An alternative to chemotherapy called immunotherapy has shown promise in treating cancer, and new artificial intelligence tools could help ensure patients have the best possible experience. there is.
First approved in 2011, immunotherapy harnesses a cancer patient's own immune system to target and fight cancer.
It doesn't work for everyone, but for the 15% to 20% of people who do, it can be lifesaving.
What is artificial intelligence (AI)?
Like other drugs, immunotherapy can have harmful side effects, which can be severe in some people.
Studies show that about 10-15% of patients develop “significant toxicity.”
An alternative to chemotherapy called immunotherapy has shown promise in treating cancer, and new artificial intelligence tools could help ensure patients have the best possible experience. there is. (St. Petersburg)
Chicago-based GE Healthcare is collaborating with Vanderbilt University Medical Center (VUMC) in Nashville, Tennessee, to create an AI model designed to remove some of the uncertainty surrounding immunotherapy. did.
Over five years of development, this AI model was trained on the electronic health records (EHRs) of thousands of patients to recognize patterns of patient response to immunotherapy, with a focus on safety and efficacy. did.
Study finds AI models may help predict lung cancer risk in non-smokers: 'significant progress'
“This model predicts which patients are likely to benefit from immunotherapy and which are not,” Jan Walber, global digital product leader for pharmaceutical diagnostics at GE Healthcare, said in an interview with Fox News Digital. It's a thing,” he said.
“We also predict which patients are likely to develop one or more significant toxicities.”
When retrieving data from a patient's health record, the model looks at demographic information, image scans, past diagnoses, lifestyle habits (such as smoking), medication history, and more.
First approved in 2011, immunotherapy harnesses a cancer patient's immune system to target and fight cancer. (St. Petersburg)
“All of this data is either already collected by the patient's oncologist or written up in advance on a form in the waiting room,” said Travis Osterman, a medical oncologist and deputy chief medical information officer at Vanderbilt University Medical Center. It's written down,” he said. he said in an interview with Fox News Digital.
(Osterman is working with Wolver to develop AI models.)
Breast cancer breakthrough: AI predicts one-third of cases before diagnosis in mammography study
“We're not asking for additional blood samples or complex imaging. These are all data points we're already collecting, such as vital signs, diagnoses, and laboratory values.”
According to an article published in the Journal of Clinical Oncology Clinical Cancer Informatics, one study showed that AI models predicted patient response to immunotherapy with 70% to 80% accuracy.
“The model is not perfect, but this is actually a very good result,” Wolver said. “In the clinic, these models can be implemented with little additional effort as no additional measurements are required.”
“Natural progression”
This type of technology is “a natural progression of what we've been doing in medicine for a long time,” Osterman said.
“The only difference is that instead of studying the patient, we're collecting the entire medical record and looking for risk factors that contribute to the outcome,” he said in an interview with Fox News Digital.
This AI model was trained on the electronic medical records of thousands of patients to recognize patterns of patient response to immunotherapy, with a focus on safety and efficacy. (St. Petersburg)
Immunotherapy generally has lower response rates than chemotherapy, Osterman noted, but some patients can have an “incredible response” and ultimately be cancer-free.
“I would be horrified to know that one of my patients for whom I did not give immunotherapy may have had a very good response,” he told Fox News Digital. .
Conversely, Osterman noted that in rare cases, immunotherapy can have serious side effects.
“I would be horrified to know that one of my patients, to whom I did not give immunotherapy, could have been one of the people who had a very good response.”
“I think about half of patients don't have any side effects, but for those who do, some of them are truly life-changing,” he said.
“We don't miss anyone, but we don't want to hurt anyone.”
According to Osterman, the core of the AI project is the ability to “put all the information into the lab” so that oncologists can advise patients about the risks and benefits of this particular treatment and keep them fully informed. About their care so they can make the best decisions based on.
Combining technology and human expertise
Dr. Mark Siegel, clinical professor of medicine at New York University Langone Medical Center and medical contributor for Fox News, was not involved in the development of the AI model but commented on its potential.
“AI models are emerging to help manage the cancer treatment response,” he told Fox News Digital.
Dr. Mark Siegel, clinical professor of medicine at New York University Langone Medical Center and medical contributor for Fox News, was not involved in the development of the AI model but considered its potential. “AI models are emerging to help manage responses to cancer treatments,” he said. (Dr. Mark Siegel)
“These allow for more treatment options and allow us to more accurately predict outcomes.”
Such AI models are an example of “the essential future of personalized medicine,” which “approaches each patient differently and uses genetic and protein analysis to accurately analyze and treat cancer. ''Siegel said.
As long as doctors and scientists continue to be in charge “rather than computers and robots,” there is “no downside,” Siegel said.
Potential limitations
Experts acknowledge that AI models have some limitations.
“It's clear that the model is not going to return 100% accuracy,” Wolver told Fox News Digital. “So there are what are called false positives and false negatives.”
New AI 'cancer chatbot' provides 24/7 support to patients and families: 'empathetic approach'
He pointed out that the tool is not a “black box” that provides definitive answers. Rather, it is a tool that provides data points to clinicians and informs them when making patient management decisions.
Osterman noted that the AI model uses a “relatively small dataset.”
“We hope to be able to refine our predictions even further by learning larger datasets,” he says.
One of the researchers said the tool is not a “black box” that provides definitive answers. Rather, it is a tool that provides data points to clinicians and informs them when making patient management decisions. (St. Petersburg)
The team is currently looking for partnerships that can test the AI model in new settings and achieve even higher accuracy of predictions.
Another challenge, Osterman said, is the need to integrate these AI recommendations into workflows.
“This is fairly new for us in the medical community, and I think we're all going to be wrestling with this issue,” he said.
Click here to sign up for our health newsletter
Looking to the future, once the AI model receives the necessary regulatory approvals, GE Healthcare plans to make the technology widely available to clinicians, and perhaps other medical fields such as neurology and cardiology. We also plan to expand this.
There is also the possibility of incorporating it into drug development.
“We don't miss anyone, but we don't want to hurt anyone.”
“One of the struggles that pharmaceutical companies have is that some of the drugs that are really helpful for some patients may be very toxic for others,” Osterman said. Told.
“Being able to select which patients can participate in trials and exclude those at highest risk of toxicity could mean the difference between whether or not the drug becomes available.”
CLICK HERE TO GET THE FOX NEWS APP
He added: “If this allows us to tailor the exact risk to the patient, I'm all for it.”
Ultimately, Osterman said, “This is a really exciting time for the medical community…We're going to look back and see this as the golden age of AI recommendations, and I think AI recommendations are probably here to stay. ” he said.
For more health stories, visit www.foxnews/health.
Melissa Rudy is a health editor and a member of the Lifestyle team at Fox News Digital. Story tips can be sent to melissa.rudy@fox.com.