Digital health leaders share predictions on what… Leave a comment
Clockwise, from top left: University of Washington professor Shyam Gollakota; WRF Capital managing director Loretta Little; Amazon vice president Taha Kass-Hout; UW professor Su-In Lee; and Hurone AI founder Kingsley Ndoh. (Amazon, Hurone AI, WRF Capital and UW Photos)
Funding for digital health startups and initiatives soared during the pandemic as entrepreneurs and consumers increasingly embraced telehealth, remote monitoring, and a suite of devices from sleep trackers to exercise bands.
Total venture capital investment in digital health hit an all time high of $29.2 billion in 2021, according to Rock Health. Funding cooled in 2022, to $12.6 billion by the end of the third quarter, but advances in technology such as artificial intelligence and the increasing interest of big tech companies are sure to propel innovation in the future.
Seattle-area startups such as CalmWave, Rippl Care, Outbound AI and Birch AI emerged in 2022 to help solve medical problems ranging from excess noise in hospitals, to mental healthcare for seniors. Larger companies also signaled major ambitions; Amazon this year announced its bid to acquire primary care company One Medical for $3.9 billion and rolled out a new online health service, Amazon Clinic.
What trends do experts see for digital health nationally and in the Seattle area for 2023? We asked five to weigh in with their predictions.
Taha Kass-Hout, vice president of technology-health AI, and chief medical officer at Amazon Web Services
Taha Kass-Hout. (Amazon Photo)
Unprecedented innovation and collaboration across the healthcare and life sciences industries is pushing the industry to move from sick care to prevention through a patient experience that is precise, personalized, and human. The industry has been experimenting with cloud for a decade and understands how technology and machine learning can enable more targeted diagnostics and treatments, known as precision medicine; personalize patient journeys; and improve health outcomes.
In 2023 and beyond, we expect healthcare and life sciences organizations to continue to make investments in modernizing their infrastructure, derive actionable insights from data, and internalize what it means to personalize health. This will involve integrating genomics and other omics data into therapeutic development, leveraging machine learning and analytics to improve clinician workflows, incorporating social determinants data into disease management at the patient or population levels, and using structured and unstructured data to predict disease with much better accuracy — helping move the industry from reactive to preventive patient care.
Kingsley Ndoh, founder and chief strategist, Hurone AI
Kingsley I. Ndoh. (Hurone AI Photo)
We should expect to see more people-centered innovations in digital health to support clinical decision making, such as diagnostic predictive technologies or tools to predict clinical outcomes for certain cancer drugs. These tools will increasingly incorporate more diversity in training datasets for machine learning models and put the specific needs of the target user at the heart of the development process, including taking into account cultural perspectives.
There will also be better integration of data generated from wearables, smartphone apps and electronic medical records to support clinical decisions, behavioral change, and personalization at scale through the power of artificial intelligence.
Loretta Little, managing director of WRF Capital
Loretta Little. (WRF Capital Photo / Mel Curtis)
Funding for most early-stage digital startups will continue to be tight in 2023, but I see opportunities for growth in several areas. We will continue to see more companies that offer access to mental health services through innovative products and approaches, such as Joon and Rippl Care, and companies focused on improving connectivity and tools for better remote care such as Valorant Health and Wavely Diagnostics.
Remote care is especially important for underserved rural communities that have limited or no access to nearby health resources. This need is only increasing, driven in part by demographic shifts in the patient population. The proportion of seniors in Washington state and across the nation is projected to grow, particularly in rural areas. This rural senior population represents a large percentage of chronic disease sufferers and will need to be linked up with services.
Shyam Gollakota, co-founder of Wavely Diagnostics and Sound Life Sciences (acquired by Google), professor at the University of Washington’s Allen School
Shyam Gollakota. (GeekWire Photo / James Thorne)
The adoption of telehealth that accelerated during COVID is likely here to stay. We may also see an increased number of remote in-home tests like COVID-19 or blood tests that will bring telehealth closer to an in-person visit. While there has been a lot of focus on using smartphones and smartwatches for mobile health, earbuds will be the next exciting platform for monitoring health and wellness as well as potentially, in the next few years, electroencephalography (EEG) signals that can open up new opportunities for brain interfaces.
We will hopefully also see a number of startups apply large language models to address various pain points in the healthcare system with the goal of improving efficiency and reducing cost. Deep learning techniques will continue improving and we will start seeing more promising results for addressing important problems like using AI to discover drugs and vaccines.
Su-In Lee, UW professor of computer science and engineering
Su-In Lee. (UW Photo)
Next year we will see AI devices with explainable AI (XAI) functionality, enabling humans to understand the reasoning process of complex, black-box machine learning models. I also see FDA approval processes incorporating XAI analysis to engender trust, transparency, fairness, and actionability of machine learning models.
Increased reimbursement by insurance providers and the U.S. Centers for Medicare & Medicaid Services will drive an increase in the number of FDA-approved AI devices. In the long term, the success and fairness of medical AI devices will rely on the extent to which FDA approval processes are updated to reflect machine learning-specific issues. For example, if there aren’t requirements to evaluate an AI dermatology device on a wide range of skin tones, it seems likely that AI devices that perform poorly on darker skin will become publicly available and disproportionately misdiagnosis people with darker skin.
