Cytology based screening programs undoubtedly reduce the mortality rate of cervical cancer; nevertheless, they are cumbersome, labor intensive and expensive. Detecting premalignant and/or malignant cells is a needle in a haystack problem; each liquid-based Pap test requires having at least 5,000 cells and in an early-stage positive sample only a few (~5-10 cells) will be cancerous.
To better understand the massive scale of the screening programs, in the United States alone, 66 million Pap tests are performed each year, of which approximately 3.5 million tests are classified as “abnormal” and require additional follow up.
To fundamentally address all these issues and shortcomings we are developing a field-portable and cost-effective computational microscopy and sensing platform that is coupled to machine learning tools to facilitate automated and accurate diagnosis of premalignant and/or malignant cells in Pap tests.
At the core of this platform, we have lens free holographic super-resolution on-chip microscopy, an emerging technology that our founding group at UCLA pioneered over the past few years, which can achieve sub-micron resolution over a very large field-of-view, e.g., ~20 mm2, which is more than two-orders of magnitude larger compared to a lens-based optical microscope having a similar spatial resolution.