Scientists who developed it tested it and it provided highly accurate estimates of patients’ life expectancy
American scientists have developed a ground-breaking Cancer Survival Calculator that harnesses the power of artificial intelligence to provide patients with a personalized prognosis.
Using a type of artificial intelligence known as machine learning, the interdisciplinary research team tested the computer on a cancer data set. According to the team, the computer provided highly accurate estimates of the life expectancy of patients diagnosed with breast, thyroid and pancreatic cancer.
Usually, predictions of the survival rate of cancer patients are based mainly on the stage of the disease. “There are many other factors that may affect a patient’s survival beyond staging criteria,” lead study author Dr. Lauren Janfscheski, ACS Cancer Programs clinical researcher. “We sought to develop this Cancer Survival Calculator to provide a more personalized assessment of their cancer prognosis,” he added.
This revolutionary prototype tool was tested on a huge national cancer dataset, with the aim of assessing five-year survival rates. After collecting data from patients diagnosed with these three types of cancer in 2015 and 2017, the research team tapped the National Cancer Data Base (NCDB), an extensive repository that contains records for 72% of new cancer diagnoses in the U.S. .P.A. Machine learning algorithms were trained on 3/4 of this data, identifying patterns linking diagnostic features to five-year survival rates. The remaining data were then used to evaluate the accuracy of the prototype.
The extensive study incorporated data from over 250,000 breast cancer patients, 76,624 thyroid cancer patients and 84,514 pancreatic cancer patients. The research reveals that a host of specific characteristics associated with patients, tumors and treatments significantly shaped survival outcomes in all three cancer types.
The researcher pointed out that many of the determining factors, such as tumor size, are part of conventional cancer staging. However, the study findings show that a wider range of factors influence a patient’s survival beyond the stage of the disease.
The researchers emphasize that the Cancer Survival Calculator takes into account specific tumor markers and treatment factors that affect a patient’s outlook, which many current tools do not.In addition, it is based on a more comprehensive data set than the NCDB. The team hopes their tool will help healthcare professionals as they can offer patients more informed and accurate prognosis estimates.
