When Artificial Intelligence hits the road
August 19, 2021
Every now and then, we remind ourselves why we enjoy our work. To us, global small cap is a wonderful asset class that offers the widest array of companies in terms of style, valuations, and growth profiles. This helps us develop a philosophy and process that fits our investment background, without having to take unnecessary market-related risks and having too much of a narrowed spectrum of investment candidates. Moreover, this investment universe is comprised of many companies that are strongly impacted by secular themes, especially the technology-driven ones.
Technologically driven secular themes usually appear in our universe following a start-up phase where venture capital, serial entrepreneurs, and bankers play god with other people’s money, through conceptual and research-driven business plans. Following that phase, we start seeing the themes when more established companies decide the technology is mature enough to turn into real products.
Artificial Intelligence (AI) is a prime example. The market is already big and getting bigger. If we were to list all of the companies in our portfolio impacted by utilizing AI, the list would be long. In healthcare alone, the AI market is expected to reach USD 58.6 billion by 2028.
Let’s dig deeper and look at a field that will greatly benefit – radiology. Radiology features a high degree of specialty mixed with high throughput that fits AI’s problem-solving capabilities. As a key part of medical practice and research, radiology interprets most human biological ailments. Its complexity has skyrocketed due to biomarkers that are now available to radiologists, especially in the field of oncology.
Radiology and machine learning will prosper in many markets, such as prostate cancer diagnostics, thanks to greater precision in comparison to the conventional PSA test. New AI-assisted markers are set to help radiologists read CT scans for colon cancer. In addition, AI is expected to increase radiology productivity (volume of readouts performed per hour) and the amount of new products (volumes from new diagnostic initiatives).
Speaking of new products, radiology and AI will be a key part of the new Alzheimer treatment paradigm, representing an estimated USD 10 billion market in 2026. The recently approved Alzheimer’s drug Aduhelm, from biotech company Biogen, requires a specialized CT confirmation scan, as well as monthly scans during treatment. As a market, radiology has an above-average growth rate of 5.2%, with a substantial market size of USD 20 billion.
Raffles Medical (RFMD: SP)
Raffles Medical (RMG) is a private healthcare provider operating in 14 cities across Asia, including Singapore, China, Japan, Vietnam, and Cambodia. In China, RMG is present in eight cities.
Hospitals are core users of radiology in the majority of departments. Radiology productivity is linked to efficiency with most procedures, especially in the private sector where RMG operates. With EBITDA margins already at 21% versus mid-teens for its hospital peers, RMG remains a first-mover when it comes to technology driven productivity. Radiology AI will have an important impact on the sustained growth in profitability going forward. RMG operates two hospitals, the 380-bed Raffles Hospital Singapore, which opened in 2003, and the 700-bed Raffles Hospital Chongqing, which opened in 2020. RMG also just opened its third hospital, the 400-bed Raffles Hospital Shanghai. According to management, the new hospitals in Shanghai and Chongqing are starting with 100-150 operational beds.
In Singapore, RMG operates the Raffles Specialist Centre, adjoined to Raffles Hospital, and it has a total bed capacity of 380. Centers of excellence in the Chongqing hospital include gastrointestinal surgery, obstetrics & gynecology, pediatrics, cardiovascular surgery, neuroscience, and oncology, which all feature significant radiology practices.
Our next holding corroborates the kind of AI productivity gains that Raffles could be able to obtain in the near future.
Radnet (RDNT: US)
RadNet is the largest provider of outpatient imaging services in the United States, with 346 centers nationwide.
The company has an AI subsidiary called DeepHealth, which focuses on developing machine-learning applications for the radiology industry. RadNet recently announced the FDA’s approval of its AI mammography triage software. This software acts as a screening tool, enabling radiologists to more effectively manage their mammography cases using AI. DeepHealth’s powerful new AI technology automatically identifies suspicious screening exam results that may need priority attention, allowing radiologists to optimize their workflow for efficiency and effectiveness.
According to the company, RadNet’s first AI approval should translate to a 25% gain in productivity covering its two million annual mammography scans. This should therefore enable the company to expand its capacity and grow without having to hire additional staff.
 Market Insight Reports, July 28, 2021
 Grandview Research, December 2020