Machine learning and artificial intelligence are in vogue right now–likely the most popular technology terms on the whole of the Internet. The applications of these technologies have been predicted to revolutionize almost anything, from what we eat to how we show to how we process massive amounts of data in the Enterprise.
Healthcare is no exception to this rule, with physicians detecting cancer with deep learning and hospitals finding new efficiencies by using the technology, but there is a deeper need driving the adoption of machine learning technology in healthcare: the need for speed.
No, this is not an intentional, cheezy reference to a 1986 cult-classic film, it’s a real need, brought on by the Amazon problem: consumers (who are also patients) need things faster and more accurately than ever before. On a less “problem” driven note, so do health care professionals. Solving highly complex healthcare workflows and science-based challenges means that the more tasks we can transfer with confidence to computers, the better.
One still has to go to medical school or nursing school. One still has to receive ACLS certification and practice in blood draws. But the more leeway healthcare professionals and institutions have to move things online, likely the better, and it’s going to make healthcare better.
Ever heard that adage “get a second opinion?” All too often today, that second opinion comes at either great cost, or is an iffy factoid absorbed from the Internet. When it comes to emerging healthcare issues on an individual basis, having accurate, actionable, preventative care information is crucial.
Machine learning and empower medical professionals to understand with extreme specificity the exact medical needs of a patient.
Let’s say you have a patient who comes in with high blood pressure, cellular stress and modest obesity. A machine learning algorithm can take these and all other patient data into account and create a heatmap of data that can inform a physician of their likelihood for future problems if these trends continue. This could eventually expand as deeply as familial genetic information, providing a much, much deeper patient picture than was ever possible before.
This makes preventative care more effective, and that lead to tremendous prevention of pain and suffering as well as cost savings. It simply makes sense as a broad-based application.
Deep Care and Research
Charges by the likes of Joe Biden and others to do big things in healthcare, such as solving cancer, are common these days. Revolutions like these are on the scale of antibiotics and vaccinations, and are considerably more complex and difficult than almost any other medical challenge tackled to date.
Machine learning is the only way forward. There is too much genetic and other data to compile or program by traditional means. Machine learning models can not just cut workloads for researchers and “deep care” physicians like oncologists down dramatically, it actually drives results that aren’t possible through human analysis.
Take clinical trials for example. Essential to the development of new drugs and treatments, there has been a century-long problem of matching trial participants with trial treatments. Machine learning can mine healthcare databases on both the drug and patient side to make this match so seamless. There is so much of this data that there is no way a given physician or trial group could hope to be more effective that a computer at this process–and it matters a lot in terms of results and outcomes.
No, we won’t solve cancer in a day this way, and we likely can’t promise remarkable results only through the application of these new technologies–but there is great confidence that they will make a difference.
At the Hospital
Hospitals are technology powered marvels, and machine learning is just the type of efficiency driver that they tend to adopt. From monitoring individual patient beds to streamlining management and billing, applications of machine learning could introduce cost savings and efficiencies that everyone could benefit from.
This is particularly important as we move away from the individual clinician model of medicine to group- and hospital-based physician care. How fantastic would it be to have even a scheduling system that understands patients and can help prioritize appointment setting and flow of information between departments and specialties.
The application of machine learning makes sense in so many ways. It will be interesting and exciting to see the applications and reap the benefits.Comments »