Automatic detection and segmentation of biological objects in 2D and 3D image data is central for countless biomedical research questions to be answered. While many existing computational methods are used to reduce manual labeling time, there is still a huge demand for further quality improvements of automated solutions. In the natural image domain, spatial embedding-based instance segmentation methods are known to yield high-quality results, but their utility to biomedical data is largely unexplored. Here we introduce EmbedSeg, an embedding-based instance segmentation method designed to segment instances of desired objects visible in 2D or 3D biomedical image data. We apply our method to four 2D and seven 3D benchmark datasets, showing that we either match or outperform existing state-of-the-art methods. While the 2D datasets and three of the 3D datasets are well known, we have created the required training data for four new 3D datasets, which we make publicly available online. Next to performance, also usability is important for a method to be useful. Hence, EmbedSeg is fully open source (https://github.com/juglab/EmbedSeg), offering (i) tutorial notebooks to train EmbedSeg models and use them to segment object instances in new data, and (ii) a napari plugin that can also be used for training and segmentation without requiring any programming experience. We believe that this renders EmbedSeg accessible to virtually everyone who requires high-quality instance segmentations in 2D or 3D biomedical image data.
- MeSH
- Algorithms * MeSH
- Humans MeSH
- Microscopy * methods MeSH
- Image Processing, Computer-Assisted methods MeSH
- Imaging, Three-Dimensional methods MeSH
- Check Tag
- Humans MeSH
- Publication type
- Journal Article MeSH
BACKGROUND: Because of its non-destructive nature, label-free imaging is an important strategy for studying biological processes. However, routine microscopic techniques like phase contrast or DIC suffer from shadow-cast artifacts making automatic segmentation challenging. The aim of this study was to compare the segmentation efficacy of published steps of segmentation work-flow (image reconstruction, foreground segmentation, cell detection (seed-point extraction) and cell (instance) segmentation) on a dataset of the same cells from multiple contrast microscopic modalities. RESULTS: We built a collection of routines aimed at image segmentation of viable adherent cells grown on the culture dish acquired by phase contrast, differential interference contrast, Hoffman modulation contrast and quantitative phase imaging, and we performed a comprehensive comparison of available segmentation methods applicable for label-free data. We demonstrated that it is crucial to perform the image reconstruction step, enabling the use of segmentation methods originally not applicable on label-free images. Further we compared foreground segmentation methods (thresholding, feature-extraction, level-set, graph-cut, learning-based), seed-point extraction methods (Laplacian of Gaussians, radial symmetry and distance transform, iterative radial voting, maximally stable extremal region and learning-based) and single cell segmentation methods. We validated suitable set of methods for each microscopy modality and published them online. CONCLUSIONS: We demonstrate that image reconstruction step allows the use of segmentation methods not originally intended for label-free imaging. In addition to the comprehensive comparison of methods, raw and reconstructed annotated data and Matlab codes are provided.
- MeSH
- Algorithms MeSH
- Cell Fractionation methods MeSH
- Humans MeSH
- Microscopy methods MeSH
- Image Processing, Computer-Assisted MeSH
- Check Tag
- Humans MeSH
- Publication type
- Journal Article MeSH
- Review MeSH
Segmentation helps interpret imaging data in a biological context. With the development of powerful tools for automated segmentation, public repositories for imaging data have added support for sharing and visualizing segmentations, creating the need for interactive web-based visualization of 3D volume segmentations. To address the ongoing challenge of integrating and visualizing multimodal data, we developed Mol* Volumes and Segmentations (Mol*VS), which enables the interactive, web-based visualization of cellular imaging data supported by macromolecular data and biological annotations. Mol*VS is fully integrated into Mol* Viewer, which is already used for visualization by several public repositories. All EMDB and EMPIAR entries with segmentation datasets are accessible via Mol*VS, which supports the visualization of data from a wide range of electron and light microscopy experiments. Additionally, users can run a local instance of Mol*VS to visualize and share custom datasets in generic or application-specific formats including volumes in .ccp4, .mrc, and .map, and segmentations in EMDB-SFF .hff, Amira .am, iMod .mod, and Segger .seg. Mol*VS is open source and freely available at https://molstarvolseg.ncbr.muni.cz/.
