Diatom Images Research Articles

A high-efficiency local and global detector for diatom-based drowning diagnosis

Detection and classification of diatoms has been considered as the most effective method for the drowning diagnosis in forensic practice. However, challenges (such as limited samples, complex background interference, and detecting multiple objects) remain during detection and classification. To address these challenges, we propose a MDMS (multi-scale dynamic multi-head self-attention)-based deep learning network for the diatom-based drowning diagnosis. This framework uses TFFT (two-stage full fine-tuning training strategy) to solve the challenges of limited diatom samples and detecting multiple objects, and builds a multi-scale dynamic multi-head self-attention module to extract the local and global features of diatom images. In the meantime, we introduce an online hard example mining strategy to attenuate the complex background interference. The experimental results show that the proposed framework can effectively reduce the missing and false detection rates of diatom objects, with the mAP (mean Average Precision) reaching 92.434%, which is better than the result using the mainstream methods.

A New Method for the Detection of Siliceous Microfossils on Sediment Microscope Slides Using Convolutional Neural Networks

AbstractDiatom communities preserved in sediment samples are valuable indicators for understanding the past and present dynamics of phytoplankton communities, and their response to environmental changes. These studies are traditionally achieved by counting methods using optical microscopy, a time‐consuming process that requires taxonomic expertise. With the advent of automated image acquisition workflows, large image data sets can now be acquired, but require efficient preprocessing methods. Detecting diatom frustules on microscope images is a challenge due to their low relief, diverse shapes, and tendency to aggregate, which prevent the use of traditional thresholding techniques. Deep learning algorithms have the potential to resolve these challenges, more particularly for the task of object detection. Here we explore the use of a Faster Region‐based Convolutional Neural Network model to detect siliceous biominerals, including diatoms, in microscope images of a sediment trap series from the Mediterranean Sea. Our workflow demonstrates promising results, achieving a precision score of 0.72 and a recall score of 0.74 when applied to a test set of Mediterranean diatom images. Our model performance decreases when used to detect fragments of these microfossils; it also decreases when particles are aggregated or when images are out of focus. Microfossil detection remains high when the model is used on a microscope image set of sediments from a different oceanic basin, demonstrating its potential for application in a wide range of contemporary and paleoenvironmental studies. This automated method provides a valuable tool for analyzing complex samples, particularly for rare species under‐represented in training data sets.

A lightweight convolutional neural network (CNN) model for diatom classification: DiatomNet.

Diatoms are a type of algae with many species. Accurate and quick classification of diatom species is important in many fields, such as water quality analysis and weather change forecasting. Traditional methods for diatom classification, specifically morphological taxonomy and molecular detection, are time-consuming and may not provide satisfactory performance. However, in recent years, deep learning has demonstrated impressive performance in this task, just like other image classification problems. On the other hand, networks with more layers do not guarantee increased accuracy. While increasing depth can be useful in capturing complex features and patterns, it also introduces challenges such as vanishing gradients, overfitting, and optimization challenges. Therefore, in our work, we propose DiatomNet, a lightweight convolutional neural network (CNN) model that can classify diatom species accurately while requiring low computing resources. A recently introduced dataset consisting of 3,027 diatom images and 68 diatom species is used to train and evaluate the model. The model is compared with well-known and successful CNN models (i.e., AlexNet, GoogleNet, Inceptionv3, ResNet18, VGG16, and Xception) and their customized versions obtained with transfer learning. The comparison is based on several success metrics: accuracy, precision, recall, F-measure, number of learnable parameters, training, and prediction time. Eventually, the experimental results reveal that DiatomNet outperforms the other models regarding all metrics with just a few exceptions. Therefore, it is a lightweight but strong candidate for diatom classification tasks.

