Fuzzy Galactic Swarm Optimization Coupled with Superpixel-Based Microscopic Image Segmentation

Abstract

Microscopic imaging is fundamental to modern medical diagnostics, offering detailed structural and morphological insights essential for studying cellular behavior and detecting diseases. However, automated analysis of such images remains challenging due to variability in cell shapes, overlapping structures, noise, and inconsistent staining. These factors complicate accurate interpretation and necessitate advanced computational techniques. A critical step in this process is precise segmentation of cellular structures, which significantly impacts downstream tasks like classification and diagnosis. Effective segmentation enhances visual clarity and improves the reliability of computer-assisted diagnostic systems by identifying well-defined regions of interest. To address these challenges, this paper proposes a novel segmentation framework, SUFGSO (Superpixel-based Fuzzy Galactic Swarm Optimization). The approach integrates type-II fuzzy logic to manage uncertainty and ambiguity, with galactic swarm optimization, a metaheuristic inspired by hierarchical swarm intelligence. Additionally, superpixel techniques are employed to group pixels into meaningful regions, reducing computational complexity and improving spatial coherence, especially in high-resolution images. The proposed method is evaluated using four established cluster validity indices to ensure a comprehensive assessment of segmentation performance. Experimental results demonstrate that SUFGSO achieves improved accuracy and efficiency, indicating its effectiveness as a practical tool for microscopic image analysis in medical applications. 

Received: 2 June 2025 | Revised: 12 March 2026 | Accepted: 15 May 2026

Conflicts of Interest

The authors declare that they have no conflicts of interest to this work. 

Data Availability Statement

The data that support the findings of this study are openly available in the Center for Research in Biological Systems at https://www.crbs.ucsd.edu/crbs-projects/highlighted-research-projects?highlight=94. 

Author Contribution Statement

Debasish Biswas: Methodology, Software, Formal analysis. Shouvik Chakraborty: Conceptualization, Writing – original draft, Supervision, Project administration. Chinmoy Ghorai: Validation, Writing – review & editing, Supervision, Resources.