This week’s brand new episode, “The one with the sixpack”, was the first episode broadcasted live from London, England, and starred Prof. Andrea Rockall, Dr. Shah Islam and Fahdi Kanawati from the Imperial College London.
Prof. Andrea Rockall started the show by presenting facts on sarcopenia and the landmarks that have to be considered when assessing it, followed by two different patient cases and the current clinical workflow involved when diagnosing sarcopenia, and addressed the unmet clinical need in detail. Prof. Rockall then took the audience through the workflow requirements when creating such an algorithm and how the database was developed, and afterwards explained the “boring bit”, as she called it: the data annotation.
It was good to have a technical expert at the studio who described in which way radiologists and AI specialists have to cooperate in order to create efficient tools and how this cooperation is crucial when developing the machine learning tools tailored to clinical needs. They provided the audience with some useful data sources, such as the Cancer Imaging Archive. As mentioned in previous episodes, the more diverse the data, the more accurate the tool.
In one of the clips shown during the live broadcast, Prof. Rockall explained, step by step, how this specific machine learning tool was developed by her team, from training the model to the statistical validation; Prof. Georg Langs joined her to explain the technical side of the process. Once again, they prove that AI really is a team effort.
The third instalment of Theo’s Technical Top Tips focused on data curation. Theo Barfoot, an MRI physicist at the Royal Marsden London, presents one valuable tip during each episode.
During this week’s episode, there was a multiple-choice live poll and the audience, through the chat feature, participated and also asked questions throughout the broadcast. The incredible hosts took extra time at the end of the show to answer and discuss these questions from the viewers.
One fact was made very clear this week: Radiologists don’t have to be afraid of being replaced by machine learning tools and AI solutions. The role of the radiologist in developing and working with these tools is, and will always be, crucial.
This season, Reasons to do AI with Friends will be broadcasted live every Wednesday at 19:30 CET until February 19 and present use cases for various clinical issues. The next episode, “The one that might be renal cancer”, will be hosted by Andrea Rockall, Wenija Bai and Say Vivekanantham from the Imperial College London on February 5 at 19:30 CET.
You can register for free and all episodes will be available on-demand for purchase on connect.myesr.org.
Read more on Episode 1, “The one where we classify breast cancer with MRI” here.
Read more on Episode 2, “The one where we tackle fibrosis” here.
Read more on Episode 4, “The one that might be renal cancer” here.
Read more on Episode 5, “The one where we find the connection” here.
Read more on Episode 6, “The one with whole body MRI” here.
Want to deepen your AI knowledge? Check out the reading recommendations from our experts here.
