Team Develops Label-Free Method to Identify Metastatic Cancer Cells

Researchers from Texas Tech University have developed a deep learning model that can classify cancer cells by type. The tool only requires a simple microscope and a small amount of computing power to produce results on par with more sophisticated techniques. Cancer cells that initiate metastasis, or the spread of the disease from its primary location, are different from cancer cells that stay in the original tumor. Distinguishing metastasis-initiating cell types can determine the cancers’ severity and help medical practitioners decide on the best course of treatment.

The current methods of categorizing cancer cells involve advanced instruments, time-consuming biological techniques, or chemical labels, all of which require resources and effort that could be spent exploring different areas of cancer prevention and recovery. Moreover, attaching magnetic nanoparticles to track cancer cells could affect their downstream analysis and the measurements’ integrity. To overcome these challenges, the researchers developed a “label-free” identification method of metastatic potential.

After feeding it an image, the team’s neural network converts the data to a probability. A result lower than 0.5 categorizes the cancer as one cell type, while a number higher than 0.5 designates another. The tool was trained to optimize the accuracy of predictions with images of two cancer cell lines, reaching over 94% accuracy across the data sets utilized in the study.

For future work, the authors hope to extend and generalize the model to include both single cells and clusters, as research shows that circulating tumor cell clusters are more responsible for the spread of cancer. The high accuracy obtained from the predictions demonstrates that the system can be retrained on a large-scale clinical dataset.

Machine learning approaches can help with the diagnostic prediction from circulating tumor cells in liquid biopsy or from a primary tumor in solid biopsy. This label-free identification method of metastatic potential could be a game-changer in cancer diagnostics, as it provides a new approach to categorizing cancer cells based on their typography. The findings were published in the journal APL Machine Learning.