Real-time fluorescence imaging flow cytometry enabled by motion deblurring and deep learning algorithms

Abstract

Fluorescence imaging flow cytometry (IFC) has been demonstrated as a crucial biomedical technique for analyzing specific cell subpopulations from heterogeneous cellular populations. However, the high-speed flow of fluorescence cells leads...

Advancing precision single-cell analysis of red blood cells through semi-supervised deep learning using database of patients with post-COVID-19 syndrome

“…The system's design fuses a YOLOv7-tiny localization block with the enhanced feature extraction capabilities of FocalNet and a tuned MobileNetV3-small classification block. Inference times of the optimal configuration yield both fast inference and high accuracy, comparable to past works on IFC 22,36,44 . The modularity of the architecture allows for the easy adaptation and replacement of individual components, enabling its application to other complex image analysis tasks beyond RBC morphology classification.…”

“…Tasks involving the identification and classification of cells and particles 21,32,36 have been slow and challenging due to the exigent nature of manual gating and the limitations of proprietary code and datasets. Variant cell morphologies and states pose additional challenges for accurate identification, thus necessitating sophisticated computational methods 37,39,40,44 .…”

“…Several works have mentioned obstacles such as out-of-focus image blurring caused by imperfect fluidic focusing of cells 39 , which requires strategies to restore image clarity and improve overall performance. Some works 36,37 have emphasized the significance of image de-blurring and reconstruction for improved imaging quality and performance metrics. Zhang and co-authors 40 demonstrated an 11% increase in variance and a 95% increase in the first-order gradient summation of cell images through machine learning algorithms, indicating the potential for data restoration.…”

Is AI essential? Examining the need for deep learning in image-activated sorting of Saccharomyces cerevisiae

“…7,8 Among these, imaging flow cytometry has seen considerable integration with AI for automated handling and classification of large volumes of cell image data. 9–18 Traditionally, AI-based cell image classification has been chiefly realized through feature extraction followed by machine learning (ML) to construct classifier algorithms. 9 Preset features ( e.g.…”

“…The benefits of utilizing complete image data over summarized statistics, coupled with the potential for improved classification accuracy using unbiased, learned features, have led to an increasing number of reports deploying CNNs for processing data from imaging flow cytometry. 13–17,22–24…”