Search for: All records

Abstract We introduce a digital inline holography (DIH) method combined with deep learning (DL) for real-time detection and analysis of bacteria in liquid suspension. Specifically, we designed a prototype that integrates DIH with fluorescence imaging to efficiently capture holograms of bacteria flowing in a microfluidic channel, utilizing the fluorescent signal to manually identify ground truths for validation. We process holograms using a tailored DL framework that includes preprocessing, detection, and classification stages involving three specific DL models trained on an extensive dataset that included holograms of generic particles present in sterile liquid and five bacterial species featuring distinct morphologies, Gram stain attributes, and viability. Our approach, validated through experiments with synthetic data and sterile liquid spiked with different bacteria, accurately distinguishes between bacteria and particles, live and dead bacteria, and Gram-positive and negative bacteria of similar morphology, all while minimizing false positives. The study highlights the potential of combining DIH with DL as a transformative tool for rapid bacterial analysis in clinical and industrial settings, with potential extension to other applications including pharmaceutical screening, environmental monitoring, and disease diagnostics. 

Abstract Obtaining in situ measurements of biological microparticles is crucial for both scientific research and numerous industrial applications (e.g., early detection of harmful algal blooms, monitoring yeast during fermentation). However, existing methods are limited to offer timely diagnostics of these particles with sufficient accuracy and information. Here, we introduce a novel method for real‐time, in situ analysis using machine learning (ML)‐assisted digital inline holography (DIH). Our ML model uses a customized YOLOv5 architecture specialized for the detection and classification of small biological particles. We demonstrate the effectiveness of our method in the analysis of 10 plankton species with equivalent high accuracy and significantly reduced processing time compared to previous methods. We also applied our method to differentiate yeast cells under four metabolic states and from two strains. Our results show that the proposed method can accurately detect and differentiate cellular and subcellular features related to metabolic states and strains. This study demonstrates the potential of ML‐driven DIH approach as a sensitive and versatile diagnostic tool for real‐time, in situ analysis of both biotic and abiotic particles. This method can be readily deployed in a distributive manner for scientific research and manufacturing on an industrial scale.