A machine learning method can identify a cancerous single cell by detecting its pH

Cancer cells exhibit several key differences from healthy cells that help identify them as dangerous. For example, the pH – the acidity level – inside a cancer cell is not the same as the pH in a healthy cell.

Researchers from the National University of Singapore have developed a method learning tool to determine if a single cell has cancer by detecting its pH. They report on their work in the journal APL Bioengineering, from AIP Publishing.

“The ability to identify single cells is of paramount importance in the field of precision and personal therapy,” said Chwee Teck Lim, one of the authors. “This is because this is the only way to describe the heterogeneity associated with any biological sample.”

Lim explained that other methods of studying a single cell can cause toxic effects or even kill the cell. Their approach, however, can differentiate cells that come from normal tissues from cells that come from cancerous tissues, as well as among different types of cancer, while maintaining they are the living cells.

The method relies on treating the cells with bromothymol blue, a pH-sensitive dye that changes color depending on the acidity of the solution. Each cell type exhibits its own distinct fingerprints of red, green and blue (RGB) based on its intracellular acidity. As a cancerous mutation alters the pH of the cell, an unhealthy cell responds to bromothymol blue, leading to normal movement of its RGB fingerprints.

By training a machine learning algorithm to map a variety of colors to the disease state of individual cells, the authors can easily identify unwanted movement. This allows them to determine cell health using just simple, standard equipment: an inverted microscope and a color camera.

“Our approach allows us to classify single cells of different human tissues, both normal and cancer, by focusing only on the levels of sexual acidity that all cell types tend to display, and using equipment simple and cheap, “Lim said.

For therapeutic implementation of this procedure, medical professionals need to obtain a sample of these abnormal cells.

“One potential use of this technique would be in liquid biopsy, where tumor cells that have escaped the primary tumor can be separated in such a minimally invasive manner from body fluid,” Lim said.

The group looks forward to further developing the concept to try to detect different levels of damage from the cells. They visualize a real-time version of the procedure, in which cells suspended in solution can be automatically identified and manipulated.