A comparative assessment of time-consuming and laborious diatom analysis：Brief experimentation with suggestion of automatic identification

Diatom testing is considered a useful method for providing supportive evidence for the diagnosis of drowning in forensic pathology. However, various factors remain controversial for recognizing diatoms, such as being time-consuming and laborious and influencing the consistency of the results. Given the absence of precise and well-defined studies on this subject, this study aimed to determine the relationship between the ability to identify diatoms and researchers with different technical backgrounds. A total of 55 samples from 18 cases, including water, lungs, liver, and kidneys, were treated using the microwave digestion-vacuum filtration-automated scanning electron microscopy (MD-VF-Auto SEM), which was used to compare diatom analyses among three groups of well-trained forensic pathologists (FPs), trained junior employees (JEs), and new trainees (TEs). In addition to achieving similar accuracy of positive findings from drowning cases, counting efficiency was evaluated based on taxonomy records and counting time after viewing more than 5500 diatom images. In contrast to the higher counting efficiency of the JE group than that of the TE group, we observed a statistically significant difference (p < 0.05) in the diatom classification between these two groups. Based on our experiments, an efficient analysis for automatically identifying and classifying diatoms is urgently required.

Cross-Correlation and Fractal Analysis in the Images Diatoms Symmetry

Diatoms are one of the main species of interest due to their possible impact on the maintenance of ecosystems, as well as their photosynthetic capacity and biotechnological applications. Reliable, easy, rapid identification of diatoms remains a highly desirable goal to avoid the difficulty for experts in the visual analysis of frustules; a method commonly used to compare the distinctive traits of diatoms, such as ornaments and the marks from images taken by optical and electronic microscopy for identification. This study proposed three simple morphometric methods based on fractal dimension measurements in conjunction with 2D cross-correlation parameters to analyze diatom images. This novel approach demonstrates the differentiation feasibility between a wide variety of diatom species with heterogeneity in their structures and shapes, which could possibly expand the application of this preprocessing process for the identification of new types of images or other biological species.

Segmentation of diatoms using edge detection and deep learning

Diatoms are photosynthesizing algae found in almost every aquatic environment. Detecting the number and diversity of diatoms is very important to analyze water quality appropriately. Accurate segmentation of diatoms is therefore crucial for this detection process. In this study, a new and effective model for the automatic segmentation of diatoms based on image processing and deep learning algorithms is proposed. In the proposed model, edge segments of a given image containing diatoms and nondiatom particles are first obtained. These edge segments are then combined, resulting in closed contours representing diatom candidates. In the final step, the diatom candidates are classified as either diatom or nondiatom using a transfer learning approach with different pretrained deep neural network models including AlexNet, ResNet18, VGG16, and their ensemble. Hence, the boundary of a diatom candidate, which is classified as a diatom, is considered to be the final output of the proposed segmentation model. Besides, a new and annotated diatom image dataset is introduced. The dataset is divided into two parts. The first part is used to train and evaluate the classification stage of the proposed model, while the second part is used for evaluating the overall performance of the proposed segmentation model. This dataset is also made publicly available at https: //www.kaggle.com/huseyingunduz/diatom-dataset for the research community. Through extensive experimental work, the performance of the proposed segmentation model is measured in terms of precision, recall, F-measure, and intersection over union values. Our work is also compared with the previous works based on various aspects such as segmentation method, dataset size, diatom varieties, dataset availability, success metric, and segmentation performance. Considering all these aspects, it is verified that the proposed model surpasses the previous works, and stands out as a well-performing approach to the automatic segmentation of diatoms.

Automated diatom detection in forensic drowning diagnosis using a single shot multibox detector with plump receptive field

The detected result of diatoms is an important indicator in forensic drowning examination, and most of the current deep learning methods have achieved greater success in detecting diatoms with simple or no backgrounds. However, diatom images captured by the high-definition electron scanning microscopy in modern forensic science contain complex backgrounds and hamper the accurate diatom detection, resulting in the omission detection of the small and marginal diatoms in multi-diatom scenario. In this paper, we proposed a Hybrid-Dilated-Convolution-incorporated Single Shot Multibox Detector (HDC-SSD) to address this problem. By adopting the merit of the plump receptive field of HDC, the proposed algorithm not only improves the detection rate but also enhances the detection ability of the small objects and the marginal objects. The proposed method was validated by using our self-established dataset. Compared with SSD, the HDC-SSD reduces the undetected rate by approximately 48.6% and almost keeps as fast as the SSD. More importantly, compared with some current state-of-the-art methods, the HDC-SSD obtains the highest Recall value at 0.9302.

