Skip to main content
LATEST INFORMATION ABOUT COVID-19

READ MORE

notice

Doctoral Thesis Defense: Muneera Alsaedi

July 31, 2019

Speaker: Muneera Alsaedi

Supervisors: Drs. T. Fevens, A. Krzyzak

Examining Committee: Drs. Maria Amer, T. Glatard, C. Laporte, C. Y. Suen,
A. R. Sebak (Chair)

Title: Computer-aided Cytological Grading Systems for Fine Needle Aspiration Biopsies of Breast Cancer

Date: Wednesday, July 31, 2019

Time: 13:00

Place: EV 11.119

ABSTRACT

According to the American Cancer Society, breast cancer is the world's most commonly diagnosed and deadliest form of cancer in women. A major determinant of the survival rate in breast cancer patients is the accuracy and speed of the malignancy grade determination. This thesis considers the classification problem related to determining the grade of a malignant tumor accurately and efficiently. A Fine Needle Aspiration (FNA) biopsy is a key mechanism for breast cancer diagnosis as well as for assigning grades to malignant cases. Carrying out a manual examination of FNA demands substantial work from the pathologist which may result in delay, human errors, and misclassified grades. In this thesis, to grade FNA images of breast cancer, six known cytological grading schemes and three novel cytological grading schemes were proposed for FNA biopsies of breast cancer based on handcrafted feature extraction and traditional machine learning algorithms. In addition, five grading systems based on convolutional neural networks (CNN) were proposed to automatically determine the malignancy grade of breast cancer. Further, the proposed systems were combined with data balancing techniques to improve the sensitivity prediction for G3 cases due to the class imbalanced problem. The proposed systems were able efficiently to classify FNA slides into G2 (moderately malignant) or G3 (highly malignant) cases. In terms of traditional machine learning algorithms, for case classification, the best results were obtained for computer-
aided CGSs based on the modified Khan et al’s and Robinson’s schemes. Meanwhile, for patient classification, the overall best results were obtained for computer-aided CGSs based on the modified Khan et al’s and modified Fisher's schemes. On the other hand, in terms of CNN models, out of the five different CNN models used in this thesis, the combination of GoogleNet Inception-v3 and the oversampling method provides the best accuracy performance. The obtained results demonstrate that computer-aided breast cancer cytological grading systems using FNA can potentially achieve accuracy rates comparable to the more invasive histopathological BR-method.




Back to top Back to top

© Concordia University