An AI-based mechanism tests cancer cells according to their acidity

Healthy and cancerous cells can look similar under a microscope. One way to differentiate them is by examining the level of acidity, or pH level, within the cells.

Tapping on this differential feature, a research team from the National University of Singapore (NUS) has developed a device that uses artificial intelligence (AI) to determine whether a single cell is healthy or cancer by analysis on its pH. All cancer tests can be completed within 35 minutes, and single cells can be classified with an accuracy rate greater than 95 percent.

The research, led by Professor Lim Chwee Teck, Director of the Institute for Health & Technology Innovation (iHealthtech) at NUS, was first published in the journal APL Bioengineering on March 16, 2021.

“The ability to analyze single cells is one of the holy graves of health innovation for precision medicine or personalized medicine. Our empirical study demonstrates the ability to use our technology as a rapid tool , cheap and correct for cancer diagnosis., ”said Dr. Lim, who is also from the NUS Department of Biomedical Engineering.

Using AI to detect cancer

Conventional methods for examining a single cell can cause toxic effects or even kill the cell. The approach developed by Professor Lim and his team can, however, differentiate between cells derived from normal and cancerous tissue, as well as among different cancer types. Crucially, all of these can be achieved as long as they keep the cells alive.

The NUS team’s method relies on the application of bromothymol blue – a pH-sensitive dye that changes color according to the acidity level of the solution – on living cells.

As a result of its intracellular activity, each cell type exhibits its own ‘fingerprint’ which contains its own unique combination of red, green and blue (RGB) components when illuminated. Cancer cells have a different pH compared to healthy cells, so they react differently to the dye, which changes their RGB fingerprint.

Using a standard microscope equipped with a digital color camera, the RGB components emitted from the color inside the cells are captured. Using an AI-based algorithm they developed, NUS researchers were able to quantitatively map the specific acidic fingerprints so that the cell types studied could be easily and accurately identified.

Thousands of cells derived from several cancerous tissues can be plotted simultaneously, and single-cell features can be extracted and studied. Compared to the usual standard methods for cancer cell imaging that require several hours, the process developed by the NUS team can be completed in less than 35 minutes.

“Unlike other cell analysis methods, our approach uses simple, inexpensive equipment, and does not require lengthy preparation and solemn tools. Using AI, we can screen cells faster and more efficiently. right.

In addition, we can monitor and analyze living cells without causing any poisoning to the cells or the need to kill them. This would allow further downstream analysis of living cells that may be needed, “explained Dr Lim.

Opening the door for faster detection

Because of their simple, low-cost, fast and high-throughput approach, the research team plans to develop a real-time version of this method where cancer cells can be automatically identified, and separated quickly. for further downstream molecular analysis, such as sequential genetics, to determine any drug-treatable mutations.

We are also exploring the possibility of conducting a real-time study of the circulation of suspended cancer cells in blood. One potential application for this is in liquid biopsy where tumor cells that have escaped primary tumors can be separated in such a minimally invasive way from body fluids as blood. “

Lim Chwee Teck, Professor and Director, Institute of Health & Technology Innovation (iHealthtech), National University of Singapore

In addition, the group looks forward to advancing their concept to detect varying degrees of malignancy from the tested cells.