Machine Learning Methods for Automated Quantification of Ventricular Dimensions

Schutera, Mark; Just, Steffen; Gierten, Jakob; Mikut, Ralf; Reischl, Markus; Pylatiuk, Christian

doi:10.1089/zeb.2019.1754

Cited by 12 publications

(13 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The zebrafish has emerged as a very powerful tool for investigating the in vivo developmental toxicity of compounds on a large scale [33][34][35]. The small size, ease of genetic manipulations, and relatively economical cost has paved the way for zebrafish to be the best organism for human disease models [36][37][38][39][40][41][42][43][44][45][46][47][48][49][50][51][52][53]. The zebrafish developmental toxicity study revealed the chronic toxicity of CM1 and CM2.…”

Section: Discussionmentioning

confidence: 99%

Cytotoxic Evaluation and Anti-Angiogenic Effects of Two Furano-Sesquiterpenoids from Commiphora myrrh Resin

et al. 2020

View full text Add to dashboard Cite

Commiphora myrrh resin (Myrrh) has been used in traditional Arabic medicine to treat various inflammatory diseases. Two furano-sesquiterpenoids, 2-methoxyfuranodiene (CM1) and 2-acetoxyfuranodiene (CM2), were isolated from the chloroform fraction of the ethanolic extract of Arabic Commiphora myrrh resin. The cytotoxicity of the compounds was evaluated using human liver carcinoma, breast cancer cells (HepG2 and MCF-7, respectively) and normal human umbilical vein endothelial cells (HUVECs) cell lines. The development toxicity and anti-angiogenic activity of both compounds were also evaluated using zebrafish embryos. Cell survival assays demonstrated that both compounds were highly cytotoxic in HepG2 and MCF7 cells, with IC50 values of 3.6 and 4.4 µM, respectively. Both compounds induced apoptosis and caused cell cycle arrest in treated HepG2 cells, which was observed using flow cytometric analysis. The development toxicity in zebrafish embryos showed the chronic toxicity of both compounds. The toxicity was only seen when the embryos remained exposed to the compounds for more than three days. The compound CM2 showed a significant level of anti-angiogenic activity in transgenic zebrafish embryos at sublethal doses. Thus, we demonstrated the cytotoxic properties of both compounds, suggesting that the molecular mechanism of these compounds should be further assessed.

show abstract

Section: Discussionmentioning

confidence: 99%

Cytotoxic Evaluation and Anti-Angiogenic Effects of Two Furano-Sesquiterpenoids from Commiphora myrrh Resin

et al. 2020

View full text Add to dashboard Cite

show abstract

“…The architecture was initially motivated by semantic segmentation tasks on small-scale biomedical image datasets. This work builds upon previous heart segmentation experiments with the U-Net [ 3 ]. The architecture is extended to expect an RGB image, not a grayscale image, as an input and to classify multiple classes instead of just one.…”

Section: Methods and Datamentioning

confidence: 99%

“…The dataset X of this work is an extension of the heartSeg dataset [ 3 ]. Each sample x ∈ X is an RGB image capturing the heart region of Medaka (Oryzias latipes) hatchlings from a constant ventral view.…”

Section: Methods and Datamentioning

confidence: 99%

See 1 more Smart Citation

Methods for the frugal labeler: Multi-class semantic segmentation on heterogeneous labels

Schutera¹,

Rettenberger²,

Reischl³

2022

PLoS ONE

Self Cite

View full text Add to dashboard Cite

Deep learning increasingly accelerates biomedical research, deploying neural networks for multiple tasks, such as