Colorectal cancer is the second most common cause of death from cancer in Denmark. A national screening program introduced in 2014 has already reduced mortality from colorectal cancer. A central criterion for success is how frequently preliminary stages of cancer are discovered and removed through colonoscopies. When more preliminary stages of cancer are removed, fewer patients die from colorectal cancer.
Unfortunately, not all patients’ bowels are sufficiently well cleaned out before the examination, and it is estimated that 42 % of all preliminary stages of cancer that are overseen because of inadequate cleaning. At the same time, there is no sufficiently objective measure for how effectively cleaning has been carried out.
“With the support from the Vissing Foundation, we will develop a Copenhagen Automatic Bowel Preparation Score (CA-BoPS). This will be based on an existing database of 2000 colonoscopy recordings,” says Kristoffer Mazanti Cold, MD and PhD student at the Copenhagen Academy for Medical Education and Simulation (CAMES), at the Rigshospital.
CA-BoPS will be an algorithm based on artificial intelligence. The researchers will train the algorithm using video recordings and expert assessments, against the background of the Boston Bowel Preparation Score, a subjective score which already exists.
“In that way we can construct an automatic score that is independent of the operator. It must be possible to use it as a tool to support the decision on when the patient should have the next colonoscopy, and to evaluate different cleaning methods,” says Kristoffer Mazanti Cold.