On-Demand Digital Pathology
OptraScan®
®
Affordable, Subscription-based System
Artificial Intelligence & Machine Learning based System for accurate, rapid and reproducible analysis of Prostate Cancer 4 3 5
Examination of histological specimens under the microscope by a pathologist is one of the most reliable methods used in detection of prostate cancer. This is carried out by examining the glandular architecture of the specimen by the most common method for histological grading of prostate tissue - the Gleason Grading System.
2
The cancer tissue is classified from 1 to 5 grades; however, in the recent times, this common method is found to be ineffective, reason being: 1
Ø Analysis on visual interpretation lacks reproducibility Ø It is limited by intra- and inter-pathologist variability
Our Machine-based scoring algorithms Our solutions to resolve the challenges appearing from Gleason Grading : Ø Fully automated solution : End to end solution
Gleason score 3+3=6. Grade 1 Gland Formation: Discrete, well formed, uniform large glands arranged back to back Legend:
Lumen
Epithelial nuclei
Epithelial cell cytoplasm
with robust and efficient algorithm modules. m Intelligent Segmentation module that works
on human perceptible color spaces to detect cell nuclei based on recognizable patterns like area, shape, intensity etc. m Automatic detection of glandular lumens
based on the clustering of identified cell nuclei and other features. Gleason score 4+4=8. Grade 4
m Robust feature extraction module to extract
structural, morphometric, texture, nucleocytoplasmic ratio and color features for detected cell nuclei and identified glandular regions.
Gland Formation: Fused, cribriform, poorly formed glands, punched out lumens Legend:
Lumen
Epithelial nuclei
Epithelial cell cytoplasm
Ø ANN (artificial neural network) based classifier : m Feature fusion and feature ranking
techniques for representation to the Neural network based classifier. m The classifier is trained to distinguish
between moderately and poorly differentiated glands. m Object level tumor grading is done using
feature characteristics for malignant and benign cell nuclei like mean intensity, area, standard deviation of intensity etc. Ø Key Differentiator : m Easily retrainable machine learning system. m High classification accuracy.
Result: Gleason Score 5+5=10. Grade 5 Gland Formation: lacks gland formation, Solid sheet of uniform neoplastic cells Legend:
Epithelial nuclei
OptraScan®
OS-15 15-slide brightfield
®
On-Demand Digital Pathology Solutions
OS-120
OS-FS
OS-FL
120-slide brightfield
7-slide frozen sections, with live view mode
15-slide fluorescence, with 6 filter cubes
IMAGEPath® Web-based Image Management and Viewing
TELEPathTM Web and Mobile Digital Conferencing
OptraASSAYSTM On-Demand Image Analysis
CLOUDPath® Laboratory Information Management System
100 Century Center Court, Suite 410, San Jose, CA 95112 OptraSCAN is an ISO13485 certified company
*All OptraSCAN systems and solutions are for research use only