By Jason Roys
When MIT artificial intelligence researcher Regina Barzilay received a breast cancer diagnosis in 2014, her first questions revolved around concerns for her son and her own survival. But soon she asked, “Why couldn’t this have been diagnosed sooner?”
Her question led to the development of one of the first AI-based systems for early detection of breast cancer. Breast cancer diagnosis is just one example of how AI applications are helping people in the healthcare sector.
Amid the feverish conversations happening these days around AI — a technopanic to borrow a term from AI author and analyst Adam Thierer — we must not lose sight of how applications of artificial intelligence are transforming healthcare.
In this article, we’ll look at applications of artificial intelligence in healthcare that are enhancing healthcare services and advancing the healthcare industry itself.
HOW AI TECHNOLOGY IS ADVANCING HEALTHCARE
AI in healthcare is rich with medical data, algorithms, predictive analysis, deep learning, neural networks and insights, and it is constantly evolving and adapting to the needs of the healthcare industry and its patients. What follows is a high-level briefing on 10 of the most important developments.
1. DIAGNOSIS
AI lets medical professionals sift through reams of health data — cases, images, patient records and notes, for example — to uncover what might be ailing their patients. Data insights are being used to diagnose cancer, triage critical findings in medical imaging, flag abnormalities, diagnose cardiac arrhythmias, determine size and location of strokes, and identify rare diseases.
There are still limitations, however. Machine learning and deep learning algorithms require data sets that are large, organized, well-classified, and accurate. Understandably, healthcare professionals may be reluctant to trust a diagnosis until there’s enough transparency to explain how the AI algorithms work.
2. WEARABLE TECHNOLOGY
Remote patient monitoring became a more urgent need during the COVID-19 pandemic, when patients couldn’t see their doctors in person. In response, the FDA issued a policy to facilitate remote monitoring devices.
You may have seen TV commercials for KardiaMobile, a one-lead personal EKG monitor that detects atrial fibrillation. Chicago-based Sibel Health has FDA clearance for its wearable AnneOne sensors for home or hospital use that detect heart rate, respiration, step count, skin and body temperature, and fall count.
The implications are mind-blowing for the management of chronic diseases, such as hypertension, diabetes, COPD, heart failure, and asthma. Remote patient monitoring can result in fewer preventable hospitalizations, a reduction in the cost of care, and better patient experience.
3. MEDICATION MANAGEMENT
More than 130 million Americans take one or more of the 20,000 or so FDA-approved prescription medications available in the marketplace. Every year, according to one study, 7,000 to 9,000 people die from medication error, not to mention those who suffer adverse reactions, both physical or psychological.
AI-powered solutions have the potential to greatly improve patient safety. Relatively simple algorithms can match medications with a patient’s condition and budget while identifying potential negative interactions. And wireless systems, for another example, can monitor whether a patient is using an inhaler or insulin pen correctly.
4. DRUG DISCOVERY
MIT and its Sloan School of Business offer an executive short course on artificial intelligence in pharma and biotechnology. “The pharmaceutical industry has begun a very significant transformation in the last few years,” says Regina Barzilay, whom we met earlier, in an accompanying video. “While previously machine learning was considered to be some kind of exotic development, now it’s really the core of the pharmaceutical business.”
In a previous article, I wrote about a Canadian company that has announced a therapeutic drug against COVID-19 designed entirely by generative AI — thought to be the world’s first.
And just recently, drug development firm Absci unveiled a paper about its novel antibodies against HER2, a gene linked to some forms of breast cancer. Here’s the amazing part: The AI model, according to an article on ZDNet, had been fed no training data on existing, successful antibodies against HER2, and no explicit information about how to bind to HER2. That makes it the first “de novo-designed antibody using zero-shot generative AI,” founder and CEO Sean said.
5. CLINICAL TRIALS
Those newly developed therapeutics, of course, must proceed to clinical trials. AI tools in clinical trials have the ability to overcome data-intensive barriers to accurate results, including unstructured data, disconnected systems, extensive manual effort, repetition, and challenges in integrating data from new sources. AI can also help to predict health outcomes for patients to reduce potential harm.
AI makes it possible to carry out decentralized clinical trials, which cast a wider net for participants in a variety of community settings. Care must be taken, however, to eliminate inherent bias from the data set of possible participants.
