Researchers at UC Irvine have developed a novel, machine learning-assisted biochip for rapid, affordable, and practical analysis of single cell tumor heterogeneity. The technology’s low cost and ease of manufacture makes it an optimal point-of-care diagnostic in developing countries, where early cancer detection is severely lacking.
·Point-of-care (POC) diagnostic for early cancer detection
·Single-cell, membrane impedance-based analysis of tumor heterogeneity
·Precise and easy to manufacture POC diagnostic for early cancer detection in developing countries that lack advanced medical infrastructure
·Novel machine learning-assisted impedance microcytometry platform for single cell analysis of tumor heterogeneity
·Requires small samples volume (nL-µL range)
~20minutes to manufacture using a commercial inkjet printer by any minimally trained end user
According to the World Health Organization, 58% of deaths from breast cancer in developing nations occur due to the lack of early detection programs and efficient diagnostic facilities. Furthermore, breast cancer survival rates in these countries are ~40%, which is significantly behind developed nations with survival rates as high as 80%.Impedance microcytometry (IC) is a popular technique for single cell characterization based on their electrical properties, which have been shown to change in health and disease. The technique has been used to study cellular heterogeneity, detect target cells in samples, isolate different cell subtypes, and test the efficacy of drugs on cells. However, the manufacturing of the devices used for IC relies on sophisticated equipment, skilled technicians, and relatively expensive and time-consuming fabrication processes, all of which are difficult to establish and maintain in resource-poor developing countries.
Researchers are UC Irvine have developed a more accessible IC platform that can be rapidly manufactured and precisely detect tumor heterogeneity with small amounts of input volume (nL-L range). Paired with machine learning classification and feature selection systems, this technology has been shown to effectively detect different cell types (test accuracy of 100% with n=2+ features) and normal vs. diseased cells (test accuracy of 86.8% with n=2 features, 99% with n=4 features). The system was also shown to effectively discriminate between cancer cell subtypes (test accuracy of 90.1% with n=2 features, 97.3% with n=4 features). Regarding manufacturability, this test can be fabricated with a commercial inkjet printer by any minimally trained end user in less than 20 minutes, eliminating the need for expensive fabrication facilities and making it an accessible POC diagnostic aid for those in developing countries.