The project focuses on analyzing drug efficacy in murine models using computational techniques. It involves segmenting specific regions where the drug interacts with biological tissues, typically through image processing or data analysis, to identify areas of drug activity. Additionally, the project quantifies the drug concentration in these regions, providing insights into dosage impact and efficacy. This work combines biological research with computational tools to enhance understanding of drug behavior in preclinical studies.

Developed an end-to-end image processing pipeline for extracting, segmenting, and quantifying luminescent signals from fluorescence-based small animal imaging datasets. The pipeline processes raw .TIF images and associated metadata files to generate ROI-level measurements suitable for computational analysis and machine learning applications.

Parsed structured metadata files (AnalyzedClickInfo.txt, IR label.txt) to extract ROI coordinates, imaging parameters, radiometric measurements, and experimental annotations.

Programmatically reconstructed ROI geometries (center position, dimensions, and shape) and generated visual overlays for validation and quality control.

Normalized high-bit-depth grayscale .TIF images to preserve radiometric integrity while enabling accurate visualization and downstream processing.

Applied transparent ROI masks and overlays to facilitate visual inspection of segmented luminescent regions.

Associated radiance and flux measurements with corresponding ROIs, creating structured datasets for quantitative analysis and predictive modeling.

Automated the complete preprocessing workflow, reducing manual effort and enabling scalable dataset generation for machine learning tasks, including segmentation and regression.

Python

tifffile

Pillow (PIL)

pandas

matplotlib

Fluorescence imaging analysis

Luminescence quantification

ROI-based signal extraction

Flat-field correction

Bias subtraction

Scientific image preprocessing

Dataset preparation for machine learning