Neuroblastoma Cells Classification Through Learning Approaches by Direct Analysis of Digital Holograms

Abstract

The label-free single cell analysis by machine and Deep Learning, in combination with digital holography in transmission microscope configuration, is becoming a powerful framework exploited for phenotyping biological samples. Usually, quantitative phase images of cells are retrieved from the reconstructed complex diffraction patterns and used as inputs of a deep neural network. However, the phase retrieval process can be very time consuming and prone to errors. Here we address the classification of cells by using learning strategies with images coming directly from the raw recorded digital holograms, i.e. without any data processing or refocusing involved. Indeed, in the raw digital hologram the entire complex amplitude information of the sample is intrinsically embedded in the form of modulated fringes. We develop a training strategy, based on deep and feature based machine learning models, in order extract such information by skipping the classical reconstruction process for classifying different neuroblastoma cells. We provided an experimental validation by using the proposed strategy to classify two neuroblastoma cell lines.

Automated Stain‐Free Holographic Image‐Based Phenotypic Classification of Elliptical Cancer Cells

“…Most of the studies regarding cancer cell classification focus on classifying single‐organ cancer cells (breast, lung, skin or lungs etc.). [ 10,24,26,32,33 ] In this study, we show that the proposed model can classify the different organ cancer cells in elliptical status that have similar morphological structures, as shown in Figure . This article uses feature‐based and deep image learning classification for three cancer cell lines of lung, breast, and skin.…”

“…[31] Most of the studies regarding cancer cell classification focus on classifying single-organ cancer cells (breast, lung, skin or lungs etc.). [10,24,26,32,33] In this study, we show that the proposed model can classify the different organ cancer cells in elliptical status that have similar morphological structures, as shown in The experiments show that deep learning outperforms featurebased classification. We show that even if plenty of effort is made to pick up the most informative features, deep learning of the phase images with a CNN is still more suitable for cancer cell classification in DHM.…”

“…[ 25 ] Images obtained from DHM are first preprocessed by performing phase reconstruction on holographic images before being processed for detecting cells. In the study by Priscoli et al, [ 26 ] holographic images without phase reconstruction obtained from microfluidic channels as well as from Petri dishes were used directly to perform classification using deep learning.…”

Tracking-based rolling angles recovery method for holographic tomography of flowing cells

“…At the same time, a lower spatial resolution is more suitable for high-throughput analysis. Recently, a classification task has been solved to discriminate different CTC lines by using the 2D holograms of flowing cells 23 . The next challenge is the development of machine-learning tools fed by a large number of 3D RI tomograms, from which more informative features can be extracted with respect to the 2D imaging, thus covering a further step towards the implementation of the liquid biopsy paradigm 24 .…”

Multi-Class Classification and Multi-Output Regression of Three-Dimensional Objects Using Artificial Intelligence Applied to Digital Holographic Information

“…The CNN is one of the major deep neural networks and a basic building block of deep learning. The ability of a CNN to handle greater numbers of convolutional layers and pooling layers in the feature-extraction stage and process n number of neurons in the classification-layer stage has made its use feasible for different kinds of tasks such as classification, autofocusing, fringe pattern denoising, image segmentation, image super-resolution, and hologram reconstruction in digital holography [ 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 , 34 , 35 , 36 , 37 , 38 , 39 , 40 , 41 , 42 , 43 ]. A CNN [ 44 ] consists of feature-extraction and classification layers.…”