Achieving a reliable and accurate biomedical image segmentation is a long-standing problem. In order to train or adapt the segmentation methods or measure their performance, reference segmentation masks are required. Usually gold-standard annotations, i.e. human-origin reference annotations, are used as reference although they are very hard to obtain. The increasing size of the acquired image data, large dimensionality such as 3D or 3D + time, limited human expert time, and annotator variability, typically result in sparsely annotated gold-standard datasets. Reliable silver-standard annotations, i.e. computer-origin reference annotations, are needed to provide dense segmentation annotations by fusing multiple computer-origin segmentation results. The produced dense silver-standard annotations can then be either used as reference annotations directly, or converted into gold-standard ones with much lighter manual curation, which saves experts' time significantly. We propose a novel full-resolution multi-rater fusion convolutional neural network (CNN) architecture for biomedical image segmentation masks, called DeepFuse, which lacks any down-sampling layers. Staying everywhere at the full resolution enables DeepFuse to fully benefit from the enormous feature extraction capabilities of CNNs. DeepFuse outperforms the popular and commonly used fusion methods, STAPLE, SIMPLE and other majority-voting-based approaches with statistical significance on a wide range of benchmark datasets as demonstrated on examples of a challenging task of 2D and 3D cell and cell nuclei instance segmentation for a wide range of microscopy modalities, magnifications, cell shapes and densities. A remarkable feature of the proposed method is that it can apply specialized post-processing to the segmentation masks of each rater separately and recover under-segmented object parts during the refinement phase even if the majority of inputs vote otherwise. Thus, DeepFuse takes a big step towards obtaining fast and reliable computer-origin segmentation annotations for biomedical images.
- MeSH
- Humans MeSH
- Neural Networks, Computer * MeSH
- Image Processing, Computer-Assisted * methods MeSH
- Check Tag
- Humans MeSH
- Publication type
- Journal Article MeSH
Increasing evidence shows that flaws in machine learning (ML) algorithm validation are an underestimated global problem. In biomedical image analysis, chosen performance metrics often do not reflect the domain interest, and thus fail to adequately measure scientific progress and hinder translation of ML techniques into practice. To overcome this, we created Metrics Reloaded, a comprehensive framework guiding researchers in the problem-aware selection of metrics. Developed by a large international consortium in a multistage Delphi process, it is based on the novel concept of a problem fingerprint-a structured representation of the given problem that captures all aspects that are relevant for metric selection, from the domain interest to the properties of the target structure(s), dataset and algorithm output. On the basis of the problem fingerprint, users are guided through the process of choosing and applying appropriate validation metrics while being made aware of potential pitfalls. Metrics Reloaded targets image analysis problems that can be interpreted as classification tasks at image, object or pixel level, namely image-level classification, object detection, semantic segmentation and instance segmentation tasks. To improve the user experience, we implemented the framework in the Metrics Reloaded online tool. Following the convergence of ML methodology across application domains, Metrics Reloaded fosters the convergence of validation methodology. Its applicability is demonstrated for various biomedical use cases.
- MeSH
- Algorithms * MeSH
- Image Processing, Computer-Assisted * MeSH
- Semantics MeSH
- Machine Learning MeSH
- Publication type
- Journal Article MeSH
- Review MeSH
xvi, 467 s. : il., tab. ; 25 cm
- MeSH
- Cardiovascular Diseases MeSH
- Brain physiology blood supply MeSH
- Autonomic Nervous System Diseases MeSH
- Central Nervous System Diseases MeSH
- Brain Diseases complications physiopathology MeSH
- Heart Diseases complications physiopathology MeSH
- Heart physiology innervation MeSH
- Publication type
- Congress MeSH
- Conspectus
- Patologie. Klinická medicína
- NML Fields
- neurologie
- kardiologie
- NML Publication type
- učebnice vysokých škol
Autoři demonstrují pět pozorování aneuryzmatických kostních cyst páteře, které byly uloženy třikrát v jejím hrudním úseku, v jednom případě ve výši Cl-2 a přilehlé části báze lehni, u jednoho nemocného byl postižen LS přechod s převládající lokalizací v segmentu S1. Shrnují nejdůležitější známky aneuryzmatických kostních cyst v MR obraze i s jejich společnými rysy s dalšími nádorovými i nenádorovými procesy páteře. Upozorňují na omezení MRI i dalších metod diagnostického zobrazování při typizaci těchto spinálních lézí i na obtíže při vymezování hranic mezi tzv. primárnimi a sekundárními aneuryzmatickými kostními cystami.
The authors demonstrate five observations of aneurysmal bone cysts of the spine which were in three instances in the thoracic portion, in one instance at the Cl-2 level and adjacent parts of the cranial base, in one patient the LS transition was affected with a predominating localisation in the SI segment. The authors summarize the most important signs of aneurysmal bone cysts in the MR image incl. features common with other tumourous and non-tumourous spinal processes. They draw attention to the limitations of MRI and other methods of diagnostic imaging in typing of these spinal lesions and to difficulties in defining the borderlines between so-called primary and secondary aneurysmal bone cysts.
