Sperm Quantification and Motility Analysis

A research-grade Python pipeline for analyzing sperm dynamics from videomicroscopy

---

About This Project

I developed this computational framework during my research into reproductive biophysics and microfluidic technologies. The pipeline combines WHO-standardized clinical metrics with physics-based trajectory analysis to provide comprehensive quantification of sperm swimming dynamics.

Key Features

✅ Advanced Tracking: Multi-object tracking using Kalman filtering and Hungarian algorithm
✅ Clinical Standards: Full implementation of WHO 2021 motility parameters (VCL, VSL, VAP, LIN, WOB, ALH, BCF)
✅ Physics-Based Analysis: Mean squared displacement, persistence length, diffusion coefficients
✅ Validation Tools: Active Brownian particle simulator for ground-truth testing
✅ Statistical Analysis: Population comparison and hypothesis testing capabilities
✅ Research-Ready Outputs: Publication-quality visualizations and data export

---

Motivation

My fascination with biophysics led me to explore the intersection of reproductive biology and microfluidic engineering. Recent studies on label-free sperm separation through microfluidic devices (Sexing by Self-Propulsion) inspired me to create this analytical framework.

What This Pipeline Does

• Quantifies swimming patterns with research-grade precision
• Supports experimental validation of microfluidic separation techniques
• Enables physics-based modeling of active particles in confined geometries
• Provides clinical assessment following WHO 2021 guidelines

Applications

• Reproductive Biology: Motility characterization across species and conditions
• Microfluidic Engineering: Device validation and optimization
• Clinical Settings: Sperm quality assessment for assisted reproduction
• Fundamental Physics: Active matter and self-propelled particle dynamics

---

Installation

Requirements

• Python 3.9 or higher
• OpenCV, scikit-image, NumPy, SciPy
• Matplotlib, Seaborn for visualization
• FilterPy for Kalman filtering

Setup

# Clone repository
git clone https://github.com/yourusername/sperm-quantification.git
cd sperm-quantification

# Create virtual environment (recommended)
python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

# Install package in development mode
pip install -e .

---

Quick Start

1. Generate Synthetic Data

from src.simulation import generate_synthetic_dataset

# Generate synthetic sperm trajectories
video, trajectories, metadata = generate_synthetic_dataset(
    n_particles=20,
    duration=5.0,
    save_path="data/synthetic/test_video.avi"
)

print(f"Generated {metadata['n_frames']} frames with {metadata['n_particles']} sperm")

2. Track Sperm from Video

from src.preprocessing import VideoReader, preprocess_frame
from src.detection import BlobDetector
from src.tracking import SpermTracker

# Initialize components
detector = BlobDetector(method='dog', threshold=0.1)
tracker = SpermTracker(max_distance=30, min_track_length=10)

# Process video
with VideoReader("data/raw/sperm_video.avi") as video:
    for frame in video:
        # Preprocess
        processed = preprocess_frame(frame, denoise_method='bilateral')
        
        # Detect sperm
        detections = detector.detect(processed)
        
        # Update tracker
        active_tracks = tracker.update(detections)

# Get completed trajectories    
all_tracks = tracker.get_all_tracks()
print(f"Tracked {len(all_tracks)} sperm")

3. Compute Motility Metrics

from src.metrics import analyze_single_trajectory

# Analyze a trajectory
trajectory = all_tracks[0].get_trajectory()  # (n_points, 2) array

metrics = analyze_single_trajectory(
    trajectory,
    fps=30,
    pixel_size_um=0.1
)

print(f"VCL: {metrics['VCL']:.2f} μm/s")
print(f"VSL: {metrics['VSL']:.2f} μm/s")
print(f"LIN: {metrics['LIN']:.2f}")
print(f"Classification: {metrics['motility_classification']}")

4. Visualize Results

from src.visualization import plot_trajectories, plot_velocity_distributions

# Plot all trajectories
trajectories_list = [track.get_trajectory() for track in all_tracks]
plot_trajectories(trajectories_list, pixel_size_um=0.1, save_path="results/trajectories.png")

# Plot metric distributions
metrics_list = [analyze_single_trajectory(traj, 30, 0.1) for traj in trajectories_list]
plot_velocity_distributions(metrics_list, save_path="results/velocity_dist.png")

