Development of a system for the automatic recognition and classification of normal and abnormal cells in human blood samples

Automation of medical diagnosis/detection process is very important in terms of enhancing diagnostic accuracy, increasing throughput, reducing costs, and training new staff. Our current goal is to go from the proof of concept stage (automatic recognition and classification of human blood images) to a complete working and optimized prototype and to start testing it in an actual clinical lab environment with the help of Calgary Laboratory Services. The prototype design will take into account user friendliness, high throughput, robustness, integration with existing laboratory workflow and reasonable cost. Our goal is to decrease the image processing time to 0.1 seconds by code optimizations and parallel programming and also increase the cell classification accuracy to be more than 95% which is an acceptable rate in comparison with manual processing.

Faculty Supervisor:

Behrouz Far

Student:

Partner:

Smart Labs Ltd

Discipline:

Engineering

Sector:

Professional, scientific and technical services

University:

University of Calgary

Program: