Artificial Intelligence and Health Care: What Premeds Should Know
With the advent of electronic medical records and wearable technology, data is becoming exponentially abundant in health care. Artificial intelligence has the potential to reshape medicine.
For example, AI can be used to aid doctors with their clinical decisions, find new tumor types in large research data sets, enhance the accuracy of diagnostic tests and improve hospital operations. The applications of AI are endless.
Premedical students can get involved in the intersection of AI and medicine by learning specific skills and working on research projects.
Advancing Health Care Through AI: Innovative Applications
Dr. Justin Norden was interested in medicine upon entering Carleton College in Minnesota, but he was also interested in computer science. What began as one computer science class to learn the basics of programming blossomed into an intellectual passion, and Norden majored in computer science along with fulfilling his premed requirements.
“My computer science courses led me to better understand programming, data structures and algorithm design,” Norden says. “All of this helped me in developing machine learning and AI applications in the future.”
Norden wanted to gain a deeper understanding of how his computer science degree could be applied to genomics and other health data applications, and he pursued a master’s in philosophy degree in computational biology at the University of Cambridge before entering the Stanford University School of Medicine.
Norden first applied his AI skills to analyze a large database of RNA sequencing data to understand the risks for developing colon cancer. He used machine learning techniques to create biomarker gene signatures in order to assess cancer risk.
At Stanford, Norden was curious about how to apply AI to digital health. He wanted to understand whether the data from wearable technologies could be leveraged to identify movement patterns in individuals and give clinicians a more precise understanding of a patient’s disease. In one research study, he looked through accelerometer data in 4,000 patients and found several features that could differentiate among normal individuals, individuals with spinal stenosis and individuals with knee osteoarthritis.
Finally, Norden desired to see how new technologies were affecting health care delivery, so he joined the Stanford Center for Digital Health. He evaluated new technologies and studied the impact of digital health solutions. For example, one study assessed whether there were differences between in-person visits and telemedicine visits in how doctors ordered prescriptions, lab tests, procedures and images.
Building Skills, AI Project Experiences as a Premed
Norden advises premed students to first learn skills that are important in AI.
“I would highly recommend learning computer science. The more technical skills you develop before starting your clinical training,” he says, “the more you will be able to bridge the fields of artificial intelligence, technology and medicine together.”
Math, data analytics and clinical medicine knowledge will be helpful, Norden adds. While he formally pursued a computer science major in college and a master’s degree in computational biology, he encourages premeds to learn through informal arenas.
“There are a lot of great, free online classes that will teach you the necessary skills online,” Norden says. “Interested premeds have more access now than ever before to learn AI.”
Once premeds gain coding skills, Norden highly recommends a next step: “My biggest advice for premeds is to find a project and a mentor – get involved in an AI project because there is so much opportunity and a great need in health care for these groundbreaking discoveries.”
Here are some avenues for premed students interested in AI projects:
One of the most common applications of AI in medicine is through precision medicine. AI can be used to understand patterns in patient attributes and then recommend a treatment plan based on AI analysis.
For example, a doctor can look at a cystic fibrosis patient’s specific genotype to help guide a personalized treatment regimen.
Patterns in Large Clinical Data Sets
AI is commonly applied to large clinical data sets to help researchers analyze patterns in data. For example, AI techniques have been used to understand gene expression data and tumor markers in cancer. In fact, researchers have been able to discover new cancer subtypes through this AI method.
One increasingly popular area among start-up companies is applying AI in radiology to more accurately identify abnormalities in patient diagnostic tests. For example, AI is now able to diagnose diseases in chest X-rays, like early stages of pneumonia.
Clinical Decision Support
Doctors and administrators can help design AI algorithms to guide physicians in clinical decision-making. For example, certain patient risk factors can alert doctors that a patient is at risk for an infection. Then the physician can either monitor the patient more closely or prescribe a medication to prevent an infection.
Hospitals are only beginning to discover how AI can be applied to hospital operations to increase patient follow-up appointments, optimize appointment slots and improve billing. For example, AI can be used to identify patients who need follow-up appointments and then send them reminders to set up one.
Another example is that hospitals can develop AI tools to calculate optimal operating room efficiency. A newer, innovative application of AI in hospitals is emerging through smart sensors, which can detect when a patient falls and alert health care providers immediately.
Natural Language Processing
Speech recognition is one nonmedical AI application that is being improved over time, and it is helpful in medicine as well. Doctors use speech recognition software to dictate their notes, which improves physician charting efficiency.
Natural language processing is also powering medical chatbots that interact with patients to provide immediate answers and a first layer of medical support.
AI is revolutionizing health care, and premeds can explore exciting ways to advance medicine and scientific research in this interdisciplinary field.