An archive of the published photographs of diatoms

An information system designed for the remote work with digital images of diatoms is described in the paper. Photographs of diatoms published in scientific articles were used as data. During taxonomic identification, hydrobiologists often refer to such publications to compare the studied organism with the typical representatives of genera or species. The created information system “Arkhip” (3d-microalgae.org/arxip) allows to index and consolidate the information about the published photos in a single database, which will help to speed up the identification of diatom species.

Segmentation of Meristem Cells by an Automated Optimization Algorithm

Meristem cells are irregularly shaped and appear in confocal images as dark areas surrounded by bright ones. Images are characterized by regions of very low contrast and absolute loss of edges deeper into the meristem. Edges are blurred, discontinuous, sometimes indistinguishable, and the intensity level inside the cells is similar to the background of the image. Recently, a technique called Parametric Segmentation Tuning was introduced for the optimization of segmentation parameters in diatom images. This paper presents a PST-tuned automatic segmentation method of meristem cells in microscopy images based on mathematical morphology. The optimal parameters of the algorithm are found by means of an iterative process that compares the segmented images obtained by successive variations of the parameters. Then, an optimization function is used to determine which pair of successive images allows for the best segmentation. The technique was validated by comparing its results with those obtained by a level set algorithm and a balloon segmentation technique. The outcomes show that our methodology offers better results than two free available state-of-the-art alternatives, being superior in all cases studied, losing 9.09% of the cells in the worst situation, against 75.81 and 25.45 obtained in the level set and the balloon segmentation techniques, respectively. The optimization method can be employed to tune the parameters of other meristem segmentation methods.

Deep learning-based diatom taxonomy on virtual slides

Deep convolutional neural networks are emerging as the state of the art method for supervised classification of images also in the context of taxonomic identification. Different morphologies and imaging technologies applied across organismal groups lead to highly specific image domains, which need customization of deep learning solutions. Here we provide an example using deep convolutional neural networks (CNNs) for taxonomic identification of the morphologically diverse microalgal group of diatoms. Using a combination of high-resolution slide scanning microscopy, web-based collaborative image annotation and diatom-tailored image analysis, we assembled a diatom image database from two Southern Ocean expeditions. We use these data to investigate the effect of CNN architecture, background masking, data set size and possible concept drift upon image classification performance. Surprisingly, VGG16, a relatively old network architecture, showed the best performance and generalizing ability on our images. Different from a previous study, we found that background masking slightly improved performance. In general, training only a classifier on top of convolutional layers pre-trained on extensive, but not domain-specific image data showed surprisingly high performance (F1 scores around 97%) with already relatively few (100–300) examples per class, indicating that domain adaptation to a novel taxonomic group can be feasible with a limited investment of effort.

Deep learning network-based recognition and localization of diatom images against complex background

ObjectiveWe propose a deep learning network-based method for recognizing and locating diatom targets with interference by complex background in autopsy.MethodThe system consisted of two modules: the preliminary positioning module and the accurate positioning module. In preliminary positioning, ZFNet convolution and pooling were utilized to extract the high-level features, and Regional Proposal Network (RPN) was applied to generate the regions where the diatoms may exist. In accurate positioning, Fast R-CNN was used to modify the position information and identify the types of the diatoms.ResultsWe compared the proposed method with conventional machine learning methods using a self-built database of images with interference by simple, moderate and complex backgrounds. The conventional methods showed a recognition rate of diatoms against partial background interference of about 60%, and failed to recognize or locate the diatom objects in the datasets with complex background interference. The deep learning network-based method effectively recognized and located the diatom targets against complex background interference with an average recognition rate reaching 85%.ConclusionThe proposed method can be applied for recognition and location of diatom targets against complex background interference in autopsy.

