Sign up for our newsletter and get the latest big data news and analysis.

Artificial Intelligence and the Move Towards Preventive Healthcare

In this special guest feature, Waqaas Al-Siddiq, Founder and CEO of Biotricity, discusses how AI’s ability to crunch Big Data will play a key role in the healthcare industry’s shift toward preventative care. A physicians’ ability to find the relevant data they need to make a diagnosis will be augmented by new AI enhanced technologies. Waqaas, the founder of Biotricity, is a serial entrepreneur, a former investment advisor and an expert in wireless communication technology. Academically, he was distinguished for his various innovative designs in digital, analog, embedded, and micro-electro-mechanical products. His work was published in various conferences such as IEEE and the National Communication Council. Waqaas has a dual Bachelor’s degree in Computer Engineering and Economics, a Master’s in Computer Engineering from Rochester Institute of Technology, and a Master’s in Business Administration from Henley Business School. He is completing his Doctorate in Business Administration at Henley, with a focus on Transformative Innovations and Billion Dollar Markets.

In October 2000, Google co-founder Larry Page made a luminary prediction: “Artificial intelligence would be the ultimate version of Google. The ultimate search engine that would understand everything on the Web. It would understand exactly what you wanted, and it would give you the right thing.” Fast forward seventeen years later, and artificial intelligence (AI) and Big Data are the new buzzwords in healthcare. According to a new Market Study report, the healthcare artificial intelligence segment is projected to see a staggering 40 percent compound annual growth rate (CAGR) between 2017 and 2024, resulting in a $10 billion market focused on medical imaging, diagnostics, robotic personal AI assistants, drug discovery, and genomics. When Larry Page compared artificial intelligence to the “ultimate search engine,” he was essentially speaking of AI’s ability to crunch massive amounts of data. AI’s deep learning algorithms are designed to detect features in huge, disparate datasets that are not discernible to entire teams of data scientists. Second, these deep learning algorithms can be trained to provide specific information, or in Page’s words, “[AI] would understand exactly what you wanted, and it would give you the right thing.”

Clinical AI implementation promises a healthcare system that is preventive rather than reactionary. Today, patients are often diagnosed with a chronic condition, such as cancer or diabetes, when it’s too late to reverse the progression of the condition. Treatment plans with late stage diseases are expensive and debilitative. Patients are poorly equipped with feedback and insights into their own health conditions, and so are less proactive about making healthy lifestyle choices and adhering to physician advice. Consequently, preventing the start of chronic disease and managing the disease post-diagnosis has become the focus of preemptive measures in healthcare. Here, AI offers a promising solution. A preventive healthcare system will capitalize on AI’s ability to collect, compile, and analyze data to facilitate three progressive and ultimately integrated stages of learning. First, AI will enable a broad scope of learning that will aid in more effective and efficient disease diagnostics based on historical data. The insights gleaned from this massive survey of Big Data will be utilized by physicians to further train AI. Second, AI will harness historical data and augment it with real-time patient data to provide feedback to patients. Finally, as AI begins to learn how patients react differently based on real-time data, it will create personalized and predictive feedback for each patient.

Broad Learning for Effective and Efficient Diagnostics

AI is distinguished by its analysis capabilities and its deep learning algorithms. These capabilities can be deployed to traverse massive amounts of data and detect a few variables across hundreds of thousands of data points that are specific to certain conditions and diseases. In the context of broad learning, AI holds the potential to aid in the diagnostic process and identify problems before they become serious.

Researchers from Sutter Health and the Georgia Institute of Technology are using deep learning to analyze electronic health records to predict heart failure before it happens. Initial results have empirically demonstrated that this AI application can accurately predict heart failure one to two years early. Philadelphia’s Thomas Jefferson University Hospital has researchers training AI to identify tuberculosis on chest X-rays, an initiative which may help screening and evaluation efforts in TB-prevalent areas with limited access to radiologists.

By leveraging public historical data sets licensed from research groups such as the Mayo Clinic or the American Heart Association, with patient-specific data such as medical history, individual symptoms, and prescribed medications, AI will enable physicians to identify a specific condition while ruling out others. Then, they’ll be able to recommend the best course of treatment based on the individual patient.

