Back in 2017, I tweeted “radiologists who use AI will replace radiologists who don’t.” The tweet has taken on a life of its own, perhaps because it has a double meaning.
On the one hand, it suggests that luddite medical imaging professionals who refuse to adopt AI might be put out of work. Of course, this seems unlikely given that imaging professionals often have led the adoption of new technologies such as ultrasound, CT, MR, and PACS. The message also intends to promote the idea that all radiologists will eventually adopt AI, thereby replacing their former selves. But regardless of which replacement scenario seems more likely, no one doubts that AI will transform how imaging professionals practice medicine.
In response to these trends, and recognizing Stanford’s incredible pre-existing strengths in machine learning, computer vision, and medical imaging, the AIMI Center was established in April 2018. Thanks to the strong support and clear vision of Sam Gambhir, Chair of Radiology, Tom Montine, Chair of Pathology, and Bob Harrington, Chair of Medicine, and the rest of our Advisory Board, the center has undergone incredible growth and transformation over the past year. The start of new year is a great time to celebrate our accomplishments.
We can be most proud in 2019 of the seminal science published from the research laboratories of our over 120 affiliated faculty. Our work has appeared in many high-impact journals, including Nature, PLOS Medicine, Radiology, Magnetic Resonance in Medicine, JAMA Network Open, npj Scientific Reports, Artificial Intelligence in Medicine, Bioinformatics, Journal of the American Medical Informatics Association, and many more. Our work on chest radiograph interpretation, led by Pranav Rajpurkar, Jeremy Irvin, Matt Lungren and Andrew Ng, has been cited over 400 times since its publication just two years ago. In another example, Akshay Chaudhari, a new faculty active in the AIMI center, recently won the ISMRM W.S. Moore Young Investigator Award for his work on a comprehensive knee MRI study that can be acquired in 5 minutes.
We are also very proud of other areas where we have grown and led in 2020. The AIMI Center has led the world in curating and releasing large medical imaging datasets for global research and education. We have worked with our hospitals and the School of Medicine to support a number of key research tools, including the Nero high-performance computing cluster, the MD.ai image labeling application, the Montage radiology report search engine, and the STARR data repository and associated tools, all of which are currently available to our investigators at no cost.
Another exciting milestone was the initiation of the $525,000 AIMI seed grant program to help investigators conduct preliminary work to catalyze extramural funding. The program was a smashing success, with 45 applications and 7 funded projects addressing diverse clinical problems. It is gratifying to see the important work being accomplished by these diverse teams. We plan another round of funding in 2020.
Finally, our successes have led to growing influence in matching AIMI Center investigators with outside partners that support their work. We recently finalized a collaboration with GE Healthcare that provides sponsored research funding for the next 5 years. We are particularly proud of our other new corporate partners and affiliates, including Bayer, Google, Hospital Israelita Albert Einstein, Kheiron, IBM, IDEXX, and Nines.ai. And special thanks go out to the Paustenbach Fund, whose gift supports Saeed Seyyedi, a post-doctoral scholar working on early detection of breast cancer.
Meanwhile, our affiliated laboratories are having a real effect on the wider world. Two additional imaging/AI companies have spun out of Stanford recently, Bunkerhill Health, an open platform for clinical validation and deployment of AI algorithms for radiology, and Subtle Medical, which is using deep learning solutions to improve medical imaging efficiency and patient experience. We expect this trend to continue.
In closing, I want to personally thank our Executive Director, Johanna Kim, who has been instrumental in our progress over the past year, Stephanie Bogdan, who manages all AIMI Center projects and relationships, and Jacqueline Thomas, who is responsible for the smooth logistics for the center’s meetings and other activities. And special thanks are due to our Associate Director Matt Lungren and all the members of our Executive committee who provide tactical and strategic feedback to keep us on track and meeting the needs of our constituent faculty and laboratories.
The secret to our success is the productive interdisciplinary collaborations we forge across a diverse set of projects, labs, departments, schools, and industry partners. I thank each and every person who contributed to the important work and collegial atmosphere of the AIMI Center over the last year, developing innovative interdisciplinary AI methods to promote human health! I wish you all a happy and productive 2020.
- Curt Langlotz, MD, PhD