Design and development of a computer vision-based breast self-examination instruction and supervision system
Breast cancer is the most prominent cause of cancer-related deaths among women globally and is the 2nd leading cause of cancer-related deaths in the Philippines. Detecting cancer early significantly affects its curability. Breast self-examination (BSE) is a cost-effective and non-invasive method whi...
Saved in:
| Main Author: | |
|---|---|
| Format: | text |
| Language: | English |
| Published: |
Animo Repository
2014
|
| Online Access: | https://animorepository.dlsu.edu.ph/etd_masteral/4760 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | De La Salle University |
| Language: | English |
| Summary: | Breast cancer is the most prominent cause of cancer-related deaths among women globally and is the 2nd leading cause of cancer-related deaths in the Philippines. Detecting cancer early significantly affects its curability. Breast self-examination (BSE) is a cost-effective and non-invasive method which can be performed by women themselves to monitor their breasts and detect the presence of lumps which can be cancerous. BSE is difficult to master especially without supervision and guidance from a professional. In response to this issue, this thesis aims to develop a computer vision-based BSE instruction and supervision system which can monitor and guide women while they perform BSE. The proposed system requires just a personal computer and a simple webcam. The proposed system is comprised of four major computer vision algorithms. The first one is devised to identify the entire areas of the left and right breasts using extensive use of color features, edges, contours, and curves. The second algorithm is devised to divide the breast area into smaller blocks, which the use will be asked to examine consecutively, using integral images and genetic algorithm. The third algorithm is devised to track the hand of the user using corner features and sparse optical flow. The fourth algorithm is devised to detect palpation motions using time series regression analysis. These algorithms were integrated together to create a complete BSE instruction and supervision with additional features such as real-time visual and audio feedback and a custom graphical user interface. This research aims to improve the proficiency of women in performing BSE through the use of the proposed interactive system in order to increase breast awareness among women and ultimately, if possible, reduce breast cancer mortality rate. |
|---|