A Baton Rouge hospital has introduced cutting-edge artificial intelligence (AI) technology aimed at improving the early detection of pancreatic cancer, a condition that is often diagnosed in its later stages. The American Cancer Society projects that in 2024, over 66,000 people will be diagnosed with pancreatic cancer, but according to this article from The Advocate, in Louisiana, where this type of cancer accounted for 7.3% of cancer-related deaths between 2016 and 2020, efforts like the AI technology used at Our Lady of the Lake Regional Medical Center are bridging the gap in screening and prevention.
In an effort to combat these statistics, Our Lady of the Lake Regional Medical Center in Baton Rouge has introducedEON technology, an advanced AI software designed to analyze CT scans and flag potential pancreatic abnormalities, even if the patient’s initial hospital visit was unrelated to pancreatic issues. For instance, if a patient underwent a CT scan following a car accident, the AI would still review the scan for signs of pancreatic cysts, potentially identifying an issue long before it developed into cancer. The software automatically notifies the hospital’s oncology team, enabling timely follow-up and treatment.
Dr. Mo Al-Efishat, a surgical oncologist at Our Lady of the Lake specializing in advanced pancreatic tumors and cysts, explained the importance of EON’s implementation. The software uses a combination of artificial intelligence and computational linguistics to analyze medical imaging reports with up to 98% accuracy. Traditional methods often miss subtle indications of pancreatic cysts, which can evolve into cancer. In contrast, EON’s sophisticated analysis ensures that cysts are not only detected but also brought to the attention of medical professionals who can schedule regular follow-ups, such as MRIs or CT scans every six to twelve months, as needed.
Dr. Al-Efishat emphasized that early detection is key in preventing pancreatic cancer. “Twenty percent of pancreatic cancers originate from mucinous cysts,” he noted. These are cysts that, if identified and monitored early, have the potential to be treated before they turn cancerous. Once a pancreatic cyst is flagged, the oncology team takes over, assessing whether further intervention is necessary. The AI system operates continuously across the hospital’s network, ensuring that any MRI or CT scan conducted within the system is evaluated for pancreatic abnormalities.
One of the critical aspects of treating this cancer is catching it at an early stage. According to Dr. Al-Efishat, when pancreatic cancer is caught in its earliest stages, the chances of survival increase substantially, with cure rates ranging from 30% to 40%. He shared a case where, after performing a robotic surgery, the biopsy results revealed high-grade dysplasia, the final stage before cancer develops. Dr. Al-Efishat expressed relief, knowing that his team had intervened at just the right moment. Once pancreatic cancer develops, survival rates drop dramatically—from 100% to 30% upon diagnosis.
Unlike breast or colon cancer, pancreatic cancer does not have an established screening method for the general population, such as mammograms or colonoscopies. There is a screening program for individuals with a family history of pancreatic cancer, but this represents a small portion of the population. The absence of a comprehensive screening program makes the need for AI technologies like EON even more crucial in bridging the gap.
Looking to the future, Dr. Al-Efishat expressed optimism about emerging research aimed at preventing pancreatic cancer. Recent studies have identified inflammation as a key driver of pancreatic cancer development. As a result, clinical trials are now underway to test the effectiveness of anti-inflammatory treatments in slowing or halting the progression of pancreatic cysts to cancer. If successful, these treatments could represent a major breakthrough in pancreatic cancer prevention, potentially reducing the number of lives lost to this devastating disease.
For more Louisiana-related articles, click here.