengths

  CardiMax 9 Ai can estimate past atrial fibrillation risk and can run on standard CPU computers, reducing hospital capex for AI use.

  Weaknesses

  The report does not provide data on its commercialization scale or profit contribution.

  Comparison

  Compared with other overseas AI diagnostic products, its advantage lies in not requiring high-end AI computers.

  Risks

  Commoditization of AI diagnostic products, adoption pace in clinical settings, and hospital budget constraints.

• JMDC (4483)

  A beneficiary of healthcare big data and AI applications, but under significant short-term investment pressure.

  Strengths

  Its healthcare big data is highly credible; AI lowers the barrier to analysis and accelerates hypothesis validation cycles, while combining data across industries creates new opportunities.

  Weaknesses

  FY2026 operating profit growth is significantly lower than revenue growth, reflecting strategic investment pressure.

  Comparison

  Synthetic data poses a relatively low substitution risk to highly credible healthcare data.

  Risks

  Hiring of AI engineers and data consultants delayed to FY2027, and infrastructure investment payback slower than expected.

• M3 (2413)

  Platform data and pharma marketing support businesses benefit from AI-driven efficiency gains.

  Strengths

  It owns proprietary healthcare platform data such as M3 DigiKar, and its pharma marketing support business is difficult for other companies to replace; AI has already delivered about 20% productivity improvement versus FY2019.

  Weaknesses

  AI products themselves may become commoditized over the long term, with advantages deriving more from data and platform.

  Comparison

  It faces competition from Medley, WEMEX, and others in the Japanese cloud EMR market.

  Risks

  Intensifying competition in cloud EMR, one-off increases in SG&A, and changes in market structure driven by the government's push for 100% EMR penetration.

• Shimadzu Corporation (7701)

  A potential beneficiary of AI software in the life sciences and analytical instruments field.

  Strengths

  Methodintelligence for LC relies on paper data and Shimadzu's proprietary analytical data, creating high barriers to entry, and is planned to be offered on a monthly subscription model.

  Weaknesses

  The new product has not yet been launched, and its revenue contribution remains unproven.

  Comparison

  Compared with hardware instruments alone, AI-based method creation, automated waveform processing, and anomalous peak detection can enhance software competitiveness.

  Risks

  Customer use of AI in R&D processes may reduce the frequency of measurement instrument usage over the long term.

• Sysmex (6869)

  AI applications in the IVD testing field are advancing gradually.

  Strengths

  Laboratory AI can be used for predicting consumable replacement, equipment failure, calibration errors, and downtime, and can also support quality control and patient risk assessment.

  Weaknesses

  The number of testing devices currently equipped with AI capabilities remains limited, and profit contribution may take considerable time.

  Comparison

  Companies such as Roche are also using AI to improve lab capacity utilization and reduce maintenance costs.

  Risks

  Slow pace of AI adoption and long commercialization conversion cycles.

Key data

• Japan hospital ICT subsidyUp to ¥80 million per hospitalApplicable to communications equipment, AI medical consultation, generative AI document creation, transport robots, monitoring equipment, and network infrastructure.
• AI electronic medical record generation speedLess than 60 secondsCaptures doctor-patient conversations through speech recognition and generates SOAP-format text.
• Claude documentation efficiency improvement1.8xUsed for drafting routine emails and explanatory materials.
• RPA administrative automation labor savings550 hours per month, equivalent to 3.6 employeesAlso identified tens of millions of yen in previously overlooked medical reimbursement claims.
• JMDC FY2026 guidance20% sales growth, 9% operating profit growthProfit growth is slower than revenue growth, mainly due to accelerated strategic investment in AI-related talent and infrastructure.
• M3 productivity improvement20% improvement versus FY2019AI is replacing parts of operational processes in its core pharma marketing support business and improving physician targeting efficiency.
• Online medical consultation penetrationMost prevalent in China and IndiaIn Japan, it is mainly used in clinics, while hospitals more often adopt physician-to-physician remote collaboration services.

Impact & implications

The investment implication is that beneficiaries of the healthcare AI theme need to be distinguished between short-term cost pressure and long-term share gains. Companies with the ability to deploy in hospital settings, proprietary data, low-capex solutions, or deep integration with existing workflows are more likely to gain sustainable advantages; by contrast, products that merely offer similar AI diagnostic functions may face commoditized competition. For investors, the key is not the AI narrative itself, but whether AI capabilities can translate into hospital procurement, increased installed base share, improved service margins, or stronger customer stickiness.

Risks

• Healthcare institution AI deployment requires higher-spec computers and system infrastructure, which may raise fixed costs and capex.
• As the number of AI medical devices and diagnostic products increases, functional commoditization may weaken individual vendors' pricing power and ability to gain share.
• Short-term spending on AI-related talent, system usage fees, and infrastructure by service companies such as JMDC and M3 may pressure margins.
• Hospital budgets, the pace of subsidy applications, or slower-than-expected recovery in capex by Japanese healthcare institutions could affect procurement of DX and AI products.
• Healthcare AI still involves issues related to accuracy, safety, ethics, data confidentiality, and fit with actual workflows.

What to watch

• The actual progress of Japanese hospitals using government subsidies from July 2026 to procure ICT, AI, and monitoring equipment.
• Whether deployment cases of Nihon Kohden DHS products in Japan can translate into increased installed base share for U.S. patient monitors.
• Whether JMDC's FY2026 investments in AI engineers, data consultants, and infrastructure are completed as planned, or delayed into FY2027.
• Changes in M3's SG&A expenses as it faces competition from Medley and WEMEX in the cloud EMR market.
• Whether low-capex AI solutions become an important screening standard for hospitals when procuring healthcare AI products.
• The launch progress of Shimadzu Methodintelligence for LC, market acceptance of its subscription model, and its contribution to software revenue.