Cancer is a devastating illness that affects not only the individual who is diagnosed but also everyone around them. The National Cancer Institute suggests that more than 18 million new cases are diagnosed each year around the world.
Fortunately, treatments have continued to improve throughout the last decade. That includes the use of radiation.
Although radiotherapy is very effective, it is a long, tedious process for both the patient and the clinicians who set it up. Google and Mayo Clinic think that there is a better way. The duo is teaming up to research how artificial intelligence (AI) can be used to speed up the planning process for radiation therapy.
When it comes to saving lives, there is no such thing as spending too much time on a patient. However, more efficient care means that doctors are able to provide high-quality treatments to more patients while using the same amount of resources.
To understand why AI can be helpful, it’s important to know the challenges that radiation therapy poses. Since radiation damages both the cancerous cells and the surrounding tissue, treatments are narrowly focused on the tumor (or tumors) to spare as much healthy tissue as possible.
This process, called contouring, is currently done by hand at most facilities. Clinicians must first collect scans of the patient’s organs before drawing lines around them to guide the radiation treatments.
According to the Journal of the American College of Radiology, this process can take up to seven hours for a single patient. That’s nearly a full shift, meaning the number of new patients that can be processed in a timely manner is limited. When dealing with something like cancer, time is of the essence.
That’s where AI comes in. Google and Mayo Clinic are working together on an algorithm that can help separate healthy tissues and organs from tumors, thus making the contouring process much faster.
Work Being Done
For now, this research is in the early phases. Google researchers are working to validate an AI model and study how it can be deployed in the hospital setting. At this point, the technology isn’t being used with patients. The algorithms are undergoing training with unidentifiable data and scans.
However, real-world implementation could come fairly soon. This job is something that AI typically excels at.
In the past, Google researchers have used the technology to analyze a variety of medical scans. That includes one study conducted in collaboration with University College London Hospitals. The team used an AI system to segment head and neck cancer from scans in nearly the same way it could be used with radiotherapy.
Given the body of work to build on, the current research is focused on radiation for head and neck cancers. Contouring these areas typically takes longer since they house a number of delicate structures. As such, the benefits of AI-driven contouring would be immense.
That being said, once the team develops a reliable algorithm, they will be able to reapply the process to other areas of the body.
In the years to come, this work could have a major impact on the lives of cancer patients around the world by making effective treatments more accessible.