Home Learn A brand new AI-based risk prediction system could help catch deadly pancreatic cancer cases earlier

A brand new AI-based risk prediction system could help catch deadly pancreatic cancer cases earlier

0
A brand new AI-based risk prediction system could help catch deadly pancreatic cancer cases earlier

A brand new AI system could help detect essentially the most common type of pancreatic cancer, recent research has found.

Pancreatic cancer is a difficult disease to detect. The pancreas itself is hidden by other organs within the abdomen, making it tough to identify tumors during tests. Patients also rarely experience symptoms within the early stages, meaning that the vast majority of cases are diagnosed at a sophisticated stage—once it’s already spread to other parts of the body. This makes it much harder to cure.

Consequently, it’s essential to attempt to catch pancreatic cancer on the earliest stage possible. A team of researchers from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) worked with Limor Appelbaum, a staff scientist within the department of radiation oncology on the Beth Israel Deaconess Medical Center in Boston, to develop an AI system that predicts a patient’s likelihood of developing pancreatic ductal adenocarcinoma (PDAC), essentially the most common type of the cancer. 

The system outperformed current diagnostic standards and will someday be utilized in a clinical setting to discover patients who may gain advantage from early screening or testing, helping catch the disease earlier and save lives. The research was published within the journal last month.

The researchers’ goal was a model able to predicting a patient’s risk of being diagnosed with PDAC in the following six to 18 months, making early-stage detection and cure more likely. To develop it, they examined their existing electronic health records.

The resulting system, often known as PRISM, consists of two AI models. The primary uses artificial neural networks to identify patterns in the information, which include patients’ ages, medical history, and lab results. It then calculates a risk rating for a person patient. The second AI model was fed the identical data to generate a rating, but used a less complicated algorithm.

The researchers fed the 2 models anonymized data from 6 million electronic health records, 35,387 of which were PDAC cases, from 55 health-care organizations within the US.

The team used the models to guage patients’ PDAC risk every 90 days until there was no more sufficient data or or the patient was diagnosed with pancreatic cancer. They followed up on all enrolled patients from six months after their first risk evaluation until 18 months after their last risk evaluation to see in the event that they were diagnosed with PDAC in that point. 

Amongst individuals who developed pancreatic cancer, the neural network identified 35% of them as high risk six to 18 months before their diagnosis, which the authors say is a big improvement over current screening systems. For a lot of the general population, there isn’t a beneficial screening routine for pancreatic cancer the best way there’s for breast or colon cancer, and the present standard screening criteria catch around 10% of cases.

Given how vital it’s to detect the disease on the earliest stage possible, the system looks promising, says Michael Goggins, a professor of pathology and pancreatic cancer specialist at Johns Hopkins University School of Medicine, who was not involved within the project.

“It will be anticipated that such a model would improve the present landscape,” he says. “But it surely really must be very early to make the most important impact.”

It’s possible that some people could have developed advanced pancreatic cancer throughout the six-to-18-month window, meaning it may very well be too late to treat them effectively by the point they’ve received a risk assessment, he says.

While this particular study is retrospective, taking a look at existing data and tasking the models with making hypothetical predictions, the team has began work on a study that may gather gather data on existing patients, compute their risk aspects, and wait to see how accurate the predictions are, says Martin Rinard, a professor of electrical engineering and computer science at MIT, who worked on the project.

Previously, other AI models built with data from a selected hospital sometimes didn’t work as well when supplied with data from one other hospital, he points out. That may very well be right down to all types of reasons, comparable to different populations, procedures, and practices.

“Because we’ve got what’s coming near data from essentially a really significant fraction of the complete population of the USA, we’ve got hopes that the model will work higher across organizations and never be tied to a selected organization,” he says. “And since we’re working with so many organizations, it also gives us an even bigger training set.”

In the longer term, PRISM may very well be deployed in two ways, says Appelbaum.

First, it could help single out patients for pancreatic cancer testing. Second, it could offer a broader style of screening, prompting people without symptoms to take a blood or saliva test that will indicate whether or not they need further testing.

“There are tens of 1000’s of those models for various cancers on the market, but most of them are stuck within the literature,” she adds. “I feel we’ve got the pathway to take them to clinical practice, which is why I began all of this—in order that we are able to actually get it to people and detect cancer early. It has the potential to avoid wasting many, many lives.”

LEAVE A REPLY

Please enter your comment!
Please enter your name here