 |
Page History
...
| Contact | Details |
|---|---|
| WebEx | https://cbiit.webex.com/cbiit/j.php?MTID=mdb5f537bde0cff01e5c7779f02680185 |
| Meeting number (access code) | 732 377 553 |
| Meeting password | tSX9U9c? |
| Join by phone | 1-650-479-3207 Call-in toll number (US/Canada) |
Agenda of
...
December 5, 2022 Webinar
Item | Topic |
|---|---|
1 Bringing AI from Hype to Reality for Routine Clinical Practice: Defining and Addressing the Gaps Dr. Eliot Siegel, University of Maryland Despite the ever-increasing number of publicly available imaging databases and oncology AI/Radiomics applications that have been curated and developed over the past more than 15 years, an extraordinarily small number of AI applications are available and in use for routine clinical cancer care by radiologists, oncologists, and other healthcare providers. This is the case despite large and carefully and expertly curated and annotated databases which have been generously funded and made available by NCI and other organizations. Mammography CAD/AI has a particularly interesting and unique history and adoption curve and while it is in widespread use throughout the US, there continues to be a large gap in accuracy between the small percentage of studies interpreted by subspecialist mammographers and the vast majority of studies interpreted by general radiologists. This presentation will discuss some of the reasons for this continuing gap and lack of adoption of mammograph CAD into clinical decision making. | Report of the MIDI Task Group about best practices and recommendations for medical imaging de-identification (Dr. David Clunie) Additionally, a combination of regulatory challenges, the lack of a paradigm for training on datasets consisting of both prior and follow-up studies, brittleness of algorithms that are not adaptive, bias due in part to lack of transparency of databases used to develop AI apps, lack of standards for consumption of on prem and off prem algorithms, multiple platforms for packaging and using applications and lack of post-market surveillance, questions about whom the algorithms should be designed for, and many other factors have hampered widespread adoption. This presentation will discuss some solutions to these challenges that could accelerate adoption of these algorithms which could substantially enhance care for oncology patients. |
2 | IDC Update Ulrike Wagner, FNLCR |
3 | TCIA Update Justin Kirby, FNLCR |
Upcoming Webinars
| Date | Agenda | December 5, 2022 | |
|---|---|---|---|
| January 2, 2023 | Canceled, New Year's Day holiday | ||
| February 6, 2023 | Title to be announced (Dr. Michael McNitt-Gray) | ||
...