Binary Masks of Concentric Rings as a Method to Approximate Identification of Diatoms Using Images

Diatoms are currently being used in different fields, and their application as bioindicators is one of their main benefits. Their identification is carried out by specialists, the main difficulty being the wide variety of existing species and the great similarity between some of them. In recent years attempts have been made to create an automatic identification system with the collaboration of experts in areas of pattern recognition. In this research we analyzed the use of binary masks of rings in images as a method to automatically identify diatoms, or render an approximative identification. Among the advantages of the proposed method is its invariance regarding the different positions in which the diatoms can be found. Invariance to illumination changes is achieved by means of different transforms, such as the power law transform. The image databases used to develop the algorithms were diatom images taken from samples collected at springs in Lake Patzcuaro, Michoacan, Mexico, at the San Lazaro basin, Baja California Sur, Mexico and from the public image database of ADIAC. In morphological and textural descriptors, the background of the image must be eliminated in order to classify the diatoms, whereas in the proposed methodology this is not necessary. The obtained results reach over than 90% even when the diatoms in the images are broken.

Image Correlation Using Fractional Hermite Transform

In this paper, we generalize the Hermite transform into a fractional case using the fractional Fourier transform and the fractional convolution. The new methodology was evaluated using phytoplankton images with different illumination patterns and fragmented images. We found that the fractional Hermite transform had a better capability to recognize images. The discrimination coefficient was evaluated for the fractional Hermite transform and the conventional Hermite transform, finding more noise tolerate with the fractional Hermite transform. The Hermite fractional transform, in combination with the extreme phase filter, showed in a study, using fragmented diatom images, a better ability to classify diatoms, even when these had little information.

An Optimal Inception Based Segmentation and Classification Model for Diatom Species Images

Diatoms act an essential contributor to the fundamental creation in aquatic ecosystem, which is positioned at the foundation of the food chain. Presently, the diatoms appear as a most important topic over the globe in studies interrelated to weather changes, and in the design of functions which enables to model of those variations. In addition, it is an efficient indicator of ecological conditions and is widely employed in water quality assessment. In a similar way, deep learning model is a widely employed technique for classifying images among diverse applications. In this paper, an optimal segmentation and classification model for diatom images particularly species images. Here, edge detection based segmentation model is employed for segmenting the images and then Inception model is utilized for classifying images. A detailed simulation process takes place on the benchmark diatom images. An overall accuracy of 99 is attained by the presented model on the applied set of test images. The outcome is compared to the state of art classification models and the results exhibited the superior performance of the presented model.

An Efficient Automated Deep Learning Model For Diatom Image Segmentation And Classification

Recently, diatoms, a type of algae microorganism with numerous species, are relatively helpful for water quality determination, and is treated as an important topic in applied biology nowadays. Simultaneously, deep learning (DL) also becomes an important model applied for various image classification problems. This study introduces a new Inception model for diatom image classification. The presented model involves two main stages namely segmentation and classification. Here, a deep learning based Inception model is employed for classification purposes. To further improve the classifier efficiency, edge detection based segmentation model is also applied where the segmented input is provided as an input to the classifier stage. An experimental validation takes place on diverse set of diatom dataset with various preprocessing models. The results pointed out that the presented DL model shows extraordinary classification performance with a classifier accuracy of 99%.

Atlas of diatoms (Bacillariophyta) from diverse habitats in remote regions of western Canada.

High-resolution LM images of diatoms from remote regions of western Canada are presented as a contribution to our knowledge of diatom floristics, ecology and biogeography in North America. Approximately 600 taxa are imaged in 132 plates. Genera with the most taxa are Cymbella (19 taxa), Cymbopleura (29), Encyonema (23), Encyonopsis (15), Eunotia (77), Gomphonema (42), Navicula (47), Neidium (20), Nitzschia (35), Pinnularia (50) and Stauroneis (34). Diatoms were collected from diverse habitats in four of North America’s major biomes: Arctic tundra, taiga, Rocky Mountains and Pacific rainforest. Many of the photographed taxa could not be identified to species and are likely new to science. Other taxa may represent new records for North America or Canada. Images of voucher specimens are keyed to individual collection sites. Detailed descriptions of the collection sites include GPS coordinates, colour photographs, vegetation, algal substrates, elevations, pH, temperature and conductivity. Samples were collected from natural substrates in fresh to brackish, flowing and standing waters. Voucher slides are deposited in the Montana Diatom Collection (Helena) and the University of Montana Herbarium (Missoula). Cleaned diatom frustules have been deposited in the Diatom Herbarium of the Academy of Natural Sciences of Philadelphia.

