Why investors are betting big on AI-powered health care

Note: This interview was recorded on Sunday, 19 October 2025. 

You can listen to the episode at either of the links below: 

• Spotify
• Apple Podcasts

Artificial intelligence is transforming healthcare diagnostics - from how images are read to how hospitals make decisions. In the latest episode of Market Mixers, hosts Ally Selby and I explore how AI is reshaping the sector and what investors should watch as innovation comes for diagnostics.

The first wave of AI in diagnostic healthcare focused on imaging. Companies like Heartflow (NASDAQ: HTFL) pioneered cardiac CT analysis, turning standard scans into 3D models that quantify plaque and blood flow. This image-based AI offered structured pixels, clear economics, and measurable value for clinicians, so they could better decide who really needs invasive procedures.

The second wave targeted workflow. Instead of simply spotting disease, new models optimised efficiency, cutting analyst time from over an hour to less than half that. Competitors such as Artrya (ASX: AYA) aim to go even further, pushing toward real-time, point-of-care automation.

Now the third wave is arriving: language and multimodal AI. These systems can interpret medical images, read radiology notes, and even generate structured reports, essentially “ChatGPT for clinicians.” Human oversight is still critical, but the trajectory points toward increasingly autonomous reporting tools.

When analysing diagnostics companies, investors should break down their research into four checkpoints: the care pathway, evidence, regulation, and reimbursement. Understanding where a product fits in the patient journey, whether it replaces, triages, or augments existing tests, is crucial. The quality of supporting data matters too: randomised controlled trials remain the gold standard, while small or single-site studies are higher risk.

Regulation defines how fast a product can reach the market. The US FDA pathways range from 510(k) for existing technology, De Novo for low- to moderate-risk innovation, to PMA (pre-market approval) for high-risk devices. Yet even the best technology can stumble without reimbursement. Category I CPT codes remain the “holy grail” for consistent insurer coverage.

In business terms, diagnostic firms typically earn through per-scan fees, SaaS subscriptions, or hardware-plus-consumables models. The less human labour involved, the higher the potential margins. Stakeholders extend beyond clinicians. IT, finance, and payers all influence adoption.

Among public and private players, RadNet (NYSE: RDNT) stands out as a “picks and shovels” success story. With nearly 400 imaging centres in the US, its DeepHealth division uses AI to improve breast, lung, and prostate screening. Studies show a 21% improvement in breast cancer detection using its DeepHealth AI.

Meanwhile, Artrya is gaining traction with a modular, software-first platform that converts cardiac CT scans into 3D artery maps. Its cleared “Anatomy” and “Plaque” modules will soon be joined by “Flow,” each leveraging prior approvals for faster expansion. Artrya’s approach, cutting turnaround times from hours to minutes, offers the scalability and margins of a SaaS business, not a traditional device maker.

Artrya’s share price has risen 8.5x in the past year, yet analysts still see headroom in its $520 million market cap, given the $5 billion addressable market. Compared to HeartFlow’s $2.8 billion valuation, we believe Artrya remains undervalued given its superior automation, patient outcomes, and hospital economics.

AI diagnostics are entering a “fintech moment”, reshaping how information flows, improving accuracy, and accelerating care. The winning mix? RadNet for stability and Artrya for growth. 

Timecodes: 

• 0:00 - Intro 
• 1:03 - How to make a "Cardio Cooler" 
• 1:19 - How investors should think about analysing healthcare diagnostics stocks 
• 1:54 - Importance of evidence and trials 
• 3:06 - Why regulation and FDA/TGA pathways matter 
• 5:01 - Do you need specialist knowledge to invest in healthcare companies 
• 6:05 - How to use AI/ChatGPT to help you with your analysis 
• 7:10 - How these businesses make money 
• 7:43 - Who are the stakeholders who are involved 
• 8:48 - A deep dive into Heartflow (NASDAQ: HTFL) 
• 9:45 - Why we like RadNet (NYSE: RDNT) 
• 11:30 - Why Minotaur is bullish on Artrya (ASX: AYA) 
• 14:24 - What we got wrong on Heartflow 
• 16:37 - Risks to be aware of when investing in healthcare diagnostics 
• 18:07 - Key takeaways