University of British Columbia
Ultrasound-based approaches to tissue classification for breast and prostate cancer diagnosis
Abstract
dc:descriptionUltrasound-based cancer diagnosis could improve the current breast and prostate cancer diagnosis methods. In this thesis, an ultrasound-based approach is evaluated as a method for breast and prostate cancer diagnosis. Ultrasound RF time series along with a comprehensive machine learning framework is used for accurate classification of tissue samples. The RF time series method requires only a few seconds of raw ultrasound data with no need for additional instrumentation. The developed method produces cancer probability/likelihood maps which show the probability of the tissue under study being cancerous. These probability maps could provide radiologists with a real-time cancer diagnosis tool which could improve cancer yield and significantly reduce the number of negative biopsies. To prove the utility of ultrasound RF time series as a tissue classification method, an in vivo breast lesion classification study of 22 subjects and an in vivo prostate biopsy core classification study involving 18 subjects is presented in this thesis. A comprehensive machine learning framework with a new semi-supervised learning technique for tissue classification is also presented in this work. An experimental study to substantiate the ultrasound RF time series hypotheses by studying the effects of ultrasound imaging parameters on animal tissue classification is also presented. Using the ultrasound RF time series method and the developed machine learning framework–we calculated the area under the receiver operating characteristics curve to be 85.6% for breast lesion classification and 91.5% for prostate tissue classification. Increasing the frame rate and the length of the time series, and decreasing the imaging depth we observed consistent improvement in tissue classification results for the animal study. The results of this thesis suggest the potential of ultrasound RF time series as a tissue classification method. Ultrasound RF time series along with other ultrasound-based methods could be a valuable and practical addition to the current cancer diagnosis procedures. It has been shown here that a high level of accuracy can be attained using these tools which are non-invasive, inexpensive and readily available to the clinician.
Degree
thesis:*- Name thesis:degree_name
- Master of Applied Science - MASc
- Level thesis:degree_level
- master's
- Discipline thesis:degree_discipline
- Electrical and Computer Engineering
- Grantor dc:publisher
- University of British Columbia
- Year dc:date
- 2014
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Uniyal, Nishant
Rights
dc:rights- Statement dc:rights
-
- Attribution-NonCommercial-NoDerivs 2.5 Canada
- Language dc:language
- eng
Identifiers
dc:identifier.*- Handle dc:identifier
- http://hdl.handle.net/2429/50039