Augmenting Broad Learning with Real-Time Patient Data

Once AI’s broad learning can identify and assist in the diagnosis of a specific condition, it can leverage historic data to develop treatments plans that are interactive, driving patient engagement. Doc.ai is using Blockchain technology to collect masses of medical data globally and generate insights from that information. Then, through machine learning, the data collected will be analyzed and processed to provide personalized feedback to users about their own medical issues. Studies have shown that ongoing feedback is a key factor in driving patient engagement. A 2012 trial found that when remote patient monitoring devices were given to patients with chronic conditions, the number of emergency room visits, hospital admissions, and one-year mortality rates decreased. The devices used in this study provided ongoing feedback for patients by reminding them when tests were due, offering educational videos, and creating a graphic chart detailing their recent clinical results.

It is just as easy to envision a heart disease patient equipped with a medical-grade wearable device that provides real-time metrics detailing the effectiveness of an exercise regime or medication based on the prior week’s metrics. This demonstrable, measurable feedback could encourage the patient to adhere better to a treatment plan or to consult with physicians between appointments to improve regimens and future results.

For AI to be truly “intelligent,” it needs to become more effective with experience, and this experience cannot occur with information pulled from historic datasets alone. AI requires copious amounts of data for optimization, and medical-grade remote monitoring technologies that continuously stream patient data are the ideal mechanism for this purpose. This is because these devices provide constant connectivity (through expanded broadband) combined with the capability to collect clinically accurate, medically verifiable data.

Specific Learning for Personalized and Predictive Feedback

Perhaps the most valued quality of AI is its ability to dynamically learn and improve over time. As AI collects individual patient data, and begins to learn how patients react differently to feedback, it can begin tailoring feedback so that it’s personalized and predictive. Such feedback is the foundation upon which a preventive healthcare system is built. Medtronic’s new IBM Watson-powered Sugar.IQ diabetes app uses real-time continuous glucose monitoring and insulin information from Medtronic pumps and glucose sensors to provide diabetes patients with personalized insights. The AI-based app is designed to learn from a patient’s own information input; its Glycemic Assist feature enables users to inquire about how specific foods or therapy-related actions and events impact their personal glucose levels. By following trends, Sugar.IQ can then help users discover the impact that these items have on their glucose levels.   Patient inputs also enable Sugar.IQ to learn and issue blood glucose level predictions by assessing the patient’s current situation and the risk of glucose levels falling outside safe thresholds.

Medical-grade wearables with AI could create predictions based on a patient’s daily biometrics. If a heart disease patient is prone to developing a rapid heartbeat after X minutes of walking, then the medical device would make a prediction and alert the patient to avoid exceeding the recommended minutes of walking. AI could also exercise predictive capabilities by learning what kinds of feedback instigate adherence for a patient and then applying that feedback to improve the patient’s disease management, almost like a personal health coach. When patients can follow their own progress, and see how certain choices have a direct impact on their health, they are more likely to adhere to treatment plants, engage in their healthcare, and change their behavior.

The Future of AI in Healthcare

Ultimately, the effectiveness of AI in healthcare will be directly predicated on its access to Big Data—both to historical data sets and EHRs as well as to real-time, continuous, patient-specific data from remote monitoring technologies. Training AI to reach its maximum potential is an interactive process in which physicians and patients are key players. AI applications must become fully integrated into existing healthcare systems, and must function within a “residency program” of sorts, in which they perform analytics on real-time patient data while being overseen by a physician. In this way, the algorithms will learn simultaneously from the data and from the physician’s oversight to hone their capabilities. AI’s ability to learn from experience and offer personalized and predictive feedback to patients and physicians is its greatest value proposition for preventive healthcare systems which improve diagnostics while catalyzing patient adherence through engagement. The integration of both broad and specific AI learning applications, the latter implemented in remote patient monitoring devices, represents the tantalizing future of preventive healthcare that beckons on the horizon.

 

Sign up for the free insideBIGDATA newsletter.

Comments

  1. yup it is the way we are going and could help professionals with EWS ( Early Warning Signal) that are classified MEWS, NEWS etc. based on the calibrated requirements…. Please visit our web site http://www.isansys.com we provide such predictions and provide patient safety using AI.

  2. Great measure to implement but I don’t think it will be an effective system for a while. Healthcare does need to evolve in the states.

Leave a Comment

*

Resource Links: