Achieving UHC with Tools (Part II): How can big data and AI help create better health for all?
Coming from the earlier part of this series, we would be exploring, at a fundamental level, some of the tools for accelerating the attainment of UHC and see how we could be thinking about some of these things better. Currently, big data and its companion, Artificial Intelligence (AI) are two examples of the transformative tools we are exploring and would be looking at them here. We would be starting with how we arrived at big data and after that, how AI is helping maximize the inherent benefits. Without waiting any further, let’s take a look right away.
In the Beginning
We have a “need”, which is “better health for ourselves and everybody else.” While we naturally need better health for ourselves and those around us, in especially way we are driven (due to our convictions) or are compelled (as a result of our positions) to help make better health available to everybody.
To help solve this, we are in search of solutions. Among other things, our most immediate answer is to try to provide adequate healthcare products and services for everybody. To be able to serve our purpose, the healthcare products and services should have some fundamental characteristics. We could call these some of the basics of Universal Health Coverage (UHC), and they are:
Good Quality: They should serve our purpose; achieve required health outcomes without causing more issues.
Readily Available and Accessible: They should be easy to reach; obtainable when needed.
Very Affordable: They should not only be cost-effective but within the purchasing power of almost everybody (institutions and individuals alike).
We want or rather “need” these fundamental characteristics within a functional healthcare system where every effort is primarily focused on the patient’s outcomes improvement.
As we always desire to make things better and seek better solutions. We realize that humans are fundamentally tool builders. Humans could build tools that could amplify natural capabilities to exponential levels. So we seek tools that could help us do better. Tools are a form of leverage. It turns out tools from advancements in science and technology could be helpful. Now, it is important to note that technologies, no matter the form, are “tools,” tools to help solve problems and make things better.
Brilliantly, tech tools when applied to healthcare help us solve some of the healthcare needs and make things better. These include all the standard and novel tools for diagnoses, telemedicine, EMRs and computers, electronic reminders and lots more. We find these helpful.
Now, because the movement and storage of information in some of these healthcare tools are usually in the digital format (i.e. expressed in digits: 1s and 0s or typically represented by values of physical quantity like voltage or magnetic polarization or two levels of electric charge stored in a capacitor as is found in the semiconductor of most modern computer), we refer to this generally as “digital healthcare” and the information as “digital information”.
We may now ask: Is the future of healthcare digital? Well, maybe yes, maybe no. However, we have digitized a good chunk of how healthcare works and so far, it seems we are having more quality, available, accessible, and affordable care with more digitization. Maybe we go with progress and adopt more of digital healthcare since digitization is helping us solve more of our problems and make things better, right? Well, the future of healthcare could be digital; if and only if digital healthcare continues to serve us better. By the way, some tools could be costly especially at the early stages. However, they usually get better and more affordable with time and continued improvements.
In any case, it essential we have an open mind towards progress. What if something different and better than what we are trying to do now comes up tomorrow? Would we be open to explore and go with whatever is better overall? We should be. Usually, it starts with overcoming our mental barriers. This suggests that it is also important not to see some of these tools as the final solution always; our use of them is not cast on stone. They could be better in some situations but maybe not in others. As a result, we should be willing and able to fine-tune products and processes to serve our unique needs better.
Also, it is not always a nice idea to try so hard to copy whatever progress we see elsewhere and try to shoehorn them into our places including where or when they may not be suitable. The recommendation here is that we should be able to properly adopt and adapt progress and then also learn to make them better. We should work to be able to take whatever we have or are given and make them better. Open mindset and creativity are essential in life and especially in dealing with progress in science and tech. We need more of these in places like Africa. We would talk more on some of these later.
Getting Big on Data
As we move and store information in digital format, giving us digital healthcare, we acknowledge that information could also be referred to as data. So technically, another word for information is data. So the books we read, our research processes and findings, our characteristics, our communication, including the things we are reading or saying now; are all data or information. This includes our health information records, our body makeup, our genetic information, the lab tests we did, the diagnostic images, our social pictures and so on.
