Automated Focusing Algorithms for Fluorescence/Brightfield Slide Imaging Microscope

Context & Background

Fluorescence and brightfield slide microscopes are used in medical pathology laboratories to detect cellular anomalies. Focusing on glass slides manually is slow, limiting laboratory throughput. Developing high-precision automated focusing algorithms can enable digital pathological screening at scale.

Problems to be Addressed

Biological slides have varied thickness and cell layers, making focus estimation difficult. Standard autofocus algorithms are slow and struggle under low-contrast or fluorescent imaging settings.

Aims and Objectives

1. Develop high-speed, automated autofocusing algorithms.
2. Test models on brightfield and fluorescence slide images.
3. Integrate focus control software directly with microscope actuators.

Methodology

The project designs sharpness-evaluation algorithms using image gradients and deep CNN models. The system predicts focus depth directly from out-of-focus slide images, sending micro-step instructions to stepper motors to achieve optical focus in milliseconds.

Expected Outcomes

A high-precision microscope autofocus software library, integration with diagnostic slide scanners, and publications in optical imaging journals.