AI improves cancer screening

Artificial intelligence improves cancer screening

Read the press release this case study was adapted from on Intellectual Ventures’ website.

Who:

A partnership led by investigators from the National Institutes of Health’s National Cancer Institute (NCI) and the Global Good Fund.

The Challenge:

Cervical cancer screening is complex, often requiring expensive screening equipment, multiple visits to a specialist and trained practitioners able to identify and detect abnormalities present in test results. In low-resource settings where high tech screening tools are not available, a low-cost and convenient method called visual inspection with acetic acid (VIA) is used to work around these challenges. Unfortunately, the method is often inaccurate and cervical cancer remains a leading cause of illness and death among women living in regions with limited health care resources.

The Opportunity:

Researchers at the NCI and Global Good developed an artificial intelligence (AI) algorithm, called automated visual evaluation, that is able to analyze digital images of a patient and accurately detect any pre-cancerous changes that might call for medical intervention. The AI approach to screening is easy to perform in a single visit and can be carried out by health practitioners with minimal training.

“Through this public-private partnership, a promising new way to screen for cervical cancer has the potential to completely transform detection of precancerous cervical changes.”

How It Works:

The algorithm for automated visual evaluation used more than 60,000 images collected during an 18-year cervical cancer screening study with upwards of 9,400 participants. The photos were digitized and then utilized to train an AI algorithm to distinguish between images that depict cervical conditions requiring treatment from those that do not.

The screening approach can be used on a cell phone or similar digital camera device and, according to Mark Schiffman, M.D., M.P.H., of NCI’s Division of Cancer Epidemiology and Genetics, “the computer analysis of the images was better at identifying precancer than a human expert reviewer of Pap tests under the microscope (cytology).”

Through this public-private partnership, a promising new way to screen for cervical cancer has the potential to completely transform detection of precancerous cervical changes, which could ultimately help significantly reduce the rate of deaths from cervical cancers, even in low-resource settings. Once further research confirms variations of these findings, the team behind Global Good will advance development of this innovation for broader use.

These types of innovation have the potential to save lives and change the trajectory of deadly diseases around the world. That’s why it is critical that innovators are supported, incentivized and protected through adequately enforced, strong intellectual property systems. Safeguarding and rewarding innovation is ultimately what makes these meaningful strides possible and accessible to patients on a global scale.

Read More >