Dissertation defence (Computer Science): MSc Ileana Montoya Perez
Time
17.5.2024 at 12.00 - 16.00
MSc Ileana Montoya Perez defends the dissertation in Computer Science titled “Data Analysis with Limited Data Availability: Prostate Cancer Prediction and Characterization as a Case Study” at the University of Turku on 17 May 2024 at 12.00 (University of Turku, Agora, Lecture hall XXI, Turku).
The audience can participate in the defence by remote access: https://echo360.org.uk/section/aed5b5e1-0864-461a-953d-ec0263b5ce21/public
Opponent: Professor Sébastien Lafond (Åbo Akademi University)
Custos: Professor Tapio Pahikkala (University of Turku)
Doctoral Dissertation at UTUPub: https://urn.fi/URN:ISBN:978-951-29-9646-9
***
Summary of the Doctoral Dissertation:
In this research, we address the challenge of analyzing limited amount of data in prostate cancer studies. We used appropriate statistical and machine learning methods to generate results that help in the diagnosis and prediction of prostate cancer. We found that features extracted from MRI scans show potential for classifying prostate tumors into low- and high-risk categories. Furthermore, we presented empirical evidence of the value that features from MRI scans along with clinical and genes data have in detecting and classifying prostate cancer. To ensure the reliability of results, we introduced a novel cross-validation method designed for small datasets, which proves to be more dependable for ROC analysis than traditional cross-validation techniques.
The impact of this research is significant for both medical professionals and researchers dealing with prostate cancer and limited data. By enhancing our understanding of MRI, clinical and genetic data from prostate cancer patients, we can better identify and treat prostate cancer, leading to improved patient outcomes. Additionally, the development of a more reliable cross-validation method can benefit researchers in various fields working with small datasets, helping them obtain more accurate results.
The audience can participate in the defence by remote access: https://echo360.org.uk/section/aed5b5e1-0864-461a-953d-ec0263b5ce21/public
Opponent: Professor Sébastien Lafond (Åbo Akademi University)
Custos: Professor Tapio Pahikkala (University of Turku)
Doctoral Dissertation at UTUPub: https://urn.fi/URN:ISBN:978-951-29-9646-9
***
Summary of the Doctoral Dissertation:
In this research, we address the challenge of analyzing limited amount of data in prostate cancer studies. We used appropriate statistical and machine learning methods to generate results that help in the diagnosis and prediction of prostate cancer. We found that features extracted from MRI scans show potential for classifying prostate tumors into low- and high-risk categories. Furthermore, we presented empirical evidence of the value that features from MRI scans along with clinical and genes data have in detecting and classifying prostate cancer. To ensure the reliability of results, we introduced a novel cross-validation method designed for small datasets, which proves to be more dependable for ROC analysis than traditional cross-validation techniques.
The impact of this research is significant for both medical professionals and researchers dealing with prostate cancer and limited data. By enhancing our understanding of MRI, clinical and genetic data from prostate cancer patients, we can better identify and treat prostate cancer, leading to improved patient outcomes. Additionally, the development of a more reliable cross-validation method can benefit researchers in various fields working with small datasets, helping them obtain more accurate results.
University Communications