Artificial Intelligence is Helping with the Fight Against Colorectal Cancer
Colorectal cancer is one of the most common and deadly forms of cancer in the United States. It affects both men and women, and it often has no symptoms until it is too advanced to treat. That's why it is important to get screened regularly and to find out the specific type of colorectal cancer you have, so you can get the best treatment possible.
But how do doctors determine the type of colorectal cancer you have? They usually take a small sample of your tumor (called a biopsy) and run various tests on it to look for genetic markers. These markers are like fingerprints that tell us more about the cancer cells and how they behave. For example, some markers can indicate how fast the cancer is growing, how likely it is to spread, or how well it will respond to certain drugs.
However, these tests can take a long time, cost a lot of money, and require specialized equipment and expertise. Wouldn't it be great if there was a faster, cheaper, and easier way to get the same information from a biopsy?
That's where artificial intelligence (AI) comes in. AI is a branch of computer science that uses algorithms to learn from data and make predictions or decisions. AI can be used for many purposes, such as recognizing faces, translating languages, or playing games. But it can also be used for medical purposes, such as diagnosing diseases, analyzing images, or recommending treatments.
Recently, a team of researchers developed a new AI-based method to predict genetic markers in colorectal cancer from routine pathology slides. (1) Pathology slides are thin slices of tissue that are stained with special dyes and examined under a microscope. They are commonly used to diagnose diseases and study their features.
The researchers used a type of AI called deep learning, which mimics the way the human brain processes information. Deep learning can learn from large amounts of data and find complex patterns that are hard to detect by humans or other methods. The researchers used a specific kind of deep learning called transformer networks, which are better at handling sequential data, such as text or speech.
The researchers trained their AI model on thousands of pathology slides from colorectal cancer patients, along with their corresponding genetic markers. Then they tested their model on new slides from different patients and compared its predictions with the results of the clinical tests. They found that their model was able to predict the genetic markers with high accuracy, speed, and reliability. Their model also performed better than other existing AI methods that use different types of deep learning.
Interestingly, even though this model was only trained on samples from resections obtained during surgery, it also performs well on biopsy tissue samples from colonoscopy, which makes it more useful for clinical practice.
AI is changing the way we diagnose and treat diseases. It can help us find out more about our health and make better decisions. It can also help us save lives.
Reference:
Wagner SJ, Reisenbüchler D, West NP, et al. Transformer-based biomarker prediction from colorectal cancer histology: A large-scale multicentric study. Cancer Cell. 2023;41(9):1650-1661.e4. doi:10.1016/j.ccell.2023.08.002
