3D Model: Breast Cancer Invasion Biomechanics

Project Goal:

Develop a compact, mechanically reliable force-feedback drivetrain that prioritizes smooth power transmission, precise tolerancing, and rapid iteration for biomedical research applications.

Role & Responsibilities:

Built a computational pipeline for 3D tissue reconstruction and analysis
Processed in vitro imaging data across multiple experimental conditions
Quantified collagen fiber alignment using Fourier-based methods
Analyzed biomechanical trends related to tissue stiffness and matrix organization

Technical Skills: Non-technical Skills

MATLAB Experimental planning
3D computational modeling Research documentation
Image analysis (Fourier) Data interpretation
Quantitative data analysis Technical communication
Biomechanics analysis Independent research

Research Context:

Breast cancer invasion through white adipose tissue is strongly influenced by the mechanical and structural properties of the extracellular matrix (ECM), particularly collagen fiber organization. However, these biomechanical interactions are difficult to quantify in vivo due to limited spatial resolution and tissue deformation. To address this challenge, this research leverages 3D in vitro model systems that mimic adipose tissue using deformable polyacrylamide beads embedded in collagen matrices of varying stiffness and concentration.

3D Reconstruction from Detected Circular Features

Modeling & Analysis Approach:

3D Reconstruction of Adipose Tissue Analogs

Two-dimensional image slices were analyzed to detect circular polyacrylamide beads representing adipocytes. Detected circles with shared centers across adjacent slices were grouped and fitted into 3D spheres, enabling reconstruction of adipose-like tissue geometry. Image scaling and interpolation were used to preserve physical dimensions and spherical morphology.

Validation of Packing Fraction

Two-dimensional image slices were analyzed to detect circular polyacrylamide beads representing adipocytes. Detected circles with shared centers across adjacent slices were grouped and fitted into 3D spheres, enabling reconstruction of adipose-like tissue geometry. Image scaling and interpolation were used to preserve physical dimensions and spherical morphology.

Voronoi-Based Structural Analysis of the 3D Model

Fourier-Based Collagen Fiber Alignment Analysis in MATLAB

Collagen Fiber Alignment Analysis

Collagen fibers were analyzed independently using Fourier-based image analysis. This approach segmented fibers into local regions and quantified alignment values ranging from 0 (unaligned) to 1 (perfect alignment). Parameter tuning was performed to optimize window size and noise filtering for reliable alignment measurements.

Results & Insights:

Increasing collagen concentration at 1 kPa tissue stiffness resulted in a negligible increase in collagen alignment
Increasing collagen concentration at 7 kPa tissue stiffness led to a significant decrease in collagen alignment
Alignment trends were confirmed through statistical analysis of mean values and standard deviation overlap
Results suggest that collagen organization depends on the interaction between matrix stiffness and fiber density, rather than collagen concentration alone

Validation of Collagen Fiber Alignment via Statistical Variability

Statistical Distribution of Collagen Fiber Alignment

Limitations:

Fourier analysis segmented fibers into local regions rather than capturing full fiber paths
Image quality and background variability limited alignment accuracy in some samples
The study analyzed a limited number of samples, requiring further data to fully generalize trends
Polyacrylamide beads, while tunable, are an approximation of native adipose tissue

Research Outcome:

This work establishes a computational framework for linking tissue stiffness, collagen alignment, and cancer invasion pathways in three dimensions. The model provides a foundation for future studies incorporating real-time cell motion and improved ECM representation to better understand biomechanical drivers of breast cancer progression.

A more detailed account of this research in the paper and presentation below:

Nowyshin_Mridula_Effect_of_Collagen_Alignment_and_Adipocyte_Stiffness_on_Breas_Cancer_Invasion_through_In_Vitro_Model_Systems.pdf

NowyshinMridula_STARSSymp-FINAL_24Jul2024.pdf