Scientists have recently developed a method to predict the survival time of cancer patients using artificial intelligence (AI) with high accuracy.
Currently, a doctor’s prognosis regarding how many years a person will live after being diagnosed with cancer can be somewhat vague and inaccurate. Some patients live much longer than expected, while others deteriorate rapidly after receiving their diagnosis and pass away within weeks.
A new study conducted by scientists at the McGaw Medical Center at Northwestern University, presented at the American College of Surgeons (ACS) Annual Clinical Congress, aims to address this issue. The researchers utilized a vast database, which includes medical records of nearly all cancer patients in the United States, to train the AI to make accurate predictions within a 10-month timeframe.
The tool analyzed the records of 400,000 patients diagnosed with breast, thyroid, and pancreatic cancers between 2015 and 2017, sourced from the National Cancer Database (NCDB). This dataset contains records of 72% of newly diagnosed cancer cases in the U.S.
The scientists used this data to create algorithms that identify patterns in symptoms, diagnosis timelines, and survival probabilities for patients over five years. Subsequently, the algorithm ranked the factors that significantly impact the prognosis for patients.
Illustration of artificial intelligence applications in healthcare. (Image: Freepik).
The factors include distinctive signs of the tumor, the age of the patient at diagnosis, tumor size, and the time from diagnosis to treatment. The lead author of the study, Dr. Lauren Janczewski, stated that the AI algorithm is highly accurate, predicting actual survival chances within 9 to 10 months.
Dr. Sarah Holmes, medical director at Marie Curie, mentioned that the device will help provide more accurate prognoses for patients, especially those with advanced-stage cancer. From a clinical perspective, Dr. Holmes believes that clear prognoses help maintain the quality of life for patients until the end of their lives.
Today, cancer is no longer a death sentence; it is gradually becoming a condition that many people can live with. The total number of diagnoses continues to rise, but the mortality rate for the disease has decreased by about a quarter since the mid-1980s. This change raises questions about how people cope with a terminal diagnosis, even when they have many years to live.