In healthcare, all the information we have given, generated, moved or stored from time immemorial are all information or “data.” So since we are recently adopting more of digital healthcare, all those info that we have given, generated, moved or stored are fundamentally digital information or data. Usually, these are called just “data.”
As we keep accumulating things (anything at all), they gradually become big, right? Same way, as we kept generating and storing digital data, gradually the data becomes “big.” So we call them “big-data” in comparison to when we had just not so much data stored. The storage is on whatever that could hold digital data, e.g., disk drives, solid state drive, optical disks and so on. As we also know, these storage devices are usually embedded into machines (e.g., computers) that help to process and store these data and bring them up when needed. Also, all of them are tools. In a nutshell, all these tools have improved our healthcare, and as they improve our healthcare, we have also generated and stored lots of data, and these data have become big, and we now call them “big-data.”
To be clear, the term “big data” refers to massive data sets that may be analyzed computationally to reveal patterns, trends, and associations, primarily relating to human behavior and interactions. It usually has the 3Vs = high volume, wide variety (structured, semi-structured and unstructured) and high velocity for processing. Is there strictly any amount of data to be called big-data? Not quite. However, usually, it is in terabytes, petabytes and even exabytes of data collected over time. We are generating more data today than ever before. For instance, every second, humanity produces 6000 tweets, 40,000 Google searches, and 2 million emails. Moreover, this keeps increasing; by 2019, global web traffic will surpass two zettabytes per year. That’s huge!
Now, for our big-data in healthcare, we are like; what do we do with them rather than let them stay there “idle,” occupying space and taking costs? Could they help us more? Could we apply them to something? Could they help our decisions, our actions or something? Meanwhile, in almost every other thing we do we refer to data (information) we have acquired and stored somehow, maybe in our brains, in books, reports or whatever. So we are used to using data for our decisions and actions. That’s not new. That’s more like life. However, for big-data, it is now too big, and we don’t bother having them all in our heads to use the analytical skills of our brains on them.
Well, the standard machines we have; calculators, computer programs and so on are already helping us somehow; at least they help us do quite a lot. However, we need something that could do more if possible. The question then is: Are there possibly other analytical tools that could do more and help us better? Maybe something that could be more “intelligent” (just as we are intelligent) in order to analyze better and derive more insights from big-data better? Good question!
If we could develop something that could help us in these “intelligent” ways, we could decide to call these “artificially” intelligent machines. So we could distinguish them from our “natural or original” intelligence! So whatever that could operate or do things humans strongly associate with intelligence; maybe machines, programs or whatever, we could call them “Artificial Intelligence.”
So we now go back to our original challenge: How do we make better healthcare available for everybody, especially where it really matters like Africa? Well, this could be a matter of urgency since Africa, according to some reports, have some of the highest global disease burdens and yet the lowest access to adequate healthcare. This stat is not sustainable and should not be acceptable at all!
So we are like; could the big-data (we already have) when married to this artificial intelligence (which is basically a machine) help our decisions and actions towards better healthcare for all? This could be a computer that can perform operations analogous to learning and decision making in humans. Put differently, this artificially intelligent “machine” would “learn” from available data and then generate insights from the data. (This is also why we would use the term “machine learning” when referring to some artificial intelligence techniques).
With likely better understanding of what our big-data is (all the information or records in healthcare especially those in digital format), we would be looking better at AI next. See you in part Trois.
About the Author
Iraneus Ogu directs the Africa Artificial Intelligence and Blockchain for Healthcare Initiative at Insilico Medicine, Inc. In addition to tech developments, he works on Longevity and Aging Interventions with his research efforts focusing on neuroregeneration. He equally works with the development team at Longenesis.com and also has a background in Pharmaceutical Sciences at the University of Greenwich, where his research focused on controlled-release dosage forms.
