Artificial Intelligence | News | Insights | AiThority
Bitcoin
$11,730.89
-145.33
(-1.22%)
Ethereum
$393.63
-6.06
(-1.52%)
Ripple
$0.31
0
(+1.52%)
Litecoin
$59.74
+0.13
(+0.22%)
EOS
$3.18
+0.11
(+3.71%)
Cardano
$0.14
0
(+0.21%)
Stellar
$0.11
0
(+0.38%)
NEO
$13.22
+0.13
(+0.99%)
NEM
$0.06
0
(+4.33%)
DigitalCash
$99.70
+1.74
(+1.78%)
Tether
$1.00
-0
(-0.14%)
Binance Coin
$22.80
-0.34
(-1.47%)
QTUM
$2.76
+0.08
(+3.07%)
Verge
$0.01
0
(+5.77%)
Ontology
$0.73
+0.01
(+1.81%)
ZCash
$94.50
-0.93
(-0.97%)
Steem
$0.21
+0.01
(+3.85%)

The Application of AI in Fighting Heart Disease

0 3

The Application of AI in Fighting Heart DiseaseThe AI market has boomed and holds great promise in various sectors with the hope that the technology will improve processes and reduce overall costs. The goal of efficiency is particularly applicable for the healthcare space where the AI market is projected to reach $34 billion worldwide by 2025. This commitment to innovate technology will help to address the challenges that the industry has long been plagued with since the adoption of the Affordable Care Act’s model of value vs. volume. While the power of this technology has yet to be fully adopted, we’re already seeing promising applications of it today, specifically within the cardiology space.

Heart disease, according to the Centers for Disease Control and Prevention (CDC), is attributed to more than 600,000 deaths every year in the U.S, and nearly half of U.S. adults have cardiovascular disease. The impact of heart health issues and heart disease is far and wide reaching and a major priority in healthcare. One of the challenges in cardiac care is testing methods have remained largely the same since the 1960s despite technology continuing to advance in other areas. The unfortunate consequence of this stagnation is that patients are often subject to multiple tests or have to undergo invasive angiograms – a procedure which requires a wire insertion into a coronary artery — and is often found to be unnecessary in retrospect.

Read More: What Is Automated Machine Learning (AutoML) and Is It About to Put Data Scientists out of Work?

With the help of AI and through innovation, we are luckily moving away from these outdated methods of diagnosis. One example is the HeartFlow Analysis, which uses advanced imaging to create an anatomic rendering of the coronary arteries, and applies known data about coronary physiology to better reveal underlying patient physiology. The process relies on deep learning that has been trained using a wealth of CT image analyses to create a high-quality computer model of the actual anatomic structure of a patient’s heart. Once this patient-specific model is extracted from the imaging data and reviewed, the HeartFlow process applies physiologic principles and computational fluid dynamics to compute blood flow characteristic values at every point in the model to help physicians understand the impact that blockages have on blood flow. With the deep learning methods employed, physicians are able to get valuable clinical data even for patients with highly calcified arteries – patients whom physicians would have had to send for an invasive coronary angiogram without HeartFlow technology.

What makes this technology so powerful is that as the deep learning algorithms are trained on more data, the performance of the next generation product improves. Similar to how humans learn and are able to keep adding to our knowledge base, deep learning algorithms continue to improve as new data is ingested, but at a much more rapid pace. Despite AI’s promise in healthcare, it should be cautioned that technology alone cannot transform healthcare. The AI/human collaboration is vital, especially in cardiology where it can mean life or death. For example, with the HeartFlow technology, we have expert human analysts who review and correct the work of our deep learning algorithms. This supervised deep learning method enables a virtuous cycle, whereby the algorithms improve through correction by humans, and successive generations of the algorithms then need less correction. With the addition of auxiliary information, for example, invasive coronary angiography data in those patients that are treated using invasive therapies, (i.e., coronary stents) further improvements can be made in the reliability of these deep learning methods.

Read More: Why Is Explainability Important and How Can It Be Achieved?

Additionally, we don’t see AI replacing healthcare jobs. In fact, we see AI as further aiding physicians in their roles and potentially creating new ones. Physicians are needed to interpret insights and understand how to apply them to patient care. This sentiment towards AI is also shared by patients. In a recent HeartFlow consumer survey of more than 1,500 respondents, 78 percent trust the technology to assist doctors with their tasks. To truly adopt and integrate AI, the healthcare industry needs to look to further refine the technology currently in their toolkit. What can we improve and how can AI help? Asking these questions and addressing them with clinical data and detailed research illustrates AI’s effectiveness, and creates trust which will motivate patients to ask for it, and physicians to use it. This consistent approach will scale services and create more adoption among healthcare providers.

We are only beginning to see how AI will transform healthcare. Particularly, with heart disease as the no. 1 killer of both men and women worldwide, we see a huge opportunity to help save millions of lives while simultaneously reducing healthcare costs by billions of dollars by applying these advanced technologies in this space.

Read More: In Appreciation of The Technologists Who Simplify Our Lives

Leave A Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.