This article is more than 5 years old. Information may no longer be current.
Progress in mobile technology can benefit cardiologists, patients
Click Here to Manage Email Alerts
Advances in digital health, artificial intelligence and mobile technologies may help physicians and patients prevent and treat CVD, according to three review articles published in the Journal of the American College of Cardiology.
Artificial intelligence
AI can help clinical practice become more convenient, efficient, effective and personalized, according to a review by Kipp W. Johnson, an MD/PhD candidate at the Institute for Next Generation Healthcare and the department of genetics and genomic sciences at the Icahn School of Medicine at Mount Sinai, and colleagues.
“Artificial intelligence has clear potential to enhance every stage of patient care — from research and discovery, to diagnosis, to selection of therapy,” Joel Dudley, PhD, director of the Institute for Next Generation Healthcare at the Icahn School of Medicine at Mount Sinai, said in a press release. “A key next step to incorporating artificial intelligence into cardiology is to align available data and technologies with clinical and business use. This way, we can prioritize short-term opportunities and understand gaps in available data or algorithms that are holding back applications of artificial intelligence in areas of high clinical need.”
Through AI, physicians can monitor patients remotely, in addition to interpreting the data and responding to it. This technology can be used for unsupervised or supervised learning, which allows physicians to discover underlying relationships within a dataset or classifying an observation into categories or outcomes.
Machine learning, another aspect of AI, can be used for feature selection, which is critical for predictive modeling. Through this, physicians can predict whether a patient with a CV condition will be readmitted to the hospital.
Although clinicians typically work with dichotomized outcomes, this can lead to a loss of useful information on relationships. Instead, clinicians should stay from making binary classifications and move onto predicting individual patient probabilities.
Several algorithms can be used when treating patients with CVD, including regularized regression, tree-based methods and support vector machines.
Future applications of AI include deep learning, which currently utilizes features of modern technology such as facial recognition and high-quality voice recognition, the authors wrote, noting that deep learning applications in cardiology are limited at this time, although a lot of applications are focused on image processing and analyses.
“The incorporation of AI into cardiology is not something that clinicians should fear, but is instead a change that should be embraced,” Johnson and colleagues wrote. “AI will drive improved patient care because physicians will be able to interpret more data in greater depth than ever before.”
Digital health technology
Researchers have yet to explore the integration of digital technologies in clinical trials, although this can lead to more streamlined and realistic trials, according to a think tank meeting held in December 2016 that was detailed in a review paper published in the Journal of the American College of Cardiology.
These technologies can be used as a diagnostic tool for common diseases, Abhinav Sharma, MD, fellow at the Duke Clinical Research Institute at the time of the study and now an advanced heart failure fellow at Stanford University, and colleagues wrote. Several studies are being conducted that focus on wearable sensing technologies such as the Apple Watch to identify atrial fibrillation.
Health outcomes may also be improved through digital health technologies by increasing patient engagement in caregiver care and self-care.
“Decision support apps allow for disease management to occur outside of clinics while empowering patients to optimize their health conditions, but a key driver in their success is how user-friendly or complex these devices and electronic communications are to the patient,” Sharma and colleagues wrote.
Trial recruitment may be enhanced through the utilization of digital health technologies, in addition to data collection.
Informed patient consent has been made easier to understand through mobile app-based consent, electronic informed consent and video consent.
“Despite the perceived advantages of these strategies, issues such as ensuring subjects do not ‘click through’ agreements, difficulties in identifying the person consenting and assessing capacity for understanding will have to be addressed before these strategies can be widely adopted,” Sharma and colleagues wrote.
There are several barriers that prevent the implementation of digital health technologies, including productivity paradox, quality of data, privacy, accessibility and the need for regulation.
The use of these technologies in clinical trials can streamline costs, although validating these apps will need to be done through early engagement with regulatory agencies.
“Close and early collaborations between stakeholders will be required to ensure that digital health technologies not only improve outcomes, but add value to health care systems, decrease cost and improve quality of care,” Sharma and colleagues wrote.
Advances in AF
Mobile health can be used to help patients improve on their amount of physical activity, which can now be measured through a number of mobile and wearable devices. Even with this convenience, patients typically overestimate the amount of physical activity they do, Michael V. McConnell, MD, MSEE, clinical professor of cardiovascular medicine at Stanford University School of Medicine and head of cardiovascular health innovations at Verily Life Sciences, and colleagues wrote.
