Artificial intelligence is increasingly becoming a part of our lives. This is by helping us automatically tag all cat pictures on our phone, analyze our sleeping patterns, or drive a car to work without our hands on the wheel. Within AI, an eminent area for application is radiology, where image analysis methods developed for completely different purposes can be applied to medical images to search for information of vital importance.
However, many radiologists are following these developments with a healthy dose of suspicion, as it is still unclear as to how and when AI will change their jobs. Without a doubt, however, AI will change the day to day tasks of the radiologist, therefore, they should not stand on the sidelines while this transition is taking place. Below, we discuss 3 arguments for why radiologists should adopt artificial intelligence today.
1. Shape the future of AI radiology
Deep neural networks classifying medical images: a few years ago, they hardly existed, but now, it is one of the major trends driving the rise of artificial intelligence in radiology. As the developments determining how AI will influence radiology in the future are happening as we speak, radiologists have every chance to influence this transition. Getting involved will not only allow radiologists to better understand what AI exactly is and what it can do, but it will also allow them to have a say in how it can be put into practice. For instance, how would a physician like to have AI software integrated into their workflows? What are the cases that radiologists would like to spend less time on? By leveraging constructive feedback from radiologists, software developers can ensure the final product seamlessly covers the needs of the physician and will truly be of value in the diagnostic chain.
Shape the future generation with AI radiology
Additionally, offering a working environment with state of the art tools will contribute to attracting top talent. The new generation of radiologists is already showing increasing curiosity for high tech radiology tools. Adopting artificial intelligence as a support tool in the diagnostic chain will give hospitals a headstart with recruiting the new radiology talent.
2. Make AI radiology fit into your budget
The healthcare information generated daily is expected to grow to 25,000 petabytes by 2020, 50 times the amount it was in 2012.1 Taking the form of physician notes, medical images, and genetic tests, manually analyzing this data would require an unimaginable increase in the amount of radiologists. Considering that the burnout rate amongst radiologists is already the highest in the US2, it is clear that automated solutions need to be made available.
But who is going to pay for the transition to AI? Will governments cover the bill? Or will hospitals have to allocate budget? Will the bill be eventually presented to the patient? Or, the scenario we are aiming at, will AI pay for itself? It is unlikely that AI radiology software will, in its current state, be able to fully support itself, mostly due to the upfront investment that needs to be made. Kurt Schoppe recently published an article in the Journal of the American College of Radiology discussing the question “Artificial Intelligence: Who pays and how?”3. He cautiously concludes: “But without the promise of governmental largesse or large inflows of reimbursements from private payers, vendors may take a pass on investing resources in radiology or health care-specific applications for AI.” In other words, it is unlikely anyone else but hospitals and clinics are going to pick up the bill in the short term. Wouldn’t it be interesting for hospitals to cut costs by getting early bird special prices for AI software?
Put your medical data to use in AI radiology
Medical institutions willing to get into a more experimental setup can try their luck on becoming a test site for vendors. Many AI radiology software development companies are eager to validate their product in the clinical workflow. Test locations at hospitals and physician feedback can offer valuable input to the software development process. Another reason why collaborations between clinics and software development companies may prove fruitful is the large amounts of medical data that hospitals have access to. Developing high quality AI based algorithms requires extremely large datasets. As a result, medical software companies are usually eager to collaborate with clinical care providers if new datasets are made available.
3. Unlock the potential of AI radiology and become the best radiologist you can be
Likely the most important reason of all: AI radiology has the ability to really support the radiologist. Either by taking over routine tasks, enabling a more differentiated diagnosis, accelerating the process,4 or increasing diagnostic accuracy. AI has every potential to become the greatest asset available to the modern radiologist.
Eager to join the transition of AI in radiology? Read more about our products or get in touch!
Bibliography
- Piai, S. Bigger Data for Better Healthcare. (2013). Available at: https://www.intel.com/content/dam/www/public/ us/en/documents/white-papers/bigger-data-better-healthcare-idc-insights-white-paper.pdf. (Accessed: 6th September 2108)
- Shanafelt, T. D. et al. Changes in Burnout and Satisfaction With Work-Life Balance in Physicians and the General US Working Population Between 2011 and 2014. Mayo Clin. Proc. 90, 1600–1613 (2015).
- Schoppe, K. Artificial Intelligence: Who Pays and How? J. Am. Coll. Radiol. 15, 1240–1242 (2018).
- Ridley, E. L. AI algorithm can triage head CT exams for urgent review. (2018). Available at: https://www.auntminnie.com/index.aspx?sec=sup&sub=aic&pag=dis&ItemID=120662. (Accessed: 6th September 2018)