What Does The Future Hold For Gamma Knife Radiosurgery & AI?

There are many different ways in which brain tumours like glioblastoma and vascular malformations, but one of the most effective and advanced treatment plans is Gamma Knife radiosurgery, a practice that was first introduced back in the 1960s and which has been gathering pace ever since.

The last few years have seen artificial intelligence (AI) really start to become commonplace as a tool in a wide range of different industries - and healthcare is no different, with new and exciting treatment opportunities emerging as a result.

Some of these opportunities have just been featured in the spring edition of Scope magazine, with valuable contributions from the Queen Square clinical team, including Ian Paddick (consultant physicist at Queen Square) and Hannah Bouas (radiosurgery physicist at Thornbury Radiosurgery Centre).

The article focuses on AI-driven auto-contouring for tumour segmentation, detailing how the tool can achieve greater levels of precision, as well as more consistent patient outcomes in stereotactic radiosurgery.

Here are just some of the key conclusions in the feature.

AI-driven auto-contouring for advanced precision

Moving away from time-consuming manual processes that increase certain risks as a result of human interpretation is essential for the healthcare sector and AI-driven auto-contouring for tumour segmentation is a significant breakthrough in this regard, ensuring more accurate Gamma Knife procedures, particularly for lesions and tumours without clearly defined boundaries.

AI frameworks mean that clinicians will be able to segment tumour boundaries with the equivalent accuracy to human annotators, the advantages of which include:

Enhanced patient outcomes

Tumour segmentation can be done with greater accuracy with AI, enabling more precise radiation delivery and reduced damage to healthy brain tissue, as well as fewer complications and side-effects.

Reduced human error

We can achieve greater levels of consistency through AI automation, significantly reducing inaccuracies, variability and oversights.

Greater efficiencies

Use of automated tools enables more efficient contour detection and significantly shorter timescales, increasing treatment times and reducing waiting lists.

Possible drawbacks of AI

While the article does agree that AI is an increasingly valuable resource in healthcare settings, there are still some limitations for models for auto-contouring, including incomplete training data that may not generalise across diverse populations. Imaging noise can also potentially lead to false positives.

With all this in mind, use of AI will still require clinical oversight, while contouring processes will still require in-depth judgement and analysis, particularly for more complex cases such as brain metastases.