---

Pipeline Architecture

┌─────────────┐
│ Video Input │
└──────┬──────┘
       │
       ▼
┌─────────────────┐
│  Preprocessing  │  ← Denoising, background subtraction
└─────┬───────────┘
      │
      ▼
┌──────────────┐
│  Detection   │  ← Blob detection, watershed segmentation
└──────┬───────┘
       │
       ▼
┌──────────────┐
│   Tracking   │  ← Kalman filter, Hungarian assignment
└──────┬───────┘
       │
       ▼
┌──────────────┐
│   Metrics    │  ← VCL, VSL, MSD, persistence
└──────┬───────┘
       │
       ▼
┌──────────────┐
│ Visualization│  ← Plots, statistics, export
└──────────────┘

---

Module Documentation

Preprocessing (src/preprocessing/)

• video_loader.py: Load videos (AVI, MP4, TIFF stacks)
• denoising.py: Gaussian, bilateral, NLM filtering
• background.py: MOG2 and median background subtraction

Detection (src/detection/)

• blob_detector.py: DoG, LoG, DoH blob detection
• segmentation.py: Watershed, adaptive thresholding

Tracking (src/tracking/)

• kalman.py: Kalman filter for motion prediction
• tracker.py: Multi-object tracker with Hungarian algorithm

Metrics (src/metrics/)

• velocity.py: VCL, VSL, VAP, LIN, WOB, ALH, BCF (WHO 2021)
• trajectory.py: MSD, persistence length, turning angles
• motility.py: High-level analysis combining all metrics

Simulation (src/simulation/)

• active_brownian.py: Stochastic ABP model
• synthetic_data.py: Generate realistic microscopy videos

Visualization (src/visualization/)

• plotting.py: Publication-quality trajectory and metric plots

Analysis (src/analysis/)

• statistics.py: Population comparison, effect sizes

---

Configuration

Customize pipeline parameters via YAML files in configs/:

# configs/default.yaml
detection:
  method: "blob"
  blob:
    threshold: 0.1
    min_area: 10
    max_area: 200

tracking:
  max_distance: 30
  max_gap: 5
  min_track_length: 10

metrics:
  velocity:
    time_window: 1.0

Load configuration:

from src.utils import Config

config = Config('configs/microfluidic.yaml')
threshold = config.get('detection.blob.threshold')

---

Example Notebooks

Explore notebooks/ for detailed tutorials:

1. 01_synthetic_data_demo.ipynb: Generate and visualize synthetic data
2. 02_basic_analysis.ipynb: Complete pipeline walkthrough
3. 03_advanced_tracking.ipynb: Tracking algorithm comparison
4. 04_statistical_comparison.ipynb: X vs Y sperm analysis

---

Testing

Run unit tests:

pytest tests/ -v

Tests cover:

• Tracking accuracy on synthetic data
• Metric computation validation
• ABP parameter recovery

---

Assumptions and Limitations

Assumptions

1. Imaging: Sufficient resolution to resolve sperm heads (~3-5 μm)
2. Frame Rate: Adequate temporal sampling (typically 30-60 Hz)
3. Contrast: Sperm appear brighter than background (or vice versa)
4. Planarity: Sperm swim primarily in focal plane

Limitations

1. 3D Motion: Pipeline analyzes 2D projections; out-of-plane motion not captured
2. Flagellum: Optional flagellum detection; focus is on head tracking
3. Dense Fields: Performance degrades with > 50 sperm/field
4. Fixed Parameters: Some WHO metrics require manual threshold tuning

---

Connection to SEB Research

This pipeline directly supports Sexing by Self-Propulsion research by:

• Quantifying subtle kinematic differences between X and Y sperm
• Validating separation efficiency in microfluidic devices
• Providing ground truth for machine learning classifiers
• Enabling parameter sweeps via synthetic data generation

Key metrics for SEB:

• Linearity (LIN): Directional persistence
• Beat parameters (ALH, BCF): Flagellar dynamics
• Persistence length: Swimming strategy characterization

---

Citation

If you find this pipeline useful in your research, please consider citing:

@software{sperm_quantification_2024,
  author = {Faheem},
  title = {Sperm Quantification and Motility Analysis Pipeline},
  year = {2024},
  url = {https://github.com/Faheem8585/sperm-quantification}
}

---

Contributing

I welcome contributions and suggestions! Feel free to:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/your-feature)
3. Commit your changes (git commit -m 'Add your feature')
4. Push to the branch (git push origin feature/your-feature)
5. Open a Pull Request

---

License

This project is licensed under the MIT License - see LICENSE file for details.

---

Acknowledgments

This project builds upon established methods in computer vision, active matter physics, and clinical semen analysis. I'm grateful to the scientific community for:

• WHO Laboratory Manual (2021) - standardized motility criteria
• Active matter physics research - theoretical foundations for ABP models
• OpenCV and scikit-image communities - excellent computer vision tools

---

Developed with passion for computational biology and biophysics research