This research investigates the intersection of applied statistics, machine learning, and healthcare to leverage the quantitative approaches as a means of improving patient quality of life after surgery. Relevant patient data prior to surgery remains relatively limited. Therefore, finding the appropriate means to maximize the limited data available to optimize patient outcomes postoperatively is imperative. This study utilizes multiple patient cohorts, drawn from a tertiary care hernia referral center in the southeastern United States, who underwent abdominal wall reconstruction surgery. Chapter 2 focuses on developing a multivariable model using unique preoperative patient features to model the relationship between these variables and the outcome of interest: patient quality of life six months following abdominal wall reconstruction surgery. Using patient cohort data from the years 2005-2017, this study successfully built and internally validated a multivariable model using 20 unique preclinical variables. These findings provide further evidence to determine what preoperative variables reliably predict patient quality of life after surgery. Ultimately, this assists clinicians in preoperative assessments to optimize early patient interventions and treatment plans. Chapter 3 shifts to investigating alternate data sources in predicting patient outcomes, i.e., patient imaging data in the form of computed tomography scans. Since these data are scarce prior to surgery, this research focuses on assessing multiple qualitative methodologies to align and improve image quality for predictive modeling. From the various methodologies analyzed, image averaging after cropping and alignment showed the most promising means of optimizing preoperative patient images for patient quality of life classification. Using these techniques aids researchers in maximizing the limited image data available for building accurate classification models in surgery. Finally, Chapter 4 explores existing predictive models in abdominal wall reconstruction surgery to determine external generalizability on other patient cohorts. This research draws on the methods and techniques explored in the third chapter to optimize patient images to successfully train and validate these models. Although not initially successful in demonstrating model external validity on outside patients, investigation in pooled validation techniques suggests successful and generalizable models are possible with further investigation into matching internal and external patient cohorts. In conclusion, this research explores the possible applications of statistical and machine learning methods in surgery and provides