The Final Part of Our Series on Health Tech Buzzwords

In the two previous posts, we discussed health tech specific terms like precision medicine and digital health platforms along with purely medical terms that made the switch from medicine to health tech. In this final post, I am looking at technical terms that apply outside purely health technology, but which are particularly important to health techs

Federated Machine Learning

From the highly medical term of genomics, we skip to the highly technical term of federated machine learning. To understand this term we need to begin with machine learning. While different machine learning algorithms use different approaches (from trial and error to guided ML which uses a mix of rules and ML), there is one constant. Machine learning requires huge amounts of data. At its heart, ML is a probabilistic approach, meaning it is based on probability and statistics. And as many of us remember from high school stats, the best way to decrease the rate of error is to grow the sample. This is one of the reasons why ML needs so much data. 

But how do you handle the need for data when the data in question is highly sensitive patient data and is stored in various different places? The risk of taking all that data and trying to store it in place is simply too great.  This is where federated machine learning comes into play and why you are beginning to hear so much about it in the health tech space. Federated machine learning is able to connect to various different data sources and use anonymised data to run its algorithms on.  

APIs

API is such a common term that it’s gone beyond a buzzword. But it is an example of a term that is thrown around by tech people so often that it can be confusing. API means application programming interface and is a way to connect different pieces of software together or to attach them to a data set. APIs allow easy connections that in the past would have required more intensive coding to build. 

Microservices 

Microservices architecture is a software architecture approach that uses a collection of small, easy to test, reliable “services” interconnected with APIs to create a complex application. It is often used in health tech because health tech software is complex, needs to connect to various different third parties (health insurers, medical practices, hospitals, pharmacies, national health services and more). Microservices are similar to lego bricks which are individually simple (and therefore less prone to bugs and security risks) but can be used to create powerful and complex applications. Breaking down complex problems in small has massive security benefits and protections, which is critical in health tech.  

Our vision at Vertrical is that technology should serve the business’ goals and that buzzwords and tech speak can confuse and distract. And, while I am happy to geek out with developers about tech, I think that we need to discuss technology in terms of business value and what it offers not what it does!