X-rAID: Computer Aided Diagnosis of X-Ray Images

Radiologists use x-ray images of the human thoracic cavity to diagnose many diseases. One
of the most common diagnoses, an enlarged heart, is detected by measure the cardiothoracic
ratio – the ratio of the heart diameter to that of the thoracic cavity. Radiologists measure the
cardiothoracic ratio (CTR) manually. This process is time-consuming and error-prone. In this
project, we aim to build a computerized system to measure the cardiothoracic ratio
automatically from chest x-rays. To this end, efficient algorithms will be designed to segment
the x-ray image into its constituent organs and to detect important points necessary for
measurements. When complete, such a computer aided diagnosis (CAD) system will help
prevents errors, save time and reduce costs. If successful, the project may be extended to
allow diagnosis of other cardiothoracic diseases.

Faculty Supervisor:

Eugene Fiume

Student:

Partner:

Medical Teknologix

Discipline:

Computer science

Sector:

University:

University of Toronto

Program:

Accelerate