Artificial intelligence can reduce costs, prompt greater medical advancement, and improve patient outcomes, but is not without risks. Integrating artificial intelligence (AI) into healthcare systems links the need for physician-inventors to protect their innovations, and for healthcare practitioners to minimize risk where unintended consequences of AI decision-making can incur significant liability.
Most healthcare-related AI patents focus on physician assistance and support rather than replacement, including diagnosis, treatment, prediction, record management, and surgical assistance. AI-based systems enhance decision-support capabilities by providing quick, accurate diagnoses and eliminating costly repetitive or unnecessary diagnostic testing. AI-based medical image diagnostic systems have improved and expedited readings of CT scans and x-rays. Stanford University researchers created an algorithm for rapid skin cancer diagnoses, leading to earlier treatment and improved patient outcomes.
Artificial intelligence has a growing role in the personalized medicine space as large-scale aggregation of patient data enables more accurate diagnoses and more effective treatments. Automated AI systems can analyze past medical history to predict the likelihood of developing a disease. AI-managed data analytics can identify classes of individuals more prone to a disease or more likely to respond to a treatment. Nevertheless, despite superior accuracy and decreased margin of error, physicians must remain alert. Liability will likely remain with the physician, suggesting that best practice will be to confirm test results and report errors to product developers.
Patent protection incentivizes AI innovations by permitting inventors to exclude others from making, using, selling, offering to sell, or importing the protected invention into the United States for a limited time. Many companies already utilize patents to safeguard their healthcare innovations. GE Healthcare has filed applications covering computer-assisted image processing diagnostic systems, most notably in the field of oncology. Siemens has protected computerized systems and automated methods that distinguish between benign and malignant breast cancer lesions. Phillips has safeguarded diagnostic and treatment systems for cancer and various other neurological disorders, such as Alzheimer's disease. Companies that traditionally provide services in the field of technology, including Samsung Electronics, IBM Corporation, and Google, seek to capitalize on this growing market and become major players in this area for the foreseeable future.
Companies specializing in artificial intelligence for healthcare have already begun building patent portfolios around their products. One such company, IDx developed IDx-DR, an AI-based diagnostic system for the autonomous detection of diabetic retinopathy, a leading cause of blindness. This device will help over 30 million Americans suffering diabetes, half of which forgo annual eye examinations. Just last month, the FDA approved IDx to market the IDx-DR, marking a historic moment for our changing healthcare system. IDx-DR is the first autonomous, AI-based diagnostic system to be granted such approval. It screens patients for diabetic retinopathy using deep learning algorithms. U.S. Patent No. 9,924,867 recently issued to IDX with claims directed to both diagnostic systems and imaging methods used therein.
AI benefits physicians, allowing them to devote more resources to individual patient needs. With the greater precision and accuracy, medical malpractice laws should become a less viable means of policing diagnosis and treatment decisions, increasing physician confidence and decreasing the exposure of physicians to lawsuits. Physicians will no longer need to run batteries of tests to avoid misdiagnoses and shield themselves from liability. The goal is defensive medicine will disappear.
Artificial intelligence also decreases healthcare costs. A recent report by Accenture indicates that clinical healthcare AI-based applications can create $150 billion in annual savings by 2026. This is significant, as healthcare is the fifth largest industry by GDP share in the United States with a GDP of over $1.2 trillion. AI-based applications could save money in such areas as robot-assisted surgery, fraud detection, and dosage error reduction. Greater accuracy and precision will in turn pass on the savings to the physician in the form of decreased malpractice insurance.
As AI gains traction, its efficacy for advancing the current system must be proven. Inventors and healthcare professionals will play critical roles in transitioning AI from bench to bedside. Inventors will need to be vigilant in protecting the systems they have developed. Doctors and nurses must be prepared to validate system accuracy to obtain a true baseline of accuracy. Doctors and nurses will need to keep up with the ever-changing legal landscape to evaluate potential liability.
Outside the healthcare context, autonomous vehicles exemplify these concerns. Uber recently ceased automated vehicle testing in Arizona when one of its cars killed a pedestrian, despite the presence of a human safety driver behind the wheel. Analogous risks exist in the healthcare industry. For example, diagnosis failures or AI-based surgical robot malfunctions could lead to increased morbidity or mortality.
The what-ifs implicated by use of AI to treat patients raise a myriad of concerns for legislators, regulatory bodies, practitioners, and patients. Who should ultimately shoulder liability if an AI system fails to diagnose a condition or disease or misdiagnoses a patient? What implications does artificial intelligence have on the role of members of the physician's supporting staff? Should a line be drawn in terms of risk?
Government agencies and industry organizations should assist in formulating approaches toward mitigating risk as the presence of artificial intelligence in healthcare increases. Diligent lobbying efforts, successful case studies, and meaningful statistical analyses will determine when the benefits outweigh the risks. The healthcare community must address the legal, regulatory, and ethical considerations that accompany these technological advances.
Registered patent attorney Alex Huffstutter practices at Patterson Intellectual Property Law primarily in the areas of electronics and the mechanical arts, with particular expertise in recent technical innovations in electrical engineering including graphene synthesis. Zachary Gureasko is a civil litigator who predominantly defends claims on behalf of various individuals and entities in the areas of: general civil liability, personal injury, premises liability, products liability, and medical malpractice. For more information, go online to iplawgroup.com.
Lagasse, J. (2018). Where AI has the most promise for reducing healthcare costs. [online] Healthcare Finance News. Available at: http://www.healthcarefinancenews.com/news/where-ai-has-most-promise-reducing-healthcare-costs [Accessed 15 May 2018].
Medium. (2018). The Impact of Artificial Intelligence in Healthcare. [online] Available at: https://medium.com/@Unfoldlabs/the-impact-of-artificial-intelligence-in-healthcare-4bc657f129f5 [Accessed 15 May 2018].
Prnewswire.com. (2018). FDA Permits Marketing of IDx-DR for Automated Detection of Diabetic Retinopathy in Primary Care. [online] Available at: https://www.prnewswire.com/news-releases/fda-permits-marketing-of-idx-dr-for-automated-detection-of-diabetic-retinopathy-in-primary-care-300628626.html [Accessed 15 May 2018].
Accenture. Artificial Intelligence: Healthcare's New Nervous system. Available at: https://www.accenture.com/t20171215T032059Z__w__/us-en/_acnmedia/PDF-49/Accenture-Health-Artificial-Intelligence.pdf [Accessed May 15, 2018].