Webinar invite: Democratize Adaptive Therapy & beyond with the latest in AI (Radiotherapy)
Radiotherapy planning and treatment, a standard for most cancers, has gone through tremendous change since the 1990's.
Back then single treatment plans for the complete duration of a patient's treatment cycle over multiple weeks with large target (i.e. cancerous tissue) margins to account for anatomical changes have been the standard.
Fast forward and today, through the remarkable advancements in imaging (simulation) and treatment delivery technology, millimeter but even sub-millimeter targeting accuracy is made possible.
Not only did these open up the possibility to decreasing target margins (in order to decrease healthy tissue toxicity), but the recent high-resolution imaging allows clinical experts to visualize anatomical changes in the shape and texture of internal tissues, as well as due weight loss and tumor shrinkage that were simply not possible with sufficient confidence before.
These innovations led to a vast amount of additional information that could be used to achieve more accurate and precise treatments but has also led to an increase in workload for medical physicists, radiation oncologists, dosimetrists and RTTs involved in steps such as organ-at-risk (OAR) and clinical target volume (CTV) contouring and treatment planning.
A few years ago with the advent of computational speed that was powerful enough to process AI algorithms quickly enough for a clinical setting, multiple steps in the treatment planning workflow can now be automated with comparable quality than before manually.
Thus, Artificial Intelligence (AI) powered (e.g. deep-learning) methods such as organ and target auto-contouring, synthetic imaging (e.g. generation of a diagnostic quality and field-of-view pseudo-CT from a low image quality cone beam CT), adaptive therapy automation and automated treatment planning (or auto-planning) are more and more becoming standards.
On one hand, this was with the goal to gain back the incremental daily clinical time spent due to the new complexity created through the decades of technology innovation, but also to allow personalized precision radiotherapy treatment with AI enabling efficient adaptive therapy.
Webinar invite
The next episode of the Human Bytes AI Academy of educational webinars will focus on how Artificial Intelligence can help democratize adaptive therapy as well as automate treatment planning and beyond.
In this webinar, you will learn from Prof. Nikos Paragios (CEO at TheraPanacea , Distinguished Professior of Applied Mathematics at Université Paris-Saclay ) about:
Zooming in on the art in ART (Adaptive Radiation Therapy)
A very recent publication by Olga Maria Dona Lemus et al. in Cancers (Basel) provides a wonderful and exhaustive overview of the various aspects, challenges and state-of-the-art of ART.
It highlights for example that "...OAR and target contouring still largely require careful inspections and frequent manual modifications by expert users...".
TheraPanacea's AI auto-contouring was not discussed and may provide a different perspective around this aspect as it was now the top pick related to auto-contouring quality and guideline compliance in three recent multi-vendor evaluations: German Oncology Center (Cyprus), Haukeland universitetssjukehus (Norway), V?stmanlands sjukhus V?ster?s (Sweden).
See here one of the many interesting sections of the publication comparing offline vs. online ART.
Can treatment planning be automated end-to-end?
By now, automation exists for various steps of the radiotherapy treatment planning workflow, to a lesser or higher degree. Results generated by automation are regarded with mixed trust.
领英推荐
Two good examples of automation are auto-contouring (of OARs and CTVs) and auto-planning. Latter is the generation of a treatment plan upon a treatment prescription with the aim to meet clinical constraints (as well as physical constraints of course, i.e. the plan needs to meet the capabilities of the linear accelerator otherwise it remains a theoretical plan) while saving significant time when compared to conventional dose calculation.
In 2023, a so-called grand challenge was setup for participants to perform automated treatment planning from scan to plan.
Participants were provided with radiotherapy CTs and a treatment prescription and were asked to produce a deliverable treatment plan using automation.
More than a dozen participated and TheraPanacea ranked 2nd (in a tie with another participant) adhering to 84.9% of the objectives when assessed with their own AI auto-contours, and 3rd with 84.7% for planning only.
The study concluded that "Automation of the full planning workflow from simulation CT to deliverable treatment plan is possible for prostate and prostate bed radiotherapy. While many generated plans were found to require none or minor adjustment to be regarded as clinically acceptable, the result indicated there is still a lack of trust in such systems preventing full automation."
Hear more about AI's benefits in adaptive therapy, auto-planning and beyond on October 23rd from Nikos Paragios.
References:
If you like to learn more or get a demo, please reach out to:
See you soon,
The Human Bytes Team