Aiforia Announces CE-IVD Mark For AI-supported Breast Cancer Diagnostics

CE-IVD marked Aiforia™ Clinical AI Model for Breast Cancer: Ki67

The CE-IVD marked AI model for Ki67 is the first model to be released as part of the Aiforia Clinical Suite for Breast Cancer, supporting pathologists in cancer diagnosis.

First-of-its kind diagnostic tool

Aiforia today announced the CE-IVD Mark for the Aiforia™ Clinical AI Model for Breast Cancer: Ki67, a widely used key biomarker in breast cancer diagnostics across pathology labs around the world. The AI Model is able to calculate Ki67 positive and negative cells specifically in invasive areas of human breast tissue in whole slide images (WSI) as well as in specific areas — making it a first-of-its kind diagnostic tool for pathologists to harness the full potential of their own expertise and make decisions with confidence.

“The Ki67 AI Model is intended to support the pathologist by enhancing confidence in their decision making and precision diagnostics when evaluating this key biomarker in breast cancer,” explains Dr. Juuso Juhila, Director of Clinical Products at Aiforia “The AI model enables the pathologist to focus on the most significant areas, the invasive ones, thus increasing the speed and accuracy in the analysis of these key breast cancer biomarkers. This is crucial to the diagnostic process for pathologists and patients alike.”

The challenge: rising rates of disease and analog tools

Rising rates of cancer create a colossal burden on pathologists, an already underrepresented group of healthcare professionals. In the US, the proportion of physicians specializing in pathology has nearly halved while their workload has increased by about 42%. Meanwhile tools at their disposal are manual and error prone. Both pathologists and patients deserve reliable and efficient tools for clinical use.

The solution: AI-assisted data and decision support

The Aiforia Clinical Suites are under development to support the diagnosis of some of the most prevalent cancers such as breast, lung, prostate, and more. They are intended for release and regulatory filings. The CE-IVD marked Aiforia Clinical AI Model for Breast Cancer: Ki67 is the first AI model to be released for diagnostic support as part of the intended upcoming release of the full Aiforia Clinical Suite for Breast Cancer.

The Aiforia Science Team, fueled by their pathology and medical expertise, is underway in developing Clinical AI Models for diagnostic support in breast cancer, including AI models for PR/ER, the more technically demanding HER2 marker and mitosis count. The intention then is to CE-IVD mark/register these along with the rest of the Aiforia Clinical Suite for Breast Cancer.

The Clinical Suites are intended to provide pathologists with a portfolio of tools to increase the efficiency and accuracy of sample review: the AI models for diagnostic support as well as for use in pre- and post-analysis screening, a case-centric browser-based viewer and a platform for quality control (QC) and flexible adaptation.

“Our aim is to empower pathologists with the means to make decisions with confidence, while the responsibility for the clinical report is within the pathologist’s remit,” explains Dr. Juuso Juhila, Director of Clinical Products at Aiforia “The AI-assisted tools for clinical diagnostics are intended to be a means to empower the pathologist in harnessing their own expertise by automating manual tasks, providing results with higher speed and consistency and ultimately helping to improve patient outcomes.”

Clinical AI Model for Breast Cancer: Ki67

The Aiforia Clinical AI Model for Breast Cancer: Ki67 for diagnostic support, was created to address current challenges in cancer diagnostics. Breast cancer biomarker scoring methods are prone to variability. For example, commonly used scoring schemes necessitate pathologists to extract quantitative parameters that are subject to inter- and intra-observer variability. This unfortunately means that analysis tends to be more qualitative than quantitative.

The Ki67 AI Model is easily accessed via Aiforia’s browser-based viewer

The need for objective and standardized tools is evident. Deep learning artificial intelligence excels at performing these types of repetitive, quantification tasks — unaffected by bias and variability. Analyses with AI can be replicated more consistently, without human condition impairing its performance. AI models such as the Ki67 have the potential to enhance biomarker detection by assisting pathologists in speed, accuracy and improving workflow efficiencies.

”Biomarkers such ER, PR, HER2, and Ki67 are not like other immunostains; a binary interpretation of positive or negative is not enough. These biomarkers have prognostic significance and drive the clinical decision on which patient can benefit from targeted therapy. Accurate and reproducible biomarker quantification can be powered by AI to improve the quality and efficiency of pathology and ultimately optimize patient care,” explains Dr. Marilyn Bui, Professor of Pathology and Scientific Director at the Moffitt Cancer Center.

Intended for use in digitized breast cancer samples stained with Ki67/MIB antibody, the Aiforia AI Model is intended to provide a wide range of applications and benefits including

• Creates a heat map assisting the pathologist in automatically finding the critical areas of the sample
• Rapidly distinguishes between normal and cancerous tissue supporting the pathologist in the quick detection of tumor epithelium
• Automatically scores the Ki67 positive and negative cells of epithelial origin from either whole slide images (WSI) or selected image areas, saving the pathologist’s time
• Provides intelligent viewing as the Aiforia Clinical AI Model for Breast Cancer: Ki67 enables viewing and selection of areas with high density of Ki67-positive cells, or hotspots, displaying the results of image analysis

The future of AI-assisted cancer diagnostics

Pathologists deserve the best assistance possible to harness the full power of their own expertise. After all, effective treatment begins with a precise diagnosis. AI-assisted image analysis has the potential to expand the possibilities of cancer therapeutics, bringing about more effective treatment and personalized therapies and  resulting in cost savings to healthcare systems. Aiforia’s mission is to provide the highest quality solutions for AI-assisted image analysis to enable personalized and democratized care for patients.

Source: Aiforia