Document
About this Digital Document
{"value":"
Osteoarthritis (OA) represents one of the most pervasive degenerative joint diseases worldwide, characterized by pain, swelling, and progressive functional decline. Current therapeutic approaches largely focus on symptom management rather than the underlying biological mechanisms driving tissue degeneration. Within this therapeutic landscape, bone marrow aspirate concentrate (BMAC) has emerged as a promising orthobiologic, capable of leveraging the body’s own regenerative and anti-inflammatory potential. Composed of platelets, growth factors, cytokines, and mesenchymal stem cells (MSCs), BMAC functions as a point-of-care biologic that simultaneously delivers multiple therapeutic agents within a single injectable. Despite its widespread clinical use, however, the proteomic underpinnings that define BMAC’s therapeutic efficacy remain poorly characterized within the literature.This project seeks to address that gap by combining the principles of proteomics and data science to explore BMAC as both a biological and computational system. From a biological standpoint, the composition of BMAC varies according to patient specific factors such as age, sex, body mass index (BMI), smoking status, and the anatomical site of harvest in addition to procedural differences. From a computational standpoint, these variations manifest as a high-dimensional dataset in which biological patterns are often non-linear, multivariate, and subtle. The central premise of this work is therefore that data-driven modeling provides a necessary and scalable framework to uncover the latent biological structure that traditional univariate analyses often struggle to capture.
The long-term goal of this project is to establish a proteomic and analytical framework that enables clinicians and researchers to characterize BMAC in a reproducible, quantitative, and accessible manner. The short-term objective is to examine whether BMAC samples derived from different anatomical regions, specifically the iliac crest and humeral head, exhibit distinct proteomic profiles that are further modulated by patient demographics. To test this hypothesis, BMAC samples were analyzed using high-throughput antibody microarray technology, allowing for semi-quantitative measurement of 109 biomarkers across cytokines, chemokines, and growth factors. The dataset produced from these analyses forms the foundation for downstream inferential statistics and machine learning.
Within this framework, probabilistic reasoning, statistical inference, and machine learning converge to create a rigorous and reproducible method for disentangling variability within patient-derived biologics. The application of these techniques enables the identification of latent protein networks and interaction signatures that might otherwise remain obscured in conventional univariate analyses. These computational approaches do not replace biological interpretation but rather extend it, providing a scaffold upon which biological meaning can be quantified and compared across diverse patient populations and clinical contexts.
The implications of this work are therefore twofold. Biologically, it provides one of the most comprehensive proteomic characterizations of BMAC to date, revealing how anatomical source and patient demographics contribute to the underlying protein composition of this therapeutic.Ultimately, this work seeks to make the study of BMAC transparent, quantitative, and clinically meaningful. By embedding the analysis of biologics within a data science framework, this work contributes to the growing field of personalized regenerative medicine.
","attr0":"abstract"}
Full Title
Leveraging Data Science to Aid Clinicians in Understanding Bone Marrow Aspirate Concentrate as a Therapeutic
Member of
Contributor(s)
Creator: Herna, Colin
Thesis advisor: Jedlicka, Sabrina
Date Issued
2026
Language
English
Type
Genre
Department name
Bioengineering
Media type
Subject (LCSH)
Citation
@mastersthesis{herna2026,
title = {Leveraging Data Science to Aid Clinicians in Understanding Bone Marrow Aspirate Concentrate as a Therapeutic},
author = {Herna, Colin},
year = {2026},
keywords = {BMAC, Data Science, Microarray, statistics, Support vector machine, SVM},
abstract = {{"value":"Osteoarthritis (OA) represents one of the most pervasive degenerative joint diseases worldwide, characterized by pain, swelling, and progressive functional decline. Current therapeutic approaches largely focus on symptom management rather than the underlying biological mechanisms driving tissue degeneration. Within this therapeutic landscape, bone marrow aspirate concentrate (BMAC) has emerged as a promising orthobiologic, capable of leveraging the body’s own regenerative and anti-inflammatory potential. Composed of platelets, growth factors, cytokines, and mesenchymal stem cells (MSCs), BMAC functions as a point-of-care biologic that simultaneously delivers multiple therapeutic agents within a single injectable. Despite its widespread clinical use, however, the proteomic underpinnings that define BMAC’s therapeutic efficacy remain poorly characterized within the literature.This project seeks to address that gap by combining the principles of proteomics and data science to explore BMAC as both a biological and computational system. From a biological standpoint, the composition of BMAC varies according to patient specific factors such as age, sex, body mass index (BMI), smoking status, and the anatomical site of harvest in addition to procedural differences. From a computational standpoint, these variations manifest as a high-dimensional dataset in which biological patterns are often non-linear, multivariate, and subtle. The central premise of this work is therefore that data-driven modeling provides a necessary and scalable framework to uncover the latent biological structure that traditional univariate analyses often struggle to capture. The long-term goal of this project is to establish a proteomic and analytical framework that enables clinicians and researchers to characterize BMAC in a reproducible, quantitative, and accessible manner. The short-term objective is to examine whether BMAC samples derived from different anatomical regions, specifically the iliac crest and humeral head, exhibit distinct proteomic profiles that are further modulated by patient demographics. To test this hypothesis, BMAC samples were analyzed using high-throughput antibody microarray technology, allowing for semi-quantitative measurement of 109 biomarkers across cytokines, chemokines, and growth factors. The dataset produced from these analyses forms the foundation for downstream inferential statistics and machine learning. Within this framework, probabilistic reasoning, statistical inference, and machine learning converge to create a rigorous and reproducible method for disentangling variability within patient-derived biologics. The application of these techniques enables the identification of latent protein networks and interaction signatures that might otherwise remain obscured in conventional univariate analyses. These computational approaches do not replace biological interpretation but rather extend it, providing a scaffold upon which biological meaning can be quantified and compared across diverse patient populations and clinical contexts.The implications of this work are therefore twofold. Biologically, it provides one of the most comprehensive proteomic characterizations of BMAC to date, revealing how anatomical source and patient demographics contribute to the underlying protein composition of this therapeutic.Ultimately, this work seeks to make the study of BMAC transparent, quantitative, and clinically meaningful. By embedding the analysis of biologics within a data science framework, this work contributes to the growing field of personalized regenerative medicine. ","attr0":"abstract"}},
language = {English},
}