The COVID-19 pandemic is running its apocalyptic course around the world. One thing is certain in the midst of total uncertainty and utter chaos: SARS-CoV-2 is a very daunting enemy and we remain totally subjugated by these viral overlords. It is easy, therefore, for we humans to lose patience and become careless as well as to decrease capacity for resilience. We need to regain our composure and to be defiant in order to for us to prevail this human vs virus struggle.
A major part of this resilience and defiance is continue our work in adoption and deployment of artificial intelligence concepts and projects in clinical medicine. Here are key takeaways from today’s AIMed Radiology webinar in collaboration with the American College of Radiology.
1. Artificial intelligence in radiology has matured with a Cambrian explosion of convolutional neural networks and now coupled with recurrent neural networks for moving images.
2. The value proposition for an AI tool in radiology now has a precedent with more discussion more in the realm of a service embedded in imaging (vs separate service).
3. While the IT infrastructure issues are daunting (lack of architecture), AI in radiology will improve accuracy and outcome so AI can serve as a North Star to overcome these challenges.
4. Model performance drift occurs with change in disease and population so a more iterative and flexible model is helpful.
5. For us to reach higher levels of AI, we should have a long term vision for the large dividends and not be distracted by the hype and potential failure to deliver short term dividends.
6. We need to balance the accuracy performance of deep learning models in medical imaging with the eventual outcome of these patients as the perfect model can have little impact.
7. Multimodal medical fusion for imaging and words (EHR) can increase the accuracy of the medical image interpretation.
8. Sharing of existing data and generating synthetic data are useful strategies to increase the volume of data for the deep learning tools.
9. Natural language processing is an essential tool to augment performance of medical image interpretation and to improve workflow of image acquisition.
10. The capabilities of artificial intelligence can extend much beyond medical image interpretation and into workflow issues and clinical practice of radiology.
Thank you faculty and attendees for your knowledge and expertise as we all learned a great deal today at AIMed Radiology!