AI-Powered Imaging System Can Tell If a Patient Is Cancer-Free

WRITTEN BY: Tara Fernandes

A patient undergoes chemotherapy and radiation for rectal cancer. How do doctors know whether all the malignant tissue has been destroyed?

Washington University in St. Louis researchers have developed a way to answer this question, using an innovative combination of photoacoustic microscopy, ultrasound, and “deep learning” artificial intelligence neural networks.

This new system, called PAM/US, allows physicians to detect the presence of residual tumors within rectal tissue, even differentiating cancer cells from other tissue abnormalities such as scars.

“Our PAM/US system paired with the deep learning neural network has great potential to better identify patients suitable for nonoperative management and improve patient quality of life,” explained Quing Zhu, lead author of the study.

“If we can tell after radiation and chemotherapy which patients may have a good response with no residual tumors, the patient may be able to avoid surgery.” This is a huge leap forward in the standard of care for patients with non-metastatic rectal cancers, many of which need to undergo invasive follow-up procedures after cancer therapies.

In the study, Zhu and colleagues describe how the technique would be used in a clinical setting. A patient under anesthesia would have a handheld probe inserted into the rectum. There, a rotating head takes a series of images, 360-degrees around the internal rectal wall. The combination of ultrasound and photoacoustic waves transmitted captures intricate details in the underlying vasculature of the rectum and any tumor growth. The whole procedure takes around 20 minutes to complete.

These images are then fed into an AI-based computational system that draws from a library of over 2,000 images captured from rectal cancer patients to make a diagnosis. The algorithm was found to be highly sensitive, correctly predicting the cancer status of 100 percent of rectal cancer patients.

As the next step, the team plans to expand the validation of their new platform by conducting a larger-scale clinical study in rectal cancer patients who have already undergone chemotherapy and radiation.