- MeSH
- Diagnosis, Differential MeSH
- Adult MeSH
- Bone Cysts diagnosis pathology MeSH
- Humans MeSH
- Magnetic Resonance Imaging methods MeSH
- Giant Cell Tumors diagnosis pathology MeSH
- Spinal Neoplasms diagnosis pathology MeSH
- Osteoblastoma diagnosis pathology MeSH
- Osteoma, Osteoid diagnosis pathology MeSH
- Check Tag
- Adult MeSH
- Humans MeSH
- Female MeSH
- Publication type
- Case Reports MeSH
BACKGROUND: Genetic focal segmental glomerulosclerosis (FSGS) is caused by pathogenic variants in a broad spectrum of genes that have a variable representation based on subjects' ethnicity and/or age. The most frequently mutated autosomal recessive gene in FSGS is NPHS2. In this study, we analyzed the spectrum of NPHS2 variants and their associated phenotype in Czech adult FSGS patients. METHODS: A representative cohort of 234 adult patients with FSGS, derived from 225 families originating from all regions of Czechia, was analyzed by massively parallel sequencing. In this study, we focused on the comprehensive analysis of the NPHS2 gene. The histological classification of FSGS followed the Columbia classification. RESULTS: We detected seven (3%) cases bearing homozygous or compound heterozygous pathogenic NPHS2 variants. A single pathogenic variant c.868G > A (p.Val290Met) was found in the majority of NPHS2-positive cases (86%; 6 out of 7) in histologically confirmed instances of FSGS. Its allele frequency among unrelated NPHS2-associated FSGS patients was 50% (6/12), and Haplotype analysis predicted its origin to be a result of a founder effect. There is an identical V290M-related haplotype on all V290M alleles spanning a 0,7 Mb region flanking NPHS2 in Central European FSGS populations. The phenotype of the p.Val290Met NPHS2-associated FSGS demonstrated a later onset and a much milder course of the disease compared to other NPHS2 pathogenic variants associated with FSGS. The mean age of the FSGS diagnosis based on kidney biopsy evaluation was 31.2 ± 7.46 years. In 50% of all cases, the initial disease manifestation of proteinuria occurred only in adulthood, with 83% of these cases not presenting with edemas. One-third (33%) of the studied subjects progressed to ESRD (2 out of 6) at the mean age of 35.0 ± 2.82 years. CONCLUSIONS: We identified the most prevalent pathogenic variant, p.Val290Met, in the NPHS2 gene among Czech adult FSGS patients, which has arisen due to a founder effect in Central Europe. The documented milder course of the disease associated with this variant leads to the underdiagnosis in childhood. We established the histopathological features of the NPHS2-associated adult FSGS cases based on the Columbia classification. This might improve patient stratification and optimize their treatment.
- Publication type
- Journal Article MeSH
Computational fluid dynamics (CFD) has grown as a tool to help understand the hemodynamic properties related to the rupture of cerebral aneurysms. Few of these studies deal specifically with aneurysm growth and most only use a single time instance within the aneurysm growth history. The present retrospective study investigated four patient-specific aneurysms, once at initial diagnosis and then at follow-up, to analyze hemodynamic and morphological changes. Aneurysm geometries were segmented via the medical image processing software Mimics. The geometries were meshed and a computational fluid dynamics (CFD) analysis was performed using ANSYS. Results showed that major geometry bulk growth occurred in areas of low wall shear stress (WSS). Wall shape remodeling near neck impingement regions occurred in areas with large gradients of WSS and oscillatory shear index. This study found that growth occurred in areas where low WSS was accompanied by high velocity gradients between the aneurysm wall and large swirling flow structures. A new finding was that all cases showed an increase in kinetic energy from the first time point to the second, and this change in kinetic energy seems correlated to the change in aneurysm volume.
- Publication type
- Journal Article MeSH
Self-assembled bilayer structures such as those produced from amphiphilic block copolymers (polymersomes) are potentially useful in a wide array of applications including the production of artificial cells and organelles, nanoreactors, and delivery systems. These constructs are of important fundamental interest, and they are also frequently considered toward advances in bionanotechnology and nanomedicine. In this framework, membrane permeability is perhaps the most important property of such functional materials. Having in mind these considerations, we herein report the manufacturing of intrinsically permeable polymersomes produced using block copolymers comprising poly[2-(diisopropylamino)-ethyl methacrylate] (PDPA) as the hydrophobic segment. Although being water insoluble at pH 7.4, its pKa(PDPA) ∼ 6.8 leads to the presence of a fraction of protonated amino groups close to the physiological pH, thus conducting the formation of relatively swollen hydrophobic segments. Rhodamine B-loaded vesicles demonstrated that this feature confers inherent permeability to the polymeric membrane, which can still be modulated to some extent by the solution pH. Indeed, even at higher pH values where the PDPA chains are fully deprotonated, the experiments demonstrate that the membranes remain permeable. While membrane permeability can be, for instance, regulated by introducing membrane proteins and DNA nanopores, examples of membrane-forming polymers with intrinsic permeability have been seldom reported so far, and the possibility to regulate the flow of chemicals in these compartments by tuning block copolymer features and ambient conditions is of due relevance. The permeable nature of PDPA membranes possibly applies to a wide array of small molecules, and these findings can in principle be translocated to a variety of disparate bio-related applications.