There is moderate to strong evidence that shows that wearable devices and mobile phones increase physical activity, although evidence on the effects of mobile health interventions on improved cardiorespiratory fitness is limited. Improving a behavior such as physical activity takes more than purchasing an app or mobile device, according to the review.
Mobile health can be used beyond prevention for the management of CVDs from hypertension to HF and CAD, especially since wearable and mobile devices can measure heart rate and rhythm.
Even with these benefits, it can be difficult to detect and manage AF because episodes can be either asymptomatic or paroxysmal. Monitoring heart rate and rhythm can aid in the management of AF without implanting devices into the patient, according to the review. Several recent studies have shown that mobile ECG- and photoplethysmographic-based detection of AF have been highly accurate.
“Despite these promising early results, a major concern for broader use of AF detection algorithms is loss of specificity, as other rhythms that create irregularity (eg, premature atrial or ventricular contractions, supraventricular tachycardia, atrioventricular block) could potentially be misclassified as AF, leading to improper diagnosis and treatment,” McConnell and colleagues wrote.
The promotion of positive and broad health benefits using mobile devices has become more widespread and accessible to many patients across the world.
“The growing involvement of CV societies, government funding agencies, technology companies and regulatory bodies will accelerate the translation of [mobile health] to improve CV disease and prevention, with a shared mission to promote equitable health and well-being and a world free of CV disease,” McConnell and colleagues wrote. – by Darlene Dobkowski
Disclosures: Johnson reports no relevant financial disclosures. Support for the think tank meeting was provided through registration fees from Amgen, AstraZeneca, Bayer AG, Janssen Pharmaceutical Companies of Johnson & Johnson, Medtronic, MyoKardia, Qualcomm Life, Sanofi and St. Jude Medical. Sharma reports he received research grant support from the Alberta Innovates Health Solution Clinician Scientist fellowship, the American Heart Association Strategically Focused Research Network-Heart Failure, Bristol-Myers Squibb/Pfizer, the Canadian Cardiovascular Society Bayer Vascular award, the European Society of Cardiology Young Investigator research grant, Roche Diagnostics and Takeda. McConnell reports he is an employee of Verily Life Sciences. Please see the studies for all other authors’ relevant financial disclosures.
Perspective
Back to Top
Jagmeet P. Singh, MD, ScM, DPhil, FACC, FHRS
These papers highlight the potential of AI and digital health technology in the way it will influence the practice of cardiology in the near future. Importantly, both AI and digital strategies will make the delivery of care both more patient-centric and cost-effective. The advances in mobile and wearable devices and sensors will help initiate the transition from conventional transactional care to a more continuous form of managing our patients.
Right now, there are multitudes of wearables and applications, and many more being developed on a daily basis. These papers highlight the importance of a thoughtful approach to validation and integration of these technologies into clinical practice.
In the immediate future, I see the greatest benefit in the management of chronic diseases such as AF and congestive HF. These two conditions are the most common CV causes for hospital readmissions and in turn constitute the greatest draw on health care resources. Mobile technologies can help monitor and manage these patients on a continuous outpatient basis and facilitate proactive interventions to prevent HF exacerbations and readmissions. Also, wearable technology can promote healthy lifestyle changes such as increased physical activity, weight loss and reduced sleep apnea; all of which help in preventing AF recurrences.
It will be a while before AI becomes integrated into clinical practice. Much of this is contingent on the ability to generate large “clean” training datasets to allow for machine learning and the development of algorithms. This is going to require academic institutions to not only work together, but also with industry.
On the mobile health front, first and foremost, we have to ensure that the data being generated by wearables and apps is “valid and reproducible.” Subsequent to the acquisition of these data, we have to ensure that this is systematically integrated into the electronic health records to enable its use for research and managing our patients. Our research there onwards needs to be geared to see if it truly impacts clinical outcomes and makes medicine more cost-effective.
The future delivery of CV care will be underscored by the integration of telehealth and m-Health strategies that are powered by the knowledge and decisions supported through AI.
Click Here to Manage Email Alerts