Artificial Intelligence Helps Cut Miss Rate of Colorectal Polyps

Intending to put a dent in colorectal cancer (CRC) cases, researchers from the US and UK applied artificial intelligence to colonoscopies in an effort to reduce the number of colorectal polyps missed due to perceptual pitfalls.

They detailed their findings in a study published in Gastroenterology.

Because of perceptual pitfalls, standard colon cancer screenings often miss several colorectal polyps. However, recent advances in AI show they can assist in locating these frequently neglected cases of colorectal neoplasia and thereby help decrease the rate of colorectal cancer cases.

The study included 230 subjects from eight different centers located in countries worldwide. All participants were undergoing CRC screening.

Researchers divided these subjects into two groups receiving a colonoscopy, one with and one without AI. However, the order of operations was different for each, as one group received their AI-assisted colonoscopy first, followed by the non-AI-assisted procedure. The other group received their procedures in the opposite order.

Researchers then calculated the differences in adenoma miss rate (AMR), which refers to the number of histologically verified lesions detected at the second colonoscopy divided by the total number of lesions detected at the first and second colonoscopies.

The results showed that AMR was 15.5 percent for the group that underwent the AI colonoscopy first and 32.4 percent for the group that received the non-AI colonoscopy first. Research also showed that the AI-first group saw lower AMR regarding the proximal and distal colon.

Researchers concluded that AI could help drive a two-fold reduction in the miss rate of colorectal neoplasia.

Using the assistance of AI when performing procedures is becoming frequent.

A recent collaboration shows how AI tools used during colonoscopy can significantly improve health equity. This project occurred through a partnership between Medtronic and the American Society for Gastrointestinal Endoscopy and intended to increase outreach to underserved communities, using AI to improve colon cancer screenings.

Another study from March 2020 published in Clinical Cancer Research describes how AI can improve cancer therapy.  By applying AI tools to CT scans, specifically algorithms created based on information from the first treatment, they can then assist in predicting tumor response, enabling a more precise approach to medicine.

The ongoing COVID-19 pandemic has also driven the development and use of AI and machine learning.

Researchers from Yale University recently developed a multiscale PHATE machine learning tool that provides a detailed analysis of millions of immune cells and information regarding which type could lead to COVID-19 mortality.