Prostate Cancer Prognosis Technology

Aureon's proprietary Systems Pathology platform integrates molecular and cellular biology, advanced image analysis and sophisticated mathematics to identify and measure clinical, micro-anatomical and molecular features which predict specific individual patient clinical outcomes.

The underlying foundation of Aureon's breakthrough technology is comprised of:

Aureon PathoMetrix™

PathoMetrix is an automated machine vision tissue image analysis system that uses advanced image processing algorithms in order to segment and measure properties of histopathological objects.

Currently, pathologists rely on the Gleason scoring system to evaluate cancer aggressiveness. The tumor is graded based on its architecture patterns as observed under a light microscope. The more aggressive the cancer, the higher its Gleason score.

(Click Image to Enlarge)

Although Gleason grading is widely considered by pathologists to be reliable, it is a subjective scoring system. The pathologist's experience, as well as other factors, can considerably affect the final Gleason scores. Moreover, the Gleason score is categorized into discrete levels to make it easier for pathologists to use. Meanwhile, new techniques in digital image processing and analysis have emerged, aided by significantly increased computational power. To date, most cancer image analysis systems have been developed for images taken from cytological specimens, which only capture cells and thus do not utilize all of the architectural information available at the tissue level. However, the structure of different pathological elements at the tissue level plays a more important role in diagnosis than the appearance of individual cells. For example, in a prostate tissue specimen, the shape and size of the gland are two of the most critical features pathologists use to determine the progression of the cancer.

PathoMetrix is one of the foundation technologies of Prostate Px and allows for the analysis of prostate digital tissue images. The features extracted from the images are then integrated with clinical and molecular information to assess disease severity and predict disease progression at diagnosis.

(Click Image to Enlarge)

Aureon M-Plex™

Multiplexed in situ protein detection consists of strategies developed at Aureon to fluorescently tag individual antibodies and then selectively analyze each [antibody: antigen] complex with spectral imaging. Our automated high-throughput imaging system allows the separation of real signal from tissue auto-fluorescence, enabling a more sensitive and quantitative measurement of biomarkers. Furthermore, an elaborate quality control system has been implemented to guarantee the day-to-day reproducibility. The captured images are processed with the immunofluorescence image technologies developed within PathoMetrix to create features representing the cellular identification and quantitative attributes from the respective biomarker profiles. Individual feature elements are then included in the predictive model and selected based on the clinical outcome.

(Click Image to Enlarge)

Aureon Discovery-Path™

Aureon has developed a novel methodology that incorporates the concepts of machine learning into the arena of clinical outcomes prediction. Machine learning is an area of artificial intelligence, which is akin to how the human brain works, and whose engineering can be explained in a statistical framework. With this methodology, we now have the ability to provide a tool that can learn from a patient's data and improve our ability to provide an accurate prognosis for patients. Discovery-Path is a supervised multivariate analytic toolbox capable of computing an optimized model to predict clinical outcomes.