How AI is Being Used in the Healthcare Industry to Save Thousands of Lives
Artificial intelligence and machine learning algorithms are revolutionizing the healthcare industry. AI’s massive processing power and data-driven predictions apply to nearly every aspect of healthcare, from administrative management to cancer treatment. In just the next few years, AI is poised to save thousands of lives.
Artificial Intelligence Comes to Healthcare
IBM’s Watson, the cloud-based AI platform most known for beating Ken Jennings on Jeopardy! in 2011, made a high-profile move into cancer research in 2014, signaling a new era in healthcare technology. Since then, researchers have explored many applications for machine learning algorithms, prompting futurists to prophesy a sea change in the healthcare industry. AI investment will likely become a major part of the healthcare industry, with a predicted $50 billion market by 2027.
“When provided with sufficient genetic data, AI has the potential to produce highly accurate predictions about an individual’s health prognosis, particularly in relation to the prevention, diagnosis and successful treatment of genetic diseases,” says Dr. Axel Schumacher, CEO and co-founder of Shivom, a genomic data-hub startup.
AI algorithms mimic a neural network and adapt through trial and error as they analyze incredibly massive datasets. This ability to “learn” and self-program promises to minimize the number of errors generated in patient care by predicting outcomes and best practices, all based on meticulous data analysis.
Applications for AI
“Machine learning algorithms can be developed to analyze almost any type of data, so the possibilities for AI integration are nearly limitless,” says Joshua Adamson, business expert for Business.org. In the healthcare field, this includes innovations for administrative offices, supply chains, surgeries, and even waiting rooms. These AI applications are already changing the healthcare industry in 2018.
Diagnosis
AI’s ability to process hundreds of data inputs without bias can help eliminate human error from the start of patient care. If a human doctor misses a red flag or a condition during an exam, the patient’s life can be at risk. Unless you’re Dr. House, it’s also impossible for anyone to commit the tens of thousands of potential diagnoses to memory.
AI is already closing the diagnosis gap — it can comb through thousands of journals, case studies, and symptoms, going far beyond human capability.
For example, out of the 12.1 million breast exams undertaken yearly in the US, nearly half end up with a false diagnosis. In 2016, AI researchers designed a machine-learning algorithm to look for signs of breast cancer in every aspect of a patient’s medical records and exam results. The AI performed analyses 30 times faster than human doctors, with an impressive 99% accuracy over 500 patients.
Treatment
AI promises an equally strong impact on treatment, especially for patients at risk of complications from prior conditions. “If you’re admitted to the ER for pneumonia, the people who are treating you may not think about the fact that you also have congestive heart failure,” says Dr. Michael Cantor, associate professor at New York University’s Langone Medical Center, in an interview with US News.
An AI program developed by Google and NYU uses retinal scans taken by high-resolution cameras to predict cardiovascular risk factors based on the characteristics of visible blood vessels. This surprising application of AI was verified across a massive dataset of over 284,335 individuals. The study leverages AI’s ability to find meaning in massive collections of raw data, with direct implications for treatment.
Post-Treatment Care
Helping patients recover after a serious procedure is another big opportunity for AI integration. Doctors, nurses, and caregivers are hard-pressed to monitor hundreds of patients for warning signs during recovery, and 24/7 monitoring isn’t practical. An endeavor led by Cheryl Reinking. chief nursing officer at El Camino Hospital, applies machine learning to one aspect of recovery: patient falls.
The AI algorithms developed by partners at Qventus analyze patient records to predict which patients might experience falls. The algorithm also employs datasets from nurse call lights and bed alarms. The end result is a risk assessment, which nurses use to flag patients who may experience a fall. In the first year, the hospital reduced falls by 39%.
Hospital Administration
Doctors depend on the accuracy and efficiency of a hospital’s administrative staff, and this key player in the healthcare industry may also get a boost from AI technology. The first major venture comes from a joint partnership between General Electric and Partners Healthcare, a network that includes Brigham and Women’s Hospital and Massachusetts General.
GE and Partners are looking to integrate AI into clinical workflows across a broad range of medical fields. “Clinicians are inundated with data, and the patient experience suffers from inefficiencies in the healthcare industry,” says David Torchiana, CEO of Partners HealthCare. The ten-year collaboration will start by employing AI in diagnostic imaging, with plans to expand applications in every aspect of patient care, from admittance to discharge.
Humans and AI Working Together in Healthcare
Despite the incredible promise of AI, the human element will always be an essential element of effective healthcare. With AI, doctors and clinicians will spend more time focusing on their roles as caregivers and less time sifting through patient data and performing administrative tasks. Tomorrow’s healthcare industry is human focused and AI driven, and patients are already seeing the long-term benefits.
