Automatic analysis of aqueous specimens for phytoplankton structure recognition and population estimation

Rodenacker, Karsten; Hense, Burkhard A.; Jütting, Uta; Gais, Peter

doi:10.1002/jemt.20338

Cited by 43 publications

(31 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It integrates methods already described [10,14] as well as new approaches into the analysis. The procedure performed by the system can be divided into the two general steps: automated imaging of the water sample and the following image processing and recognition.…”

Section: Discussionmentioning

confidence: 99%

“…Additionally, microscopes are currently used for the established procedures for plankton analysis, thereby allowing first an adaptation and then an easy comparison of both systems. One reported automated microscopic system is PLASA, which was developed with the goal of classifying different phytoplankton organisms with the use of automated microscopy and image analysis for an ecotoxicological microcosm study [10]. To allow a better differentiation between different phytoplankton taxa and between phytoplankton and other objects in the sample (zooplankton, detritus and inorganic particles) fluorescence imaging for phycoerythrin and chlorophyll was integrated into the system.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

PlanktoVision - an automated analysis system for the identification of phytoplankton

Schulze¹,

Tillich²,

Dandekar³

et al. 2013

BMC Bioinformatics

View full text Add to dashboard Cite

BackgroundPhytoplankton communities are often used as a marker for the determination of fresh water quality. The routine analysis, however, is very time consuming and expensive as it is carried out manually by trained personnel. The goal of this work is to develop a system for an automated analysis.ResultsA novel open source system for the automated recognition of phytoplankton by the use of microscopy and image analysis was developed. It integrates the segmentation of the organisms from the background, the calculation of a large range of features, and a neural network for the classification of imaged organisms into different groups of plankton taxa. The analysis of samples containing 10 different taxa showed an average recognition rate of 94.7% and an average error rate of 5.5%. The presented system has a flexible framework which easily allows expanding it to include additional taxa in the future.ConclusionsThe implemented automated microscopy and the new open source image analysis system - PlanktoVision - showed classification results that were comparable or better than existing systems and the exclusion of non-plankton particles could be greatly improved. The software package is published as free software and is available to anyone to help make the analysis of water quality more reproducible and cost effective.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

PlanktoVision - an automated analysis system for the identification of phytoplankton

Schulze¹,

Tillich²,

Dandekar³

et al. 2013

BMC Bioinformatics

View full text Add to dashboard Cite

show abstract

“…For instance, in situ imaging instruments can detect and enumerate symbioses directly in the environment (Olson and Sosik 2007 ;Malkassian et al 2011 ;Erickson et al 2012 ), which can be particularly useful for rare and delicate planktonic symbioses. Automated light microscopy, including automated acquisition and classifi cation by sophisticated algorithms, is another promising, not yet applied approach to quantify symbiotic associations in the environment (Rodenacker et al 2006 ;Zeder et al 2011 ;Schulze et al 2013 ). Coupling automated microscopy and FISH assays is a highly interesting combination to rapidly visualize and quantify free-living and associated forms of described partnerships.…”

Section: Perspectivesmentioning

confidence: 99%

Photosymbiosis in Marine Planktonic Protists

2015

View full text Add to dashboard Cite

Some of the most enigmatic components in the plankton are the diverse eukaryotic protists that live in close association with one or more partners. Mutualistic and commensal planktonic interactions are most commonly encountered in the oligotrophic open ocean at tropical and subtropical latitudes. They are functionally and ecologically distinct, and involve a great taxonomic diversity of single-celled partners. Protists like Foraminifera, Radiolaria, dinofl agellates and diatoms can all harbor microalgal symbionts of eukaryotic and prokaryotic origin inside (endosymbiosis) and/or outside (ectosymbiosis) their cytoplasm. Such symbioses (photosymbioses) combine phototrophy, heterotrophy and sometimes dinitrogen (N 2 ) fi xation (reduction of N 2 to ammonium). Symbiotic microorganisms therefore represent an important component of marine ecosystems and play a role in the food web and biogeochemical cycling (e.g., carbon and nitrogen). Despite their important ecological function and early recognition in the late nineteenth century, our knowledge about the diversity, distribution, and metabolic exchanges for many of the photosymbioses remains rudimentary compared with the other marine and terrestrial symbioses. Recent technical advances in single-cell genomics and imaging have greatly improved our understanding about planktonic symbioses. This review aims to present and compare many eukaryote-eukaryote and eukaryote-prokaryote photosymbioses described so far in the open ocean with an emphasis on their ecology and potential function in the ecosystem.

show abstract

“…Culverhouse et al [28] studied the classification accuracy achieved by the classification-committee-based (consisting of RBF networks) automated system DiCANN [29] and argued that accuracy of about 72% achieved by the system in a six-class phytoplankton categorization task was similar to the accuracy achieved by the trained personnel. Rodenacker et al applied fluorescence imaging in their image acquisition system, to capture more information for discrimination between five classes of phytoplankton [30]. Sosik and Olson [25] presented perhaps the most elaborate study regarding multiclass phytoplankton categorization using data obtained from Imaging FlowCytobot [31].…”

mentioning

confidence: 99%

An Integrated Approach to Analysis of Phytoplankton Images

Verikas

Gelžinis

Bačauskienė

et al. 2015

IEEE J. Oceanic Eng.

View full text Add to dashboard Cite

Abstract-The main objective of this paper is detection, recognition, and abundance estimation of objects representing the Prorocentrum minimum (Pavillard) Schiller (P. minimum) species in phytoplankton images. The species is known to cause harmful blooms in many estuarine and coastal environments. The proposed technique for solving the task exploits images of two types, namely, obtained using light and fluorescence microscopy. Various image preprocessing techniques are applied to extract a variety of features characterizing P. minimum cells and cell contours. Relevant feature subsets are then selected and used in support vector machine (SVM) as well as random forest (RF) classifiers to distinguish between P. minimum cells and other objects. To improve the cell abundance estimation accuracy, classification results are corrected based on probabilities of interclass misclassification. The developed algorithms were tested using 158 phytoplankton images. There were 920 P. minimum cells in the images in total. The algorithms detected 98.1% of P. minimum cells present in the images and correctly classified 98.09% of all detected objects. The classification accuracy of detected P. minimum cells was equal to 98.9%, yielding a 97.0% overall recognition rate of P. minimum cells. The feature set used in this work has shown considerable tolerance to out-of-focus distortions. Tests of the system by phytoplankton experts in the cell abundance estimation task of P. minimum species have shown that its performance is comparable or even better than performance of phytoplankton experts exhibited in manual counting of artificial microparticles, similar to P. minimum cells. The automated system detected and correctly recognized 308 (91.1%) of 338 P. minimum cells found by experts in 65 phytoplankton images taken from new phytoplankton samples and erroneously assigned to the P. minimum class 3% of other objects. Note that, due to large variations of texture and size of P. minimum cells as well as background, the task performed by the system was more complex than that performed by the experts when counting artificial microparticles similar to P. minimum cells.

show abstract

Automatic analysis of aqueous specimens for phytoplankton structure recognition and population estimation

Cited by 43 publications

References 23 publications

PlanktoVision - an automated analysis system for the identification of phytoplankton

PlanktoVision - an automated analysis system for the identification of phytoplankton

Photosymbiosis in Marine Planktonic Protists

An Integrated Approach to Analysis of Phytoplankton Images

Contact Info

Product

Resources

About