Holotomography in 2025: From Morphometric Imaging to AI-Driven Multimodal Phenotyping

TL;DR

Holotomography by 2025 has evolved into a versatile, AI-integrated imaging platform enabling detailed, label-free 3D cellular analysis with applications in biomedicine, pathology, and high-throughput screening.

Contribution

This review highlights recent advances in holotomography, emphasizing AI integration, extension to complex tissues, and multimodal approaches, outlining future research directions and clinical applications.

Findings

Deep learning enhances virtual staining and classification.

HT applied to organoids and thick tissues.

Integration with Raman and polarization microscopy improves specificity.

Abstract

By 2025, holotomography (HT) has matured from a niche optical modality into a versatile platform for quantitative, label-free imaging in biomedicine. By reconstructing the three-dimensional refractive-index (RI) distribution of cells and tissues, HT enables high-resolution volumetric imaging with low phototoxicity and minimal sample perturbation. This Review surveys recent advances in the field and highlights three emerging directions: (i) the incorporation of deep-learning approaches for virtual staining, phenotypic classification, and automated analysis; (ii) the extension of HT to structurally complex biological systems, including organoids and thick tissue specimens; and (iii) the integration of HT with complementary modalities, such as Raman and polarization-sensitive microscopy, to enhance molecular and biophysical specificity. We summarize current HT applications spanning subcellular phenotyping, metabolic and mechanical profiling, and early-stage clinical studies in areas such as infectious disease and pathology. Finally, we discuss remaining technical and translational challenges and outline a roadmap for the prospective integration of HT into digital pathology and high-throughput screening workflows.