Health Care AI Technology In Clinical Practice

This week, Microsoft continued its latest acquisition streak by announcing the friendly take-over of Nuance Communications. The company specializes in voice transcription and related artificial intelligence software. In particular, Nuance is active in the health care niche, aiming to provide software to facilitate clinical documentation.

This made me wonder, how this niche of health care AI solutions is composed and who other players might be.

General Overview

First of all, the health care industry is huge. In the U.S. alone, health care spending amounted to $3.8T. It’s no wonder that Big Tech companies such as Microsoft or Apple heavily invest in their health care devices and services, still having lots of room to grow.

So how does Artificial Intelligence fit into this picture? In a recent article, Forbes divides the use of AI in health care into three distinct areas. They differ between clinical, administrative and R&D use. In this post, I am going to focus on the clinical applications.

Use of AI in the Clinical Practice


StartUs Insights forecasts that software will be especially disruptive in areas which today would require human intelligence. Consequently, the implementation of AI will have a major impact on diagnostics. Essentially, software algorithms will automate diagnostics tasks and at the same time reduce human error, e.g. through advanced image recognition capabilities. In a research white paper, Geoff Hinton declares in 2016 that “people should stop training radiologists now. It’s just completely obvious that within 5 years, deep learning is going to do better than radiologists.” Examples of emerging start-ups would be Caption Health and Paige. Although Hinton’s 5 years are due this year, the adoption of the technologies still lacks behind Hinton’s expectations. However, the benefits are apparent and it will be interesting to see how the implementation evolves.

Radiology Scan showing a left human hand
Photo by Owen Beard on Unsplash

Patient Intake and Engagement

Another important development stems from progress in Natural Language Processing. Chatbots become increasingly capable to automate the intake of patients as well as the initial engagement like patient screening and care navigation.

Remote Health

The combination of both mentioned advancements fuels another application: Telehealth. A previous study by McKinsey, the authors identify a quarter-trillion dollar post-COVID-19 opportunity. In a recent study, McKinsey reports on virtual health care visits during the pandemic. They find that telehealth utilization has increased 10-fold in Germany, 25-fold in the U.S. and 50-fold in France. The soaring demand definitely justifies for the technology to stay around post-pandemic.

Precision Medicine

In short, precision medicine aims to provide individualized treatments, based on a patient’s very own genetic, environmental and behavioral context. The benefits of such a service are obvious. Precision medicine yields a dramatic efficiency increase of treatments with potential implications for the design and manufacturing of drugs.

However, the success of the technology depends on data and there is plenty of that. Per year, the sector captures about one trillion gigabytes of data and that’s doubling every two years. Although the technology has yet to deliver on its promises, there are some promising players who might be able to create such services eventually.


Advancements in clinical applications of AI yield promising and fascinating ways to treat patients in the future. The use of AI in diagnostics and remote health could drastically decrease error rates due to advanced image processing. Recent developments in Natural Language Processing might enable a streamlined patient intake which is less prone to human administration. At the same time, the amount of data for treatments is growing at a fast pace. Relatedly, precision medicine might leverage the historical data and unlock individualized patient treatment.

Closing Remarks

As much as I am interested in the sector, the implementation of the technology in the field is still lacking behind. It will be very interesting to see, whether people like Geoff Hinton will eventually prove to be right and especially, in what time frame they might do so.

As always, thank you very much for reading. Please feel free to leave feedback!

Leave a Reply