by John R. Fischer
, Senior Reporter | March 26, 2018
Creating and training algorithms to analyze and infer cancer diagnoses from tissue samples can now be done with no programming experience, following the launch of Fimmic’s Aiforia Cloud.
The AI imaging analysis cloud software, formerly known as WebMicroscope, enables pathologists to train convolutional neural networks (CNNs) to detect and quantify specific findings in tissue images with self-service and automatized pathology image analysis capabilities. The solution, renamed in tribute to these newly added AI features, is designed to quicken the diagnosis process and production of treatments.
“With a rising prevalence of cancer – and an aging population overall – the health care industry faces a growing number of tissue samples to analyze in order to make diagnoses and lifesaving drugs,” Kaisa Helminen, CEO of Fimmic, told HCB News. “The technology for digital pathology exists today, making sample management and viewing easier. But deep learning AI opens up an entirely new window of opportunity for the industry by enabling fast and consistent sample analysis, even for complex tasks that have not been possible to automatize with conventional machine vision methods.”
Nearly 22 million new cases of cancer are expected to occur by 2030, placing greater pressure on health care systems and physicians to quickly and correctly analyze tissue samples for cancer diagnoses and develop lifesaving drugs.
The SaaS platform is equipped with Aiforia Create tools, a feature that enables pathologists with little to no programming skills to train CNNs and produce algorithms to perform laborious image analysis tasks in less time, requiring no investments in hardware or software.
Users can deploy and use Aiforia Cloud through a browser with training taking place through an easy to use cloud service.
The use of the software enables algorithm creation to take place in a matter of days, compared to months with conventional machine learning approaches, and allows pathologists to devote more time to high-value work, such as the analysis of complex or rare samples.
“AI stands to transform the industry at large, but can often feel intimidating to those who don’t have a technical background,” said Helminen. “We believe in empowering users with powerful, yet simple, technology, which is why the self-service capabilities are so revolutionary. Moreover, pharmaceutical research and medical research benefit the most from this tool since they typically work with rapidly changing applications and image analysis needs, and have large sample collections. With our tool, they can design new algorithms for any image analysis tasks they may ever wish and obtain quantitative information from samples, enabling completely new research approaches.”
The solution is compatible with all 2-D images and can be used for purposes outside of microscopy.
Fimmic is currently demonstrating the uses of Aiforia Cloud this week at the USCAP Annual Meeting in Vancouver, Canada.