🧬 LFA Automatic Analyzer: High-Throughput Quantification for Lateral Flow Assays

📌 Overview

The LFA Automatic Analyzer is a robust, research-grade bioinformatics tool designed to automate the quantification of Lateral Flow Assay (LFA) strips. Unlike subjective visual interpretation, this tool leverages Computer Vision and Signal Processing algorithms to extract precise Test/Control (T/C) ratios, enabling quantitative analysis for applications ranging from clinical diagnostics (e.g., hCG) to toxicology (e.g., Xylazine detection).

This platform features an interactive Streamlit interface, allowing researchers to process batch datasets or individual biological samples with real-time visualization of signal profiles.

📸 Visual Demo

Interactive Dashboard	Automated Analysis Output
User-friendly Streamlit Interface for batch processing	Robust peak integration with automatic baseline correction

✨ Key Technical Features

1. Robust Signal Processing Pipeline

• Auto-Orientation & Segmentation: Algorithms automatically detect strip boundaries and correct orientation for consistent profiling.
• Background Detrending (Baseline Correction): Implements a polynomial fitting algorithm to remove membrane background noise (e.g., "smiling effects" or uneven lighting), ensuring a flat global baseline for integration.
• Adaptive Smoothing: Utilizes Savitzky-Golay filters to reduce high-frequency sensor noise while preserving the integrity of peak morphology.

2. Dual-Mode Workflow

• 📂 High-Throughput Batch Mode: Process hundreds of pre-cropped strip images simultaneously, exporting a consolidated Excel report with Pivot Tables for statistical analysis.
• ✂️ Interactive Single-Image Mode: Upload a raw photo of a multi-strip board. The integrated ROI selector allows users to define active zones, and the system automatically slices and analyzes each strip individually.

3. Advanced Peak Integration

• Dynamic Assay Sensitivity:
  • Traditional Mode: High sensitivity (prominence=0.5) for detecting faint positives.
  • Competitive Mode: High specificity (prominence=1.8) for analyzing competitive inhibition assays (e.g., small molecule detection).
• Global Baseline Integration: Calculates Area Under the Curve (AUC) using a unified baseline floor anchored to the signal minima, preventing bias in T/C ratio calculations.

🛠️ Installation & Usage

Prerequisites

• Python 3.8+
• pip package manager

Quick Start

1. Clone the repository:

   git clone [https://github.com/YourUsername/LFA-Automatic-Analyzer.git](https://github.com/YourUsername/LFA-Automatic-Analyzer.git)
   cd LFA-Automatic-Analyzer

2. Install dependencies:

   pip install -r requirements.txt

3. Launch the Dashboard:

   streamlit run app.py

   The application will automatically open in your default browser at http://localhost:8501.

⚙️ Algorithmic Configuration

Researchers can fine-tune the following parameters in app.py to adapt to different membrane materials:

Parameter	Default	Description
SMOOTH_WINDOW	15	Window length for Savitzky-Golay filtering. Increase for noisy nitrocellulose membranes.
BASELINE_METHOD	'lower'	Strategy for baseline anchoring ('lower' recommended for maximum sensitivity).
T_DIST_NEAR	30	Minimum pixel distance upstream from Control line to define the Test zone.

📊 Outputs

• Structured Data: Summary_Analysis.xlsx containing raw and adjusted T/C ratios.
• Visualization: High-resolution plots (.png) for every processed strip, showing the raw signal, fitted baseline, and integration boundaries.

🤝 Contribution

Developed by Minhao Liu (Lead Developer).

• Algorithm Design: Signal processing logic and T/C quantification.
• UI/UX: Streamlit dashboard implementation.

Acknowledgment:

• Pu Sun (Technical Contributor): Contributed to the conceptual design of the processing pipeline and provided insights for code refinement.