Learning-based automatic sensing and size classification of microparticles using smartphone holographic microscopy.

Area of Science:

• Optics and Photonics
• Biomedical Engineering
• Environmental Science

Background:

• Accurate microparticle size classification is crucial for environmental monitoring and biomedical applications.
• Conventional methods often rely on bulky equipment and offer limited performance.
• A need exists for portable, smart platforms for efficient particle analysis.

Purpose of the Study:

• To develop a novel, portable sensing platform for automatic microparticle identification.
• To integrate smartphone-based digital in-line holographic microscopy (DIHM) with machine learning algorithms.
• To achieve high accuracy in microparticle size classification using a compact device.

Main Methods:

• A smartphone-based DIHM system was constructed using a laser, pinhole, sample holder, 3D-printed attachment, and smartphone camera.
• Holograms of polystyrene microparticles (2-50 μm) were captured with a wide field-of-view and high resolution.
• Machine learning algorithms, including support vector machines, were employed to classify particles based on extracted holographic features.

Main Results:

• The developed platform is compact (4 × 8 × 10 cm³) and lightweight (220 g).
• A support vector machine model, utilizing six key features (three geometrical, three light-intensity), achieved >98% accuracy in classifying microparticle sizes.
• The system demonstrated high performance on both training and test datasets.

Conclusions:

• The developed handheld smartphone-based platform offers a portable and smart solution for microparticle size classification.
• This technology has significant potential for applications in mobile healthcare and environmental monitoring.
• The synergistic integration of DIHM and machine learning provides a powerful tool for advanced imaging needs.