Automated detection of basal cell carcinoma using polarization sensitive optical coherence tomography
Abstract/Contents
- Abstract
- According to recent estimates, one in five Americans develop skin cancer in their lifetime. Basal cell carcinoma (BCC) accounts for more than 80% of total incidences of skin cancer, and while not fatal, it can grow to cause significant disfigurement and morbidity if left undiagnosed and untreated. In the current workflow of the US medical system, visual assessment of suspicious skin lesions by a primary care physician (PCP) serves as the basis for referring a patient to a dermatologist for positive determination of BCC. Biopsy and subsequent histology still remain the gold standard for skin cancer diagnostic in dermatology offices. The current medical workflow for detecting BCC suffers from the subjectivity and potential error of visual evaluations by PCPs, the invasiveness of biopsies, and the time-consuming nature of histology-based evaluations. Therefore, there is a significant clinical need for a non-invasive, real-time, automated and reliable technique for BCC detection. In this dissertation we report the development of highly sensitive classifiers, which successfully detected BCC in ex-vivo mouse and human skin. In ex-vivo mouse skin the classifier demonstrated 94.4% sensitivity and 92.5% specificity, while in ex-vivo human skin the classifier achieved 95.4% sensitivity and specificity. Our classifiers utilize features extracted from a large number of images captured by polarization-sensitive optical coherence tomography (PS-OCT), which is a form of non- invasive infrared imaging, paired with machine learning techniques that fully automate the detection process. Our proposed classifiers could be used in PCPs' and dermatologists' offices for real-time and automated detection of BCC in intact skin.
Description
Type of resource | text |
---|---|
Form | electronic; electronic resource; remote |
Extent | 1 online resource. |
Publication date | 2017 |
Issuance | monographic |
Language | English |
Creators/Contributors
Associated with | Marvdashti, Tahereh |
---|---|
Associated with | Stanford University, Department of Electrical Engineering. |
Primary advisor | Bowden, Audrey, 1980- |
Thesis advisor | Bowden, Audrey, 1980- |
Thesis advisor | Solgaard, Olav |
Thesis advisor | Tang, Jean (Jean Y.) |
Advisor | Solgaard, Olav |
Advisor | Tang, Jean (Jean Y.) |
Subjects
Genre | Theses |
---|
Bibliographic information
Statement of responsibility | Tahereh Marvdashti. |
---|---|
Note | Submitted to the Department of Electrical Engineering. |
Thesis | Thesis (Ph.D.)--Stanford University, 2017. |
Location | electronic resource |
Access conditions
- Copyright
- © 2017 by Tahereh Marvdashti
- License
- This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).
Also listed in
Loading usage metrics...