Evolution of Data and Artificial Intelligence in Healthcare
Tabi Campagna, MBA and BSN RN | 11/5/2020
Imagine yourself standing inside a store. In today’s world you are laser focused, list in hand and a mask covering your nose and mouth. You generally know what aisle you need and what you are purchasing. Shopping has turned into more of a tactical mission. No more promises of dreamily floating through the aisles of Target with a mocha-latte in your hands and exiting the store with goodies you did not even realize you needed. While bent over scanning what brand of toothpaste best suits your needs, someone nearby coughs. The hairs on the back of your neck stand up a little– you think to yourself- are they sick? Do they have the ‘rona? You slowly start backing up as you make eye contact with the “offender” … they are wide-eyed looking right back. “It’s allergies, I promise!” they say, but how do you know–who do you trust? What if you already had the ability to assess the risk of the situation sitting in the palm of your hand? According to science, we should trust data and artificial intelligence (AI), that is who!
The promise of data and AI providing faster and more accurate diagnosis to many of today’s most common disease processes is not a new concept. Back in the 1960’s, Dendral became the first problem-solving program created at Stanford University. This led to MYCIN, a backward chaining expert system in the 70’s. MYCIN utilized AI to identify bacteria, recommend antibiotics, and would adjust dosage for the patient’s body mass index appropriately. While this wasn’t widely adopted by physicians at the time, it did kickstart the movement through the 80’s and 90’s that would lead to incredible advancements and improvements through modern day in computing power, growing genomic sequencing databases, the widespread use of electronic health records, robot-assisted surgeries, improving deep learning techniques and data logs for rare diseases, as well as enabling machines to replicate human perceptual processes with improvements in natural language processing and computer vision.
Fast forward to today with analytics taking the forefront in screening and diagnosis. Utilizing AI, scientists at the Massachusetts Institute of Technology (MIT) have just built an algorithm that can distinguish subtle differences in the cough of a person with coronavirus whether they are symptomatic or not. Because the virus has a specific impact upon the lungs and vocal cords there is a slightly different sound from a healthy person’s cough that is generally considered inaudible to a human. According to the MIT study, the model has a 98.5% sensitivity with subjects diagnosed with an official test, and a 100% sensitivity with asymptomatic subjects. It concludes that that AI could augment the current approaches of containing the spread of COVID-19 in a “free, non-invasive, real-time, any-time, instantly distributable, large-scale COVID-19 asymptomatic screening tool.”
What does this mean to us? This means that right now the promise of an application being built that can be used to screen people in a more reliable and meaningful way than temperature screening is on the horizon. Practical use cases of this could include private, daily pre-screening for students and employees, for use prior to running errands and engaging in leisure and sporting activities, public use for transportation systems, and best of all for quickly recognizing, diagnosing and reducing outbreaks in groups. Use of this algorithm and potential application can easily allow for responsible quarantining to contain and then stop the spread of this life-altering virus. This gives us hope that there could soon be an end to the OK Corral-style showdowns in the aisles at Target.
While my depiction above takes a break from the heaviness of illness and takes on a lighthearted tone, the threat of coronavirus is real. Continued advancements within AI has the potential to reduce the prevalence of the disease, its severity and can guide policy for interventions.
While many organizations dream in AI, Tuatara helps make that real, helping you wring every ounce of value out of your data. Capabilities ranging from strategy to business analytics, to predictive, data visualizations, data warehousing, and more help bring unprecedented insights to life and maximize the value of data and business impact. Tuatara can help you leverage machine intelligence and AI to capture the unexploited areas of business data.
Is this an opportunity within your company? Book a 30-minute meeting with our team to discuss how we can help your organization.
Tabi Campagna is Tuatara’s Partner Executive who is responsible for developing strong partnerships with Big Tech companies and customers from small to enterprise level organizations.