Anyone who has ever entered a phrase in the search box of Google and seconds later received direct links to millions of highly-specific references to that phrase, has experienced the power of algorithms.
Algorithms are also used in almost every service and product which requires data computation – social networks such as Facebook, online retailers such as Amazon, smart phones, and WiFi systems, to name a few.
According to an article from the University of Utah Health Sciences Center – "Algorithms are Changing Healthcare"– the next frontier for harnessing the power of data to predict outcomes will be in the delivery of healthcare.
This trend towards the use of algorithmically based systems is fundamentally changing the way medicine is (and will be) practiced, and it represents a confluence of engineering, clinical medicine, and physiology.
An Expert on Mathematics and Healthcare
In an effort to shed some light on this medical "mega-trend" of algorithmically-based healthcare, here are some insights as to how mathematics is affecting medicine.
Everything I do is about modeling non-linear, dynamic systems. I work to understand the data that govern their behavior and use that knowledge to design models to solve whatever problems they are encountering.
I applied my engineering expertise among a wide-range of high-profile companies and have done pioneering work on model-based therapeutics.
In my opinion, modern healthcare has a 'manufacturing' problem, not unlike what I dealt with at General Motors. We need better efficiencies in its delivery and this can only occur when we employ the tremendous advances in computer technology. What I did in my early years – making models for the manufacture of automobiles – I now do in the area of metabolic systems.
What is an Algorithm?
While algorithms can be immensely intricate, they are also elegantly simple.
Whenever your grandmother made her wonderfully delicious and distinctly different cookies she was using an algorithm. There was a recipe and a series of steps she followed and she probably had some unique aspects that made them special.
The best algorithms have unique insights into the problem to be solved. In healthcare, the word 'algorithm' is not used as often as in the technology world. The preferred word for this tool in healthcare is 'protocol.' What a model-based algorithm does is use data from the patient to encapsulate a model of his/her dynamics – in this case, their response to insulin – so that the physician and clinicians can personalize care.
If healthcare can be delivered in an automated way via this algorithm, the patient will receive better, more consistent care with less effort and expense. Using this automation allows us to realize better productivity in probably the last area where computing and technology have not initiated significant gains in productivity and quality: healthcare. With this approach, the highly-trained and valuable resources – nurses and clinicians – are allowed to focus their efforts on the more difficult aspects of patient care.
The Future of Medicine
Mathematics is a tool to help you understand the world around you. In applying this to healthcare, it can help you describe this delivery of care. In the future, you will likely see mathematics creeping into healthcare more and more. I see tools, such as EndoTool for inpatient glycemic control being used to measure and monitor factors that are difficult to calculate manually.
You will likely see data collection being used to offer more personalized patient care and to monitor the outcome of that care. Through the use of algorithms, we will see technology moving into diagnosis, treatment, and monitoring of patients.
Another important area for mathematics in the future of healthcare involves aggregation of this data. As this data begins to accumulate over a group of patients with some similarities, we will see more datamining for research and improved treatment. This is no different from what Amazon does with the data of millions of individuals who are buyers. It uses data, captured by its algorithm and gathered from all of these customers to enable a better buying experience. With medicine, we will be able to better treat chronic conditions, such as diabetes, by using a comprehensive model built from data collected.
As noted in a recent New York Times article ("Your New Medical Team: Algorithms and Physicians") Technology is not "forgetful, fallible, or prone to distraction."
About Dr. Chase:
Professor Geoff Chase is a Distinguished Professor of Engineering at the University of Canterbury in New Zealand. His research focuses on the development, validation and clinical application of physiological models, primarily in intensive care, with a focus on saving lives and cost to increase the quality and equity of access to care. He is a member of the Medical Advisory Board for Monarch Medical Technologies. He is also a Fellow of the Royal Society of NZ (FRSNZ), the American Society of Mechanical Engineers (FASME) and IPENZ (FIPENZ) with over 1000 scientific publication and patents, and has founded 2 venture funded start-ups.