Material Information

Biomedical and Health Informatics Part I: Graduate Scholarship
SUNY Oswego
Bearce, Lindy ( Speaker )
Pacleb, Ryan ( Speaker )
Bichindaritz, Isabelle ( Speaker )
Publication Date:
Copyright Date:


Subjects / Keywords:
Quest 2021
Biomedical and Health Informatics


A Comparison of Regression Models: Performance and Prediction of Distant-Staged Colorectal Cancer in Populations Under Age 50 by Lindy Bearce. Colorectal Cancer (CRC) is the 2nd most common cause of cancer death in the United States. In populations under age 50, incidence and mortality rates have steadily been increasing. Regression modelling is essential to the field of cancer research as understanding interactions between co-variables is key to predicting patient survival. While the Cox Proportional Hazards Regression model is the most common used regression model for cancer survival data, machine learning models have proven to be equally effective. The aim of this study is to compare performance and predict outcomes utilizing different regression model methods with a focus on survival. The models chosen for this analysis are the popularized Cox Proportional Hazards Regression model, a Linear Regression model using Least Absolute Shrinkage and Selection Operator (LASSO) and a Random Survival Forest (RSF) ensemble machine learning method. The data for this study comes from the Surveillance, Epidemiology and End Results (SEER) database and consists of Colorectal Cancer cases, excluding Appendix, of Distant-Staged disease among non-pediatric patients under age 50 (ages 20-49) diagnosed during 2010. Data is preprocessed and modelled utilizing Python and R software packages. During data pre-processing, categorical data is transformed into binary values and the data is split into training and testing sets. The Concordance index (c-index) performance metric is calibrated utilizing variable importance and models analyzed utilizing additional performance indicators. Finally, the best performing model is used to predict survival outcomes on new data. ( ,,,,,,,,, )
Analysis on Departmental budget and Actual budget on of Canton-Potsdam Hospital by Ryan Pacleb. Canton-Potsdam Hospital is part of the St. Lawrence Heath System and is located in St. Lawrence County in the northern part of New York. From a rural hospital to a fast-growing, advanced, and sophisticated hospital in the northern area that provides topnotch quality healthcare to patients. To remain true to its mission and continue to provide clinical quality of healthcare. Canton-Potsdam Hospital needs to have visualization, systematic assessment, analysis, and optimization of its process and procedure in creating on its department budget and expenses. The absence of visualization and systematic assessment between actual budget which is the expenditures and budget which is the estimated revenues and expenses for the fiscal year makes department leadership incapable to evaluate and track spending, unable to give strategic management of hospital operation, and incapable to estimate possible future financial needs. This inability affects the performance of Canton-Potsdam hospital on delivering clinical standards. The aim of this study to evaluate the actual budget and estimated budget of Canton Potsdam Hospital. To identify the process of how every department creates a budget. Understand the procedure and different calculations to obtain the variance between the estimated budget and the actual budget. To be able to optimize the workflows, processes, and procedures in budget creation. Data Qualitative research method was used by the researcher for this study. The researcher will only focus on a method of making the department budget and department actual budget, the factors that affect the consideration of the department budget, and optimization on the budget process and workflow. The researcher used the data is available in the data repository of Canton-Potsdam hospital. The researcher pulled worth of 5 years of data from the data repository. The researcher will heighten the use of the existing software called Dimensional Insight that is used for reporting and creation of dashboards in other hospital operations. Dimensional Insight is an analytic software is now in the market and helping small and large business. Visualization of actual budget and an estimated budget of Canton Potsdam Hospital with Dimensional Insight help department leaders evaluate and track spending, give strategic management of hospital operation, simplify the decision-making process, identify obstacles, and capable to estimate possible future financial needs. This project marks the performance of Canton-Potsdam hospital on delivering clinical standards.
Session Chair: Isabelle Bichindaritz
Collected for SUNY Oswego Institutional Repository by the online self-submittal tool. Submitted by Zach Vickery.

Record Information

Source Institution:
SUNY Oswego Institutional Repository
Holding Location:
SUNY Oswego
Rights Management:
All applicable rights reserved by the source institution and holding location.

OswegoDL Membership

Added automatically


This item is only available as the following downloads: