header logo image

AI tool created to guide colorectal cancer care with more precision – Scope

October 6th, 2020 4:55 am

A new modeling tool may be able to help doctors assess which treatments are best for individual patients with colorectal cancer. The artificial intelligence program analyzes a patient's disease details -- such as the stage of cancer and other chronic conditions -- and compares those details to other colorectal cancer cases to predict the patient's chance of surviving past 10 years.

"Predicting survival of cancer patients as a means to help determine treatments is not new," said Jean Emmanuel Bibault, MD, PhD, a radiation oncologist who led the study. "But current standard techniques are not very accurate, and we're hoping that by using AI we can bring more precise information to doctors as they make crucial decisions about care."

Predicting a cancer patient's survival time lends valuable insight into the best course of treatment for both the long and short term, helping to determine what is likely most suitable.

The online tool works by assessing 32 details about an individual patient, such as age, the stage of the cancer, exercise habits, cholesterol levels, history of chronic disease and much more.

After these details are input into the tool, the algorithm predicts how long that person might live and reports a number in years. The tool also provides context, citing the top reasons for its calculation, such as the stage of the cancer, the patient's age at diagnosis, or how the patient was initially treated.

"From a physician's point of view, we want to know how well our patients are going to do from the get-go. We're looking at two main things: how to choose the right therapy, and if we can alter their destiny," said Daniel Chang, MD, professor of radiation oncology, who is an author of the study.

"Some folks have a bit of a nihilistic point of view," he continued, "that survival is determined by the genetics of your cancer and of your body. But the question is: Can anything we, as doctors, do change that outcome if we do it sooner or do it differently from the start? That's where I see a lot of value for this research."

An abstract on the research appeared online inGut.

Bibault, Chang and professor of radiation oncology, Lei Xing, PhD, devised the algorithm powering the prediction tool with data made available through the National Institutes of Health, from thousands of de-identified patients who have or had colorectal cancer at various stages and are of varying ages.

The team trained the algorithm to track survival of thousands of patients, in conjunction with the details of their disease and some details about their course of treatment. In this way, the algorithm uses the outcomes and survival rates of past cohorts to calculate the chance of survival for future patients.

So far, the tool has been about 90% accurate in predictions it made on 472 patient cases that were not used to train the tool. The tool has not been used in a clinical setting.

"The treatments that we have nowadays are becoming more and more specialized, targeted, in many cases intensified. And the reality is that not everybody is going to benefit from new treatments, therapies or technologies in the same way," said Chang.

"This algorithm could allow us a better shot at personalized medicine, and enhance our ability to tailor the treatments to be as appropriate as possible," he added.

Although patients could use the tool on their own, Bibault said the ideal application would be for doctors and patients to use the tool together. That way, doctors would be able to contextualize the result and answer any patient questions.

The team's goal is to enhance the algorithm's accuracy and to find other applications for it.

"We have laid the foundation for this model," said Bibault, "and we're hopeful it can apply to other cancer types as well."

Photo by National Cancer Institute

Read more:
AI tool created to guide colorectal cancer care with more precision - Scope

Related Post

Comments are closed.


2024 © StemCell Therapy is proudly powered by WordPress
Entries (RSS) Comments (RSS) | Violinesth by Patrick