- Tomorrow's Dose
- Posts
- Edition 4 - Emerging Radiotherapy Techniques & AI Advances: A New Era in Cancer Treatment
Edition 4 - Emerging Radiotherapy Techniques & AI Advances: A New Era in Cancer Treatment
Explore the first clinical implementation of Minibeam Radiation Therapy, AI's role as a secondary reader in mammography, and innovative prediction models for brain metastasis treatment response.
First clinical implementation of Minibeam Radiation Therapy (MBRT).
AI as a secondary reader in mammography.
Prediction of treatment response after stereotactic radiosurgery of brain metastasis.
Featured follow of the week
Top posts of the week across social
Meet the editor
Want a featured article?
Specialty: Radiotherapy // Sub-Specialty: Techniques // Body Site: Skin
1. First clinical implementation of Minibeam Radiation Therapy (MBRT):
Minibeam radiation therapy (MBRT) is characterized by the delivery of submillimeter-wide regions of high “peak” and low “valley” doses throughout a tumor. A clinical orthovoltage unit was commissioned for MBRT, with the 180 kVp output spatially separated into minibeams using a tungsten collimator with 0.5 mm wide slits. Two patients were treated, each receiving 2 fractions, and both experienced prompt improvement in symptoms and tumour response.
Read Full Article
Paul’s Thoughts:
In spatially fractionated radiation therapy a grid is applied above the target area with the intention of delivering of non-uniform dose distribution to the tumour, instead of delivering a uniform dose distribution to the tumor. Preclinical studies have long shown the promise of this technique, but this report of the first clinical implementation of MBRT is very interesting. It remains to be seen whether the benefits of the tumour response justify the efforts required to deliver this technique - patient-specific 3D-printed collimator holders are required to ensure conformality to the unique anatomy of each patient and are affixed directly to the body.
Timescale: Mid | 5 Years
Specialty: Radiology // Sub-Specialty: Diagnosis // Body Site: Breast
2. AI as a secondary reader in mammography:
AI provides an option as a secondary reader in mammography. Following the publication of Gommers in late 2023 that discussed the optimisation of pairs of radiologists when reading mammography, a letter was sent from Dr Batan, making some interesting suggestions regarding the use of AI. There is a suggestion to investigate which group performs best between randomly paired radiologists, optimised pairs, and human-with-AI pairs. Suggestions are also made to the potential of eye tracking technology and the use of more personalised metrics beyond cancer detection rate and abnormal interpretation rate.
Read Full Article
Paul’s Thoughts:
It is clear there are many unanswered questions in the AI and radiology space right now, and it is important that major journals continue to publish letters such as these so that the topics can be openly discussed. The potential for eye-tracking technology to be used to determine how AI influences the reading by the radiologist is an intriguing one.
Timescale: Early | 2 Years
Specialty: Radiotherapy // Sub-Specialty: Prognosis // Body Site: Brain
3. Prediction of treatment response after stereotactic radiosurgery of brain metastasis:
This study showed prediction of treatment response after stereotactic radiosurgery of brain metastasis using deep learning and radiomics on longitudinal MRI data. Maximum axial diameter (Dmax), radiomics, and deep learning (DL) models were generated for comparison. 194 patients with BM and 369 target lesions were used for for training and testing, and a further 172 MRI scans from 43 patients with BM and 62 target lesions were used for external validation. It was found that a 2D gated recurrent unit (Conv-GRU)-based CNN model outperformed the 3D Conv-GRU, Dmax, and radiomics models.
Read Full Article
Paul’s Thoughts:
Further application of CNN models to treatment prognosis in radiotherapy add weight to its use in in the clinic. It will be interesting to see if new insights can be garnered from the deep learning models.
Timescale: Early | 3 Years
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