Low-cost oblique illumination: an image quality assessment.

We study the effectiveness of several low-cost oblique illumination filters to improve overall image quality, in comparison with standard bright field imaging. For this purpose, a dataset composed of 3360 diatom images belonging to 21 taxa was acquired. Subjective and objective image quality assessments were done. The subjective evaluation was performed by a group of diatom experts by psychophysical test where resolution, focus, and contrast were assessed. Moreover, some objective nonreference image quality metrics were applied to the same image dataset to complete the study, together with the calculation of several texture features to analyze the effect of these filters in terms of textural properties. Both image quality evaluation methods, subjective and objective, showed better results for images acquired using these illumination filters in comparison with the no filtered image. These promising results confirm that this kind of illumination filters can be a practical way to improve the image quality, thanks to the simple and low cost of the design and manufacturing process.

Large-Scale Permanent Slide Imaging and Image Analysis for Diatom Morphometrics

Light microscopy analysis of diatom frustules is widely used in basic and applied research, notably taxonomy, morphometrics, water quality monitoring and paleo-environmental studies. Although there is a need for automation in these applications, various developments in image processing and analysis methodology supporting these tasks have not become widespread in diatom-based analyses. We have addressed this issue by combining our automated diatom image analysis software SHERPA with a commercial slide-scanning microscope. The resulting workflow enables mass-analyses of a broad range of morphometric features from individual frustules mounted on permanent slides. Extensive automation and internal quality control of the results helps to minimize user intervention, but care was taken to allow the user to stay in control of the most critical steps (exact segmentation of valve outlines and selection of objects of interest) using interactive functions for reviewing and revising results. In this contribution, we describe our workflow and give an overview of factors critical for success, ranging from preparation and mounting through slide scanning and autofocus finding to final morphometric data extraction. To demonstrate the usability of our methods we finally provide an example application by analysing Fragilariopsis kerguelensis valves originating from a sediment core, which substantially extends the size range reported in the literature.

Diatom Valve Three-Dimensional Representation: A New Imaging Method Based on Combined Microscopies.

The frustule of diatoms, unicellular microalgae, shows very interesting photonic features, generally related to its complicated and quasi-periodic micro- and nano-structure. In order to simulate light propagation inside and through this natural structure, it is important to develop three-dimensional (3D) models for synthetic replica with high spatial resolution. In this paper, we present a new method that generates images of microscopic diatoms with high definition, by merging scanning electron microscopy and digital holography microscopy or atomic force microscopy data. Starting from two digital images, both acquired separately with standard characterization procedures, a high spatial resolution (Δz = λ/20, Δx = Δy ≅ 100 nm, at least) 3D model of the object has been generated. Then, the two sets of data have been processed by matrix formalism, using an original mathematical algorithm implemented on a commercially available software. The developed methodology could be also of broad interest in the design and fabrication of micro-opto-electro-mechanical systems.

Diatoms image database of India (DIDI): A research tool

Abstract Biogeographical and environmental heterogeneity govern the high algal diversity of India. As of yet, no serious efforts have been made to explore Indian algae for ecological and industrial uses. The scattered nature and discrepancy in taxonomical information, as well as morphological variations in algae are hampering such studies. For example, these morphological variants may lead to ambiguous algae identification that may result in erroneous ecotoxicological assessment and also hinders the bioprospecting aspect of algae for industrial applications. To improve the situation, the Diatoms image database of India (DIDI) ( http://indianalgae.co.in ) is being developed to document the diatom diversity of India. In addition to identification aids and diatom images, the web site also includes data for environmental variables. Our database is the first web-based photographic documentation of diatoms from the sub-tropical region. DIDI is originally based on seven years of rigorous collection, examination and identification of the algal flora from different parts of India. All the diatom images given in the database are validated using Indian monographs, published papers and online image databases.