INIFY Prostate® shown to accurately outline suspicious prostate cancer in study with two prominent medical centers

STOCKHOLM – October 18, 2021 – ContextVision, a medical technology software company specializing in image processing, image analysis and decision support tools for digital pathology, today announced that results from a US clinical evaluation were presented at the Pathology Visions Congress in Las Vegas, USA.

The evaluation was performed together with two independent pathologists, Anil Parwani, M.D., Ph.D., M.B.A. at Ohio State University Comprehensive Cancer Center (Columbus, Ohio), and Ming Zhou, M.D., Ph.D., Chair and Pathologist-in-Chief at Tufts Medical Center (Boston, Massachusetts). It showed that for images with small focus of cancer (≤1 mm), the INIFY algorithm identified all cancer glands and did not miss any. The sensitivity was 100% in these cases. Also, for images with larger cancer areas, outlining was very precise, with a sensitivity of 99.3 – 99.8%. For a decision support tool to be an effective companion in the clinical workflow, it is important that it does not identify too many false positives. In this evaluation, INIFY Prostate showed a high specificity of 96 – 98%.

“Many pathologists find signing out benign cases to be accompanied by the anxiety of having missed something important. Decision support tools that are reliable in finding even the smallest cancer foci can play an important role in increasing diagnostic confidence,” says Dr. Zhou.

Prostate biopsy slides from 30 unique patients from both Wexner and Tufts medical centers were scanned using three different scanners. Drs. Parwani and Zhou reviewed INIFY Prostate’s predictions of suspicious cancer areas in all scanned slides. The aim of the evaluation was to see whether the product’s performance differed depending on pathology lab (i.e., variations in slide preparation such as staining) or scanners used. It was concluded that INIFY Prostate works equally well on images from both labs, rendering very similar outcomes in terms of sensitivity and specificity. Thus, the INIFY predictions are robust.

“This is the first quality assurance study that has compared the performance of a prostate algorithm on images scanned on three different scanners side by side,” says Dr. Parwani.

INIFY is a powerful AI-based software that precisely outlines suspicious cancer areas in prostate biopsies, with unique, detailed pixel-level focus – a result of MasterAnnotation, a patented annotation method used in training of the algorithms.

“This result adds to several studies that confirms INIFY’s ability to support the pathologists in their daily clinical work,” says Magnus Aurell, VP Business Unit Digital Pathology at ContextVision.

###

For further information, please contact ContextVision’s CEO, Fredrik Palm, at Fredrik.palm@contextvision.com or visit www.contextvision.com.

###

About ContextVision ContextVision is a medical technology software company specialized in image analysis and artificial intelligence. As the global market leader within image enhancement, we are a trusted partner to leading manufacturers of ultrasound, X- ray and MRI equipment around the world.

Our expertise is to develop powerful software products, based on proprietary technology and artificial intelligence for image-based applications. Our cutting-edge technology helps clinicians accurately interpret medical images, a crucial foundation for better diagnosis and treatment.

ContextVision is now entering the fast-growing digital pathology market. We are re-investing significantly in our product portfolio of decision support tools, and we are dedicated to becoming a leading resource for pathologists to radically develop cancer diagnosis and improve patient care.

The company, established in 1983, is based in Sweden with local representation in the U.S., Japan, China and Korea. ContextVision is listed on the Oslo Stock Exchange under the ticker CONTX.

This information is subject to the disclosure requirements pursuant to Section 5-12 the Norwegian Securities Trading Act

Attachment