Legendary surgeon, and my long time mentor and friend, Dr. William Norwood, sadly passed away in December. He was a preeminent force in innovation who embodied the unrelenting drive to break the status quo and unparalleled courage to develop new strategies. So as a special tribute to the great man, here are my predictions for 2021 in artificial intelligence in clinical medicine and healthcare:
Deep learning for deciphering protein folding will only be the beginning of AI making sizable impact in biomolecular science and drug discovery.
Embedded AI will be coupled with telehealth and monitoring devices to improve patient care not only during the next few months of the pandemic but beyond.
Natural language processing and its exponential rise in capabilities with GPT-3 will create a myriad of opportunities in healthcare-related communications and data analytics.
Robotic process automation will be utilized much more as a productive AI tool for many aspects of hospital administration operations and logistics.
Data sharing was ineffective especially during the early period of the COVID-19 pandemic and this will be focused upon as an area for improvement in healthcare.
Synthetic data generation will be more commonplace in biomedicine especially with rare diseases and uncommon conditions that are so common in biomedicine.
Convolutional neural network will be combined with recurrent neural network and other methodologies to render moving images a more robust domain for deep learning.
Explainable AI will be discussed with increasing frequency and progress will be made in making AI both more interpretable and explainable.
Automated workflow is more essential than ever before and this application of AI will start to make impact in subspecialties like pathology and cardiology after its maturation in radiology.
Cognitive architecture will be increasingly appreciated as a necessary element in current deep learning in AI in healthcare, especially in decision support areas.
