Scholars from the Cedars-Sinai recently developed an innovative medical tool that is based on an artificial intelligence system. The device could accurately predict the health condition of patients who are afflicted with the effects of pancreatic cancer.
AI-Based Pancreatic Cancer Detector by Cedars-Sinai
The new AI-powered cancer detector can quickly analyze the state of a patient just by going through records of an individual's CT scans. Through the history of an individual's pancreatic images, the tool can predict whether or not a subject will develop cancer in the future.
The findings could help many specialists and even the patients detect the state of pancreatic cancer as early as possible, giving more time to alleviate the development of the condition before becoming severe and leading to death.
Cedars-Sinai's Biomedical Imaging Research Institute, professor of Biomedical Sciences and Imaging director Debiao Li, who also co-authored the study, explained that the AI device is able to get ahold of subtle symptoms appearing in early pancreatic cancer.
These hints commonly appear during the manifestation of pancreatic ductal adenocarcinoma (PDAC). The particular symptom manifests in CT scans even years before the diagnosis of pancreatic cancer in patients, Li continued.
Pancreatic cancer has many types aside from the common ductal adenocarcinoma. However, previous findings over the history of oncology studies show that PDAC is the most prevalent and the most deadly among the rest.
According to a report by EurekAlert, pancreatic ductal adenocarcinoma patients are most likely to suffer compared to those who have other types of pancreatic cancer. Less than ten percent of the individuals who were diagnosed with PDAC survive a maximum of five years after the diagnosis or initial treatments.
The latest pancreatic cancer studies, on the other hand, shows that early detection of the illness could improve the state of a patient and raises their survival rate by as much as 50 percent. The only problem that prevents people from obtaining this rate is that there is no current technology available to detect early signs of the cancer's development.
Early Detection Through CT Scan
Individuals who are developing pancreatic cancer in their systems are commonly afflicted with symptoms including unexplained weight loss and extreme abdominal pain. But due to being similar to other basic illnesses, the signs are often overlooked.
Cedars-Sinai's Basic and Translational Pancreas Research and program director of the Gastroenterology Fellowship Program director Stephen Pandol, who also served as co-author for the study, said the PDAC has no unique symptoms that could be used for early diagnosis.
The new AI cancer detector, according to Pandol, could assist in the early screening of patients at risk. The device could work through the use of images and other information collected from a person after going for a CT scan due to abdominal pain and other related issues.
The device was tested on previous CT scans of patients who were diagnosed in the last 15 years in order for the authors to know its accuracy. The scans were carried out three to six years prior to the cancer detection. About 36 of these subjects had hit the criteria. All of these patients were admitted by ER facilities due to abdominal pain.
The AI can detect even the slightest changes in the scans provided to it. Details that the tool could pick up include the pancreas molecular disparity and even the surface of the organ itself.
The experts noted that the AI-based cancer detector reached an 86 percent accuracy in predicting the development of the illness. The study was published in the journal Cancer Biomarkers, titled "Predicting pancreatic ductal adenocarcinoma using artificial intelligence analysis of pre-diagnostic computed tomography images."
Check out more news and information on Cancer in Science Times.
