MRIs, Machine Learning, and the Future of Health Tech

Healthcare systems around the world are faced with an epidemic of metabolic and obesity-related diseases. Accurate and direct measurement of body composition has been prohibitively expensive in the past. Forcing clinicians and researchers to use indirect and inaccurate metrics. Such as BMI or waist circumference to assess a patient’s disease risk.

As Americans continue to become more conscious of their health and take greater ownership over their personal healthcare. Now tech companies are jumping in to provide more and better resources and tools.

I was recently introduced to a company that uses four decade old MRI technology, but is reinventing it by making it available to consumers in modern ways. Marcus Foster is founder and CEO of Klarismo, a company that builds 3D body profiles through an automated quantitative image analysis pipeline.

What is automated quantitative image analysis? Foster gave me a break down on how this tech will shape the future of health-conscious consumers:

The Importance of Understanding Body Composition

Understanding body composition in detail allows for the definition of Image Derived Phenotypes (IDP). Knowing precise volumetric measurements of different tissue types and organs. We can begin to understand what genetic and environmental factors contribute to certain phenotypes.

Foster explains that with the advances in MRI technology, we can now correlate the volume of certain muscles with other observations (e.g. a person’s mobility or strength). To understand how changes in muscle volume influence other areas.

“Most importantly we can learn how fat volume in different areas of the body, such as the subcutaneous fat underneath the skin vs. the visceral fat inside the abdominal cavity influences health outcomes,” says Foster. “This way we can give people more accurate guidance on how much or how little fat they should be carrying around with them.”

Significance of MRI Profiles Over Time

Imagine if we could give someone a detailed prediction and 3D rendering of what their body would look like in five years time based on different lifestyle choices. What they will look like at 40 if they don’t change anything, versus what they would look like if they started cycling to work every day and stopped eating wheat, or meat, or sweets, etc.

“If we can better understand in detail how bodies change over time, plus how these changes are different for people with a different genetic profile and different lifestyles. We would be able to make much more personalized predictions of health outcomes,” says Foster.

He imagines people regularly scheduling MRI scans to keep track of physical changes to their bodies over time – either as a result of ageing or lifestyle changes such as diet and exercise.

The Future of MRI Tech is Machine Learning

Foster is hopeful that the growing demand for medical imaging will lead to a reduction in the cost of the equipment and maintenance.

“Klarismo envisions a future where MRI scanning is becoming much more routine than it currently is,” says Foster. “To make this possible we need to have the ability to analyze very large numbers of images without human intervention.”

The established methods to perform quantitative analysis of MRI scans and to understand changes between two scans all involve a lot of manual work. Which simply does not scale when you want to look at thousands of images simultaneously.

Foster says his approach has always been focused on eliminating any involvement of people in the analysis. By manually annotating a large sample data set of whole body scans and training machine-learning models how to read the data. It has become possible to rapidly quantify an ever-growing number of features in the body. This computation is performed in the cloud it is highly scalable and cost effective.

“If MRI scanners would cost closer to $300,000 instead of $1.2m we should see the cost of scans drop dramatically. Which in turn will make regular non-diagnostic scans affordable for many people. Thus increase the volume of data that becomes available to help us understand how our bodies actually develop,” asserts Foster.

Additional Applications for this Research

Klarismo’s technology is enabling large, population-level imaging studies to perform quantitative analysis quickly and reliably.

“The UK Biobank project is in the process of scanning 100,000 people from neck to knee,” says Foster. “To derive body composition data or quantify organ volumes from these scans would normally require many years of analysis. However, using our technologies we can deliver results almost in real time. As soon as the subjects have been scanned for a fraction of the cost of traditional methods. Similar studies are ongoing in other countries.”

Did you find this article helpful? Click on one of the following buttons
We're so happy you liked! Get more delivered to your inbox just like it.

We're sorry this article didn't help you today – we welcome feedback, so if there's any way you feel we could improve our content, please email us at

Written by:
Sheila Eugenio a tech enthusiast and Digital Marketing Consultant. Sheila has consulted with several mediums to large businesses on how to position their brand and attract the right audience.
Back to top