6. ROBOTIC SURGERY
Surgical robots supplement and augment a surgeon’s skills. Robotic surgery arms are fitted with tiny laparoscopic clamps that can hold tools for minimally invasive surgical procedures. This can include everything from coronary bypass to joint replacement.
We’re a long way from autonomous, AI-driven surgical robots that operate without human intervention. However, computer-manipulated devices driven by machine learning can allow surgeons to focus on the complex aspects of a surgical procedure.
Asensus Surgical has a performance-guided laparoscopic AI robot that provides information back to surgeons, such as size of tissue, rather than requiring a physical measuring tape. What’s more, every surgical procedure is analyzed, gleaning insights that will become available in real time during future procedures.
7. VIRTUAL NURSING ASSISTANTS
Many healthcare systems and insurance plans offer nurse-staffed call centers. With the current shortage of healthcare workers, these are expensive to operate and reach only a small percentage of patients.
Now, there is AI nurse assistant technology that proactively identifies and addresses patients’ concerns. Accessible by voice, text and chat, these machine learning systems can improve patient care while reducing labor costs.
One such innovative system, Care Angel, announced in December 2022 that it had significantly increased clinical capacity of a major insurer while reducing readmissions and labor costs. “Angel is a data-driven platform that helps deliver high touch experiences to bend the cost curve while ensuring faster, easier access to care with the appropriate interventions,” said its CEO, Bud Flagstad.
8. INDIVIDUALIZED MEDICAL TREATMENT
Precision medicine requires understanding disease based on a patient’s individual biological data, including medical diagnoses, electronic health records, clinical presentation, laboratory studies and imaging, along with environmental, demographic and lifestyle factors. The power of machine learning lies in its ability to analyze data from these disparate sources, find patterns and contribute to more precise treatments.
AI is rapidly changing the landscape for cancer patients. In 2022, Louisiana State University researchers announced a cancer drug discovery engine powered by AI that soon could match any type of cancer, based on a small cell sample from a patient, with the drug most likely to be effective.
While machine learning and precision medicine hold much promise, more work needs to be done to test and validate AI-driven clinical decisions.
9. DATA MANAGEMENT AND RECORD KEEPING
Health data is derived from myriad sources: patient data in electronic medical records, insurance claims, genomic and diagnostic tests, clinical documentation and notes, medical devices that make up the Internet of Medical Things (IoMT), among others. Data science is being used to seek, collect, store and standardize medical data regardless of the format, assisting repetitive tasks and supporting clinicians.
One example is Prognos Health, which uses machine learning to compile and analyze existing data from prescriptions, medical claims, lab results, and other sources. The Prognos Factor data management platform enables companies to quickly acquire health data to detect diseases, underwrite policies, and note gaps in care.
10. HEALTHCARE SYSTEM ANALYSIS
With ever-tightening operating margins, healthcare organizations are on the hunt for efficiencies that will save a few bucks. One of the areas ripe for AI is in analyzing a healthcare system’s revenue cycle to find opportunities to optimize revenue, contain costs and improve the patient experience.
One example is prior authorization, which often involves time-consuming siloed and manual processes, leaving healthcare providers and patients wondering whether a procedure has been approved.
MidLantic Urology, a practice in Pennsylvania, adopted an AI solution and has found that it has improved workflows and increased gross revenue while providing medical professionals and patients with real-time actionable insights about approved treatments and claim status.
However, adoption of these technologies has been relatively slow because there’s still a “black box” feel to it. More demonstrable outcomes like those above will be needed to tip the balance.
MAKING THE MOST OF AI APPLICATIONS
Healthcare organizations contemplating adoption of AI technology are being bombarded with pitches from companies that promise their artificial intelligence application will speed up diagnosis, improve patient outcomes and save money. Few organizations have the expertise on staff to give these applications the scrutiny needed to make a good decision.
A technology consulting firm like SDV INTERNATIONAL can open up the “black box” and help an organization decide on the right technology. Our expertise in AI and our dedication to our clients means your healthcare organization can trust us to make the best recommendations to serve your patients and move your organization forward.
Contact SDV INTERNATIONAL today at our website, by calling 800-738-0669 or by emailing info@SDVInternational.com.