We believe in the importance to raise awareness to heart diseases and strokes and hope we can help AHA/ASA raise funding, a medical research community looking to raise funding for NIH heart and stroke research, through this tribute page to provide better patient outcomes.
Please help us reach our goal by supporting our fundraising efforts with a donation today. Your tax-deductible gift will make a difference in the lives of many! It is faster and easier than ever to support the fight against heart disease and stroke - America's No. 1 and No. 4 causes of death. You can make your donation online by clicking the donation link.
Thank you for your continuous support! In words of Jackie Alfau, "Don't forget to smile always!".
With love,
My Personal Web Log
Return to the main discussion page
Can Machine Learning And AI Head Off Heart Attacks And Strokes?
More than 33 million people worldwide have an irregular heartbeat, and the vast majority of them probably don't know it. A person with this condition is five times more likely to suffer a stroke, which happens every 40 seconds in the U.S.
Machine learning and artificial intelligence more broadly could help reduce strokes and heart attacks by analyzing patient data to spot trends and make predictions. But those tools are only beginning to be applied in healthcare, and such a widespread effort will take planning, coordination, and education.
In a recent discussion with Dr. Jonathan Baugh, D.O., an emergency medicine physician and cofounder and chief medical officer of Remote Health Solutions, a provider of telemedicine devices and services, I asked him to help diagnose the situation.
Dr. Baugh: Remote patient monitoring is a big push in healthcare right now. We send a patient home with a personal tablet that has peripherals for monitoring blood pressure, oxygen saturation, glucose level, body mass index, and other health data. That data is updated daily by the devices and sent to the cloud, where it's analyzed for possible intervention.
Say it's a congestive heart failure patient and his blood pressure is up and he gained several pounds overnight. That data is flagged, the monitoring technician is alerted, and the technician or physician connects with the patient remotely. That's a telemedicine opportunity: Let's recheck vital signs, change medications based on physicians' orders--increase Lasix, increase blood pressure med, etc. Let's see if we can keep this person at home, where patients typically do better, especially with proper monitoring and care.
Oracle Healthcare
ORACLE
Walker: Basic vital sign information can tell us a lot, especially over time. By using Internet of Things devices to stream data to the cloud, and machine learning to look for patterns or trends in the data that may reveal an underlying condition such as arrhythmia, we can prevent a heart attack or a stroke.
Dr. Baugh: We've treated patients in their homes using remote monitoring techniques for years, and now that number is going to increase significantly. Changes to Medicare by the CMS (Centers for Medicare & Medicaid Services), effective January 1, make remote monitoring financially viable for many more potential patients. It's because CMS sees the benefit of preventive care instead of reactionary care.
Walker: It used to be that reimbursement for remote care was tied only to video-based telehealth. Now Medicare patients can be reimbursed for using data monitoring devices. Predictions I've seen put the number of remote-monitoring patients at more than 50 million by 2021.
Dr. Baugh: When you're talking about monitoring a couple of patients, that's easy--humans can do that. But when it comes to monitoring thousands, hundreds of thousands, millions of patients, that's where we have to use emerging technologies: artificial intelligence, machine learning, data analytics. It's about quantity as well as individual care, because we know we can improve outcomes through remote monitoring.
Walker: There's a business case as well. Benefits include reducing hospital length of stay, readmissions, and costs associated with patient transportation. In 2017, the CMS penalized about 80% of the hospitals it examined for excessive readmissions within 30 days of discharge.
Dr. Baugh: Recently, our company began work on a research project about incorporating additional remote monitoring capabilities into our telemedicine platform for warfighters who are injured and stuck at the front lines--what's called prolonged field care, or PFC. That same technology will be used increasingly with civilian patients in their homes or in facilities to provide better access to care and help improve outcomes.
Walker: If you had talked about "the virtual hospital" a couple of years ago, you'd have probably been laughed out of the room. But this is exactly what we're talking about: visibility and treatment of the patient wherever that patient resides. It's good for the patient and it's good for the healthcare system.
Dr. Baugh: In my current practice, connected care is becoming more and more common. We talk about EKG machines or telemetry monitors and how they're able to flash up some very simple rhythm acknowledgments.
But that's only simple data recognition, not really telling us a lot about the patient. As we're able to figure out machine-learning processes to monitor for and recognize risk factors related to strokes or heart attacks or an escalating disease process, that's the way we will provide increasingly better outcomes. That's directly changing and saving lives.
Walker: Machine learning is a relatively new technology that has great potential in healthcare to help harness data and address situations that can be extremely time sensitive.
The American Heart Association estimates that fewer than 30% of those who suffer strokes get needed treatment within the recommended 60-minute timeframe. Connected care reduces door-to-needle times for remote patients by alerting hospitals, neurologists, EMS teams to the critical condition.
Machine learning might take that care one step further, for example, by helping to identify the root cause of the stroke. If the root cause isn't identified, the patient might be released from the hospital, go home and have another stroke.
Dr. Baugh: Conventional wisdom when it comes to telemedicine is simple: a video connection with a physician. And that's OK, that's a base. But when we can combine that with data collection and analytics, then we're really able to change outcomes.
That's a lot of where our forward drive is--not just to deploy devices and connect patients with remote physicians, but also improve outcomes by applying more AI and ML to the evaluation of patients. If you're monitoring 50 million patients in their homes across America, it can't just be through human technicians reacting to alerts. You have to have that data being processed and analyzed, the system recognizing factors, watching for and pulling out opportunities.
Walker: The concept of virtual, connected care is becoming a reality. Now that remote patient care using streaming vital-sign data will be reimbursed, telemedicine will advance into the 21st century.
By connecting patients with clinicians regardless of location, health issues can be evaluated faster, specialists can provide treatment sooner, and patterns can be detected earlier. And now we have the ability to use emerging technologies such as IoT, ML, Autonomous Data Warehouse, and analytics to enable effective data-driven care.
It's really a combination of technologies that allows for these new ways to improve access to care and provide for better healthcare outcomes. The impact on patients and their providers will be profound--and sooner than you might imagine.
Embracing digital transformation in healthcare
Michael Walker, global lead for healthcare in Oracle's Industry Solutions Group, has more than 25 years of experience in healthcare, with both payers and providers, in supply chain, product strategy and operations executive positions.
by Jackie's Daughters on Fri, May 17, 2019 @ 3:26 PM
Permanent Link
No Comments