Early Detection of Lung Cancer Using Machine Learning: Creating Algorithms to Identify CT Scans of Lung Cancer Nodules
Principal Investigator
Name
Michael Blueglass
Institution
Yorktown High School
Position Title
Science Research Teacher
Email
About this CDAS Project
Study
NLST
(Learn more about this study)
Project ID
NLST-225
Initial CDAS Request Approval
Jun 14, 2016
Title
Early Detection of Lung Cancer Using Machine Learning: Creating Algorithms to Identify CT Scans of Lung Cancer Nodules
Summary
To Whom it May Concern,
My name is David Livadhi and I am a sophomore in Yorktown High School in Yorktown Heights, NY. Currently, I am enrolled in my school’s three year Science Research Program. The purpose of the program is to afford students the opportunity to participate in the community of scientific scholarship and research as a part of their high school experience. Further, a primary goal of the research program is to allow students early exposure to the skills necessary to conduct research as well as carry out a research project under the guidance of a professional mentor.
In previous articles, it was proven that the detection capabilities of an image processing algorithm would allow for earlier detection compared to current diagnostic methods. Additionally, I saw how small data sets hamper the robustness of these algorithms due to the lack of data. I plan on using the data you provide to train and improve accuracy of machine learning models. My research will be differ from previous studies because the increase in the data sample size will allow for more credible results, increased early detection and reduced false-positive rates. Current research indicates that machine learning algorithms are efficient in binary classification of lung cancer through analysis of CAT scans. The goal of this research is to design machine learning algorithms to classify CAT scans with the objective of early lung cancer diagnosis. Additionally, I would like to reduce the high false-positive rates, which may increase the use though out the medical community.
Overall, with your help, I hope to make a small difference in the field of lung cancer research. In the future, I would like to use image processing algorithms to further classify these images into clinical stages. This will improve treatment specificity, survival rates and quality of life for those diagnosed with lung cancer. Thank you in advance for your time and I look forward to hearing from you soon.
Sincerely,
David Livadhi
My name is David Livadhi and I am a sophomore in Yorktown High School in Yorktown Heights, NY. Currently, I am enrolled in my school’s three year Science Research Program. The purpose of the program is to afford students the opportunity to participate in the community of scientific scholarship and research as a part of their high school experience. Further, a primary goal of the research program is to allow students early exposure to the skills necessary to conduct research as well as carry out a research project under the guidance of a professional mentor.
In previous articles, it was proven that the detection capabilities of an image processing algorithm would allow for earlier detection compared to current diagnostic methods. Additionally, I saw how small data sets hamper the robustness of these algorithms due to the lack of data. I plan on using the data you provide to train and improve accuracy of machine learning models. My research will be differ from previous studies because the increase in the data sample size will allow for more credible results, increased early detection and reduced false-positive rates. Current research indicates that machine learning algorithms are efficient in binary classification of lung cancer through analysis of CAT scans. The goal of this research is to design machine learning algorithms to classify CAT scans with the objective of early lung cancer diagnosis. Additionally, I would like to reduce the high false-positive rates, which may increase the use though out the medical community.
Overall, with your help, I hope to make a small difference in the field of lung cancer research. In the future, I would like to use image processing algorithms to further classify these images into clinical stages. This will improve treatment specificity, survival rates and quality of life for those diagnosed with lung cancer. Thank you in advance for your time and I look forward to hearing from you soon.
Sincerely,
David Livadhi
Aims
I have two specific goals my project will be investigating.
1) To train and improve accuracy of machine learning models with the hope to reduce false-positives
2) Provide a method for early detection in both small cell lung cancer (SCLC) and non-small cell lung cancer (NSCLC)
Collaborators
David Livadhi (High School Student at Yorktown High School)