In recent years, artificial intelligence has been one of the most talked about and popular technology trends in...
healthcare. There have been many success stories around the use of AI for medical imaging, data mining and implementation of bots that have conversations with patients. AI also offers hospitals an opportunity to enhance quality of care and improve patient outcomes. However, there are several obstacles standing in the way of AI adoption in healthcare.
Many key drivers behind the use of AI applications in healthcare stem from the need for hospitals to improve patient outcomes and deliver more effective treatments and care. Mayo Clinic was one of the early adopters of IBM's flagship AI platform Watson and implemented it as part of its clinical team to help match patients to the appropriate cancer treatments and trials. Their implementation proved that AI could provide valuable insights into accurately matching patients with specific needs to treatment plans that match their conditions. Another use was at the Cleveland Clinic, which adopted Microsoft AI technology to help determine high-risk patients in their ICU. The process included the use of machine learning and advanced analytics to determine, based on different indicators, who needed attention and would be considered high risk.
Despite these successful use cases for AI applications in healthcare, several hospital IT executives find it challenging to implement this technology in their current environment. The difficulties hospitals face when implementing AI are the result of a few challenges that healthcare as a whole is dealing with. Here are six common barriers to AI adoption in healthcare.
An incomplete digital platform
It may be hard to believe, but the use of paper and faxes is still alive and well in some hospitals. This makes it challenging for organizations to claim that all patient data is available for mining in a discrete format and AI can tap into all of it. Despite the use of optical character recognition, which can convert scanned paper-based documents and extract text from them, AI may still not have access to all relevant patient data since not all of it is in a digital and readable format.
Lack of interoperability is a huge limiting factor for AI
To date, several healthcare technology experts continue to highlight that interoperability plays a significant role in supporting data sharing. Without access to a patient's comprehensive data, AI will not be able to offer its full benefits to healthcare. With primary care physicians, specialists and hospitals working with different EHR platforms, it is very challenging for any one entity to be able to access a patient's full record. This limits the information that AI can see and causes incomplete analysis of the medical record.
Limited use of AI applications in healthcare
Advanced image processing and predictive analytics tend to be the most popular implementations of AI applications in healthcare. However, there is certainly a lot more that AI can offer healthcare organizations. Natural language processing, interactive bots, robots and machine learning are just a few examples that only a limited number of hospitals have engaged in.
Shortage of AI talent causes more delays in adoption
With the increasing demand for AI adoption across all industries and limited talent pool, healthcare organizations are finding it very difficult to initiate AI-based projects due to the lack of resources. Many of them rely on third-party vendors and costly solutions. Unfortunately, this also means that hospitals are likely to limit internal experimentation and innovation because of the talent shortage.
Lack of cloud adoption slows down AI use
When considering some of the AI services available to date for organizations, many of the ones that come to mind are likely to be hosted and offered in the cloud. Vendors like Amazon, IBM, Google and Microsoft have been known to offer several different AI solutions through their cloud service. Unfortunately, some healthcare organizations are still hesitant to move data to the cloud. This results in some organizations abandoning the use of cloud-based AI applications in healthcare and resorting to on-premises solutions that may have limited capabilities and potentially more complexity due to the IT environment requirements.
Limited knowledge and understanding of AI
AI means different things to different people; some see it as the software that powers the robot that cruises the hospital hallways delivering different supplies to nurses, others consider it the platform that can perform deep analysis of large data sets to detect abnormalities in patient records. The fact remains that AI can be used in many ways in a healthcare setting, but because of this limited understanding of what it can and can't do, there is little buy-in from some stakeholders in hospitals that will hinder artificial intelligence adoption.
With AI already introduced in some areas of healthcare and successfully supporting patient care, hospitals are beginning to recognize the need to further invest in the technology to improve care, costs and quality. But to take advantage of its full potential, healthcare organizations must be able to overcome the highlighted obstacles above and help facilitate the adoption of AI applications in healthcare. Those who embrace the use of AI will certainly set themselves apart in the marketplace with a competitive edge that will differentiate them from the rest.
Current best use cases for AI are small and targeted