❓ Is Radiation Oncology ready for AI?
Take our short, anonymous AI knowledge quiz to find out!
✨ Just 10–15 min
⚡ Instant feedback
🎓 Help guide future training
🚀 Start here: airadoncquiz.streamlit.app
🙏 Please share with colleagues!
🎉 A record year for #SLRON at #ESTRO25! 25 accepted abstracts, including 1 talk, covering a range of topics including AI in #RadOnc, clinical trials, re-irradiation & open-face masks in H&N RT. A true #MDT effort! 🚀
Thanks to #SLICR & @friendsofstluke for their support!
EU opinion piece on data processing for #AI in #MedPhys & #RadOnc:
✅ Criteria for AI models to be considered anonymous
✅ Use of 'legitimate interest' to justify data use
✅ Implications of using unlawfully obtained personal data in model development
#GDPR ++
www.edpb.europa.eu/system/files...
A step closer to true human-AI synergy: This study shows how integrating human feedback into AI planning tools enhances treatment planning, reducing hotspots while preserving clinical quality. Further integration of AI as a complement, not a replacement. #AI #MedPhys #Radonc
shorturl.at/5DHlT
Acdc/Irish trad mix...wow! #latelatetoyshow
Loving how @bsky.app is the new place for positive #latelatetoyshow Christmas vibes.
Promising results from Lauren Conway on 5-point open face masks for #H&N patients using #SGRT:
- Fewer patients needing anti-anxiety meds
- Highlights the importance of the #SLRON OPEN trial which evaluates both 3- and 5-point masks
Towards greater comfort and more options for our patients!
Interesting talk from Michael Tallhammer on the use of Cherenkov imaging during RT delivery to provide real-time visualization of dose delivery. Nice mix of the physics and the clinical!
#SGRT #MedPhys #RadOnc
Looking forward to attending the SGRT Annual European Meeting today in London.
Particularly interested in:
- How others use #SGRT for open #H&N RT facemasks
- How others use SGRT to measure intra- inter-fraction motion
- Different ROI approaches/strategies
#medphys #Radonc #VisionRT #SLRON
Some interesting work from Glocker et al. demonstrating the need for open weight models after investigating the risk of #Bias on closed weight models:
doi.org/10.1148/ryai...
Exciting news from #Google Research: they're releasing foundational health models as open-weight models! This gives others a strong base to train for their own needs, reducing #bias and improving performance on local datasets. Hope more follow their lead! #AI #Health
shorturl.at/zQaPQ
Done!
Fully automated #radiotherapy planning: progress meets skepticism. This study from the Physics #ESTRO workshop '23 shows #AI can handle prostate cancer planning with minimal edits, yet trust in automation remains low. A promising future,if confidence can catch up to capability.
tinyurl.com/mwpcmv3s
go.bsky.app/P1zYwRs
Please share with others in the Radiation Oncology/Cancer Research space in Ireland.
If you would like to be added to the radonc starter pack just let me know!
#radonc #radiotherapy #radiationoncology
go.bsky.app/2YKENEm
Article from ESTRO meets ASIA 24: Accumulated Dose in Clinical Practice – are we there yet?
Dose mapping in RT is advancing but still faces challenges like DIR uncertainties and limited tools. Collaboration and better communication are key to progress.
Full article here:
tinyurl.com/3erdux87
This approach seems both easier and more cost-effective to implement in a clinic (compared to a full 'AI' based planning approach), as it keeps the human in the loop to oversee the process.
Interesting paper I read today re: automated RT breast planning:
➪ ML-based system predicts ideal fluence maps.
➪ Automated plans match clinical quality, cuts planning time & allows for personalised fine-tuning.
➪ Supports planners reaching optimal planning solution.
tinyurl.com/Scaop
Education is key:
- Reviewing auto-contours is a different skill from manual delineation and requires additional training.
- Staff must maintain their manual delineation skills.
Be aware of biases:
- Automation bias can affect how clinicians/RTTs review auto-contours.
- Anchoring bias and review fatigue may lead to missed edge cases or poor system performance.
Technique and intent matter:
- Manual editing of auto-contours should only be needed if it impacts the treatment plan.
- Understanding clinical intent and treatment technique is crucial for evaluating acceptability.
When choosing a system:
- Involvement of the multidisciplinary team in evaluating and commissioning is essential.
- Commissioning should use local data representative of your clinical cohort.
- Collect common errors during commissioning for training purposes.
QA of autocontouring systems is not quite like traditional QA on a Linac:
- Interestingly, for ongoing QA day-to-day, only qualitative evaluation is recommended.
- However, broader Post-implementation monitoring is also recommended using Geometric, dosimetric, and qualitative metrics.
Auto-contouring systems should only be used as intended: staff must be aware that performance may vary across different tumour types, immobilisation equipment, and imaging modalities.
Some key takeaways from my initial read:
- Healthcare professionals approving auto-contours are ultimately responsible for their clinical use.
- Auto-contouring performance depends on how the system is trained and operates.
Some RadOnc news! 📢 The RCR has just published new guidance on auto-contouring systems for radiotherapy treatment. Essential reading for all #RadOnc #MedPhys and #RTT healthcare professionals. Check it out here: tinyurl.com/RCRAISeg #Radiotherapy #RCR #AI
Have you added @clairepoole.bsky.social :)
Thanks for sharing! The great thing about starter packs is that they are dynamic, so hopefully this list will grow and become a valuable resource!
@ciaralyons.bsky.social @gerryhanna.bsky.social
