Microsoft is hoping to leverage artificial intelligence (AI) to spearhead the fight against cervical cancer in India. Through a collaboration with SRL Diagnostics, the company is helping to improve early detection rates with help from an AI algorithm.
Cervical cancer is a major killer in India, with 67,000 fatalities each year. Indeed, the country accounts for 25% of 260,000 yearly deaths from cervical cancer. Part of the problem is detections; if found early the cancer can be treated and/or prevented.
Microsoft and SRL developed an AI-powered screening program that provides faster analysis. Through cytoanalysis, the companies reviewed 100,000 Pap smear samples each year. In a release, the company's say 2% of the samples require further analysis because of abnormalities.
“We were looking for ways to ensure our cytopathologists were able to find those 2% abnormal samples faster”- Dr. Arnab Roy, Technical Lead for New Initiatives & Knowledge Management at SRL Diagnostics.
With AI, the technology can find discrepancies between analysis and “create a consensus on the areas assessed,” says Manish Gupta, Principal Applied Researcher at Microsoft Azure Global Engineering.
Microsoft and SRL Diagnostics trained the AI based on annotations of smear tests made by cytopathologists from several labs. The company explains how any discrepancies were moved to further review by trained personnel.
“The images for which annotations were found to be discordant — that is if they were viewed differently by three team members — were sent to senior cytopathologists for final analysis.”
Microsoft has now released the Cervical Cancer Image Detection API. This Azure-powered tool has numerous benefits, such as being able to decide between pathological cytological slides. Smear slides can also be compared across seven subtypes of cervical cytopathology.
“The API has the potential of increasing the productivity of a cytopathology section by about four times. In a future scenario of automated slide preparation with assistance from AI, cytopathologists can do a job in two hours what would earlier take about eight hours!” -Dr. Roy.