Artificial Intelligence (AI) has revolutionized oncology care as a whole1 and is beginning to show great potential to transform the role of radiotherapy oncologists. According to Matthew A. Manning, MD, the future of AI in radiation oncology could lie in diagnostics and clinical decision support.
“AI can take in large amounts of data and process more data than a human could in a reasonable time frame. We expect clinical decision support tools to come to radiation oncology to assess individual patient records and refine treatment recommendations based on the fact that artificial intelligence can also look for patterns that humans who want to marginalize information might not recognizable, helping us to recognize patterns in patients may be at higher risk of complications or higher risk of failure unless treatments improve,” said Manning, chief of oncology at Cone Health.
The main application of AI in radiation oncology is to support radiation. Manning explained that adaptive radiation therapy and motion management can be more precise with AI. In addition, AI has been brought into important clinical tasks such as data sharing and response assessment.
“Even in radiation oncology, there are various platforms for medical records necessary for the practice, which are often separate from the hospital medical record. It is important to create these interfaces that make it possible to reduce the redundancy of work for both clinicians and administrative staff. Tools using AI and business intelligence are accelerating our efforts in radiation oncology,” said Manning.
In the interview, Manning discussed the many tasks of radiation oncology that have been enhanced by AI and their likely utility in cancer diagnosis and treatment decisions.
Targeted Oncology: Can you talk about the different ways AI is being used in radiation oncology?
Manning: AI is the backbone of some of our clinical decision support tools. It helps with driving automation. We can reach oncology or oncology in general [with AI]. The amount of medical science information is growing exponentially, doubling about every 2 months. While it would not be possible for physicians to update knowledge with the latest science, the use of knowledge-based AI and non-knowledge-based AI provides clinical decision support.
Radiation oncology practices also require different medical record platforms that are often separate from the hospital medical record. It is important to create these interfaces that make it possible to reduce the redundancy of work for both clinicians and administrative staff. Tools using AI and business intelligence are accelerating our efforts in radiation oncology.
By accelerating radiation oncology treatment planning and administrative burdens, we are able to accelerate the onset of cancer treatment for patients with growing cancer. Delays in treatment improve outcomes and cure rates.
©heitipaves – stock.adobe.com
What about the role that AI plays in data sharing? What tools are available or being developed?
I think the role of AI for any technology is just beginning. I think we’ll see a lot fill up over the years. One area where AI is currently helping is in software platforms related to billing and coding or fees for treatment. This data needs to be checked against what has been documented, and by processing natural language through AI evaluating existing documentation, we have software platforms that can verify what we bill correctly for weeks. It flags cases and notifies you when necessary.
There are recent advances such as image-guided and adaptive radiation therapy. What do you know about how AI is used for these purposes?
In image-guided radiation therapy, the position of the target is checked while the patient is lying on the treatment table. The patient is repositioned millimeter by millimeter to ensure that the target is in the cross hairs of the radiation. There are many ways to perform image-guided radiation, including prior to radiation just making sure patients are properly set up. Then there is even intrafraction monitoring, in which the image guidance takes place during the irradiation itself. It only takes a few minutes and allows you to adjust the irradiation position.
There is also adaptive radiation therapy, where the patient arrives and their images are taken and their planning is updated to account for changes in their anatomy. Adaptive radiation therapy in its early days in the United States. In the future, we expect not only to customize radiation treatments for each patient, but that we will customize radiation treatments for each patient every day based on their anatomy. We will be able to see anatomical changes based on what patients eat, drink and their body weight. Such things can be adjusted to optimize radiation therapy.
Can you talk about motion management by AI? Why are these tools important?
Movement management is very important. People who live move. Their lungs move as they breathe their heartbeats, and things like lung cancer and even liver cancer move up and down the respiratory cycle. In the past we had to account for this movement and only treat the entire potential range of tumor locations.
But nowadays with movement management we can have tools to reduce movement by compressing the abdomen and tools to track movement such as: With motion management, we are able to reduce the amount of birthmark tissue exposed to high-dose radiation and potentially reduce complications from radiation necrosis reactions.
What role does AI play in reaction evaluation?
Some tumors change when treated with radiation. Ideally, they shrink, sometimes they initially increase. With the evaluation of the response, we are able to assess the target and revise the radiation treatment with adaptive radiation therapy to ensure that the radiation reaches the tumor.
One instance where this is relatively common is in central cancers that block healthy lung tissue. When we first meet the patient, it can be difficult to distinguish their cancer from their congested collapsed lung. Since radiation precedes the tumor, the airways in the lung areas open up. Now we may be able to refine the radiation to just cover the tumor and not treat healthy lungs with high-dose radiation.
How do you think AI will be helpful in the future?
AI is growing in strength and in terms of training, and we’re starting to see use cases for AI. AI is currently finding its way into computer-aided recognition on diagnostic films. If we look at a chest CT and try to find more knowledge, we see that the AI is very powerful at finding spots that the human eye might miss. With regard to radiation oncology, the ultimate use will be in the context of clinical decision support.
AI can take in large amounts of data and process more data than a human could in a reasonable time frame. We expect to see clinical decision support tools for radiation oncology come to market that assess individual patient records and refine treatment recommendations based on this artificial intelligence, which can also look for patterns that are useful for people who want to marginalize information. may not be apparent, and help us identify patterns that patients may be at greater risk of complications or at greater risk of failure if treatments do not improve.
I think in the current state we are starting to see artificial intelligence invading oncology diagnostics, in terms of therapeutics, no more people will rely on artificial intelligence to support radiation oncology.
REFERENCES:
Luchini C, Pea A, and Scarpa A. Artificial intelligence in oncology: current applications and future prospects. Br J Cancer. 2022 Jan;126(1):4-9. doi: 10.1038/s41416-021-01633-1
