How Artificial Intelligence and Software can defeat COVID-19
June 11, 2021
How big data impacts the finance and banking industries
June 12, 2021
One-third of hospitals and imaging centers report using artificial intelligence, machine learning, or deep learning to aid tasks associated with patient care imaging or business operations, according to a survey from Definitive Healthcare.
“Radiologists, in particular, use AI technologies to sift through millions of images and screen for potential abnormalities and patterns – thereby saving time (and potential lives),” said researchers.
“By using these technologies to screen images and identify abnormalities in patient scans, radiologists can significantly reduce the risk for both misdiagnosis and oversight in their analysis. With AI algorithms detecting abnormalities quicker than the human eye, AI has the ability to drive more efficient workflows for radiologists.”
Imaging centers reported a slightly higher use rate of AI technology compared to hospitals, with 34.7 percent of imaging centers saying that they use these tools and 31.9 percent of hospitals saying the same.
Artificial intelligence has the potential to transform medical imaging, saving providers time and effort when diagnosing and treating patients.
“This difference in reported usage is most likely due to distinctions in the primary objectives of each facility type. Imaging centers, for instance, are heavily based in diagnostic imaging services and, therefore, may show greater willingness to explore AI technologies that have shown promise in image analysis and operational tasks,” researchers said.
Nearly one-third of responding organizations not currently utilizing AI said that they will be using the technology within the next two years. Hospitals reported fewer plans for implementation than imaging centers, with 34.0 percent of imaging centers planning for implementation compared to just 28.3 percent of hospitals.
When combining current use rates with planned implementation numbers, 53 percent of responding organizations say they will be using AI within the next two years to either perform or assist in patient imaging or business operations.
Most participants who currently use AI are leveraging the technology for computer-aided detection of disease states, with 92.6 percent of organizations citing this as their area of use. The second most common use area was process or workflow improvement, at 26.5 percent, with no other area surpassing a usage rate of 16 percent.
In comparison, those who said they plan to use AI indicate that they will employ the technology more frequently across their organization, with 54.8 percent reporting they will use AI to improve processes and workflows and 50.0 percent planning to leverage AI for image detection of fractures and musculoskeletal injuries. Overall, current AI adopters averaged 1.79 out of a possible 6.00 areas of use, while those planning to adopt AI averaged 2.57 areas of use.
With the adoption of AI technology, imaging centers and hospitals are confident that they can improve care delivery as well as business processes.
“The implementation and use of AI is seen as highly valuable towards enhancing both patient care and business operations, although respondents feel there is a greater upside towards enhancing patient care across all data segments,” researchers stated.
More than half of respondents are confident that AI will have the greatest impact on patient care by improving or assisting in the accuracy of diagnosis. Seventeen percent of participants also said that the technology could offer operational improvements, while 15 percent said AI could speed up diagnosis time and enable earlier detection of disease.
The results show that while AI use is not yet ubiquitous in medical imaging, the development of new skills and tools will help facilitate adoption in the future.
“Artificial Intelligence Study shows only a limited existing establishment of AI technologies across imaging centers, the widespread interest in these technologies will likely result in a rapid growth in AI implementation across all facility types—particularly as these technologies improve and become more widely-accessible in the next few years,” researchers concluded.
