- Tomorrow's Dose
- Posts
- Edition 9 – AI in Radiotherapy: Automation, Device Surge, & Nobel Prize Win
Edition 9 – AI in Radiotherapy: Automation, Device Surge, & Nobel Prize Win
Explore the rise of automated radiotherapy, the surge in AI medical devices, and celebrate AI pioneers awarded the Nobel Prize in Physics.
Fully automated radiotherapy treatment planning: A scan to plan challenge
Number of AI medical devices spikes
Nobel prize in Physics given to AI researchers
Featured follow of the week
Top posts of the week across social
Meet the editor
Want a featured article?
Specialty: Radiotherapy // Sub-Specialty: AI // Body Site: All
1. Fully automated radiotherapy treatment planning: A scan to plan challenge
A challenge to perform automated treatment planning for prostate and prostate bed radiotherapy was set up as part of an ESTRO Physics Workshop in 2023. Participants were provided with simulation CTs and a treatment prescription and were asked to use automated tools to produce a deliverable radiotherapy treatment plan with as little human intervention as possible. Plans were scored for their adherence to the protocol when assessed using consensus expert contours. Thirteen entries were received. The top submission adhered to 81.8% of the minimum objectives across all cases using the consensus contour, meeting all objectives in one of the ten cases. The same system met 89.5% of objectives when assessed with their own auto-contours, meeting all objectives in four of the ten cases. The majority of systems used in the challenge had regulatory clearance (Auto-contouring: 82.5%, Auto-planning: 77%). Despite the ‘hard’ rule that participants should not check or edit contours or plans, 69% reported looking at their results before submission.
Read Full Article
Paul’s Thoughts:
The authors conclude that there seems to be a lack of trust in full automation, since the participants in the study made manual reviews of the resulting contours and treatment plans. This 'skepticism' seems perfectly normal and is analogous to reviewing a plan either by a clinician or an independent physicist/planner. As deep learning models evolve and improve, there will be fewer human 'corrections', but it will still be necessary for humans to review and develop new treatment concepts. One of the challenges highlighted in the paper is the definition of a 'good' plan, because objectives are often in competition. Often the selected plan will rely on other background information (previous irradiations, response to other treatments), which is unlikely to be of knowledge to the TPS. There is thus a need for the TPS to produce multiple plans, some prioritising taregt coverage, others focusing on OARs, to allow the clinician to select the most appropriate for the given case. The question for current TPS developers is to determine when is the best first point of contact for the human when selecting the plan?
Timescale: Acute | 1 Year
Specialty: All // Sub-Specialty: AI // Body Site: All
2. Number of AI medical devices spikes
The number of medical devices with artificial intelligence technology has risen sharply in the past decade. The Food and Drug Administration has authorized 950 AI or machine learning-enabled devices as of Aug. 7, 2024, according to the agency’s database. The FDA authorized the first AI-enabled device in 1995, in 2015 the FDA authorized six AI medical devices, while in 2023 the agency authorized 221 devices.
Read Full Article
Paul’s Thoughts:
The trend has been driven by more connected devices, more investment into AI and machine learning and growing familiarity with how software is regulated as a medical device, experts said in interviews. Additionally, we’re seeing huge increases in investment. Large medtech firms such as GE Healthcare, Siemens Healthineers and Medtronic are incorporating AI into equipment and building standalone software tools. And startups such as Aidoc, RapidAI and Butterfly Network are creating targeted solutions to help flag health conditions and improve ultrasound imaging. We are also seeing companies outside of the medical device space getting involved, with Apple developing features that use data from its watches to detect heart arrhythmias and Nvidia partnering with medical device companies including Medtronic and Johnson & Johnson to help build out their use of AI.
Timescale: Acute | 2 Years
Specialty: Radiology // Sub-Specialty: Radiographer // Body Site: All
3. Nobel prize in Physics given to AI researchers
The 2024 Nobel prize in physics has been jointly awarded to John Hopfield and Geoffery Hinton ‘for foundational discoveries and inventions that enable machine learning with artificial neural networks’. John Hopfield and Geoffrey Hinton’s work led to artificial neural networks that power AIs, such as the protein structure prediction program AlphaFold.
Read Full Article
Paul’s Thoughts:
When most of us think of AI, we think of chatbots like ChatGPT, of image generators like DALL-E, or of scientific applications like AlphaFold for predicting protein folding structures. Very few of us, however, think about physics as being at the core of artificial intelligence systems. But the notion of an artificial neural network first came as the result of physics studies across three disciplines — biophysics, statistical physics, and computational physics — all fused together. It’s because of this seminal work, undertaken largely in the 1980s, that the widespread uses of artificial intelligence and machine learning that permeate more and more of daily life are available to us today.
Timescale: Acute | 2 Years
I had a great few days in London last week representing German Oncology Center!
We had our first in-person meeting for the IPEM AI Group, where we discussed results from our survey and how they could shape our group's work for the future, among many other things. Thanks to Richard M. for his vision and leadership in this group, as well as the other members Sofia Michopoulou, Lydia Davidson, Mike Hutton, Jonathan Taylor, Nick Vennart, James Leighs, Virginia Marin and James McLaughlan.
In the shadows of Big Ben and the birthplace of the ICRU, the Institute of Physics and Engineering in Medicine (IPEM) #STEF2024 conference took place at Westminster Central Hall. I presented the IPEM AI group's survey on attitudes and perspectives towards AI in the Medical Physics and Clinical Engineering community, which served as a great conversation starter for the following days.
I was able to discuss:
- The potential of a platform for testing AI tools with Jaishree Naidoo.
- The potential for developing medical physics AI training with Nadia Smith and Rebecca Nutbrown at TÜV SÜD.
- How the results could shape National Physical Laboratory (NPL) research interests with Catharine Clark, Spencer Angus Thomas and Russell Thomas.
- How the results impact on industry attitudes in the panel discussion with Jaishree Naidoo, Stephen Chalkly-Pereira, Nathan Flannery and Wayne Martel, chaired by Anna Barnes, PhD and Richard M..
Lots of new connections and opportunities for collaboration make this an exciting few months ahead! And also a few more of you joining me here every week too!
A round-up of some of the best posts we found online this week.
Was this email forwarded to you?
Our weekly email brings you the latest health trends and insights, combining top news and opinions into a straightforward, digestible format.
 |  |  |
Want an article featured?
Have an insightful link or story about the future of medical health? Reach out below, and we may include it in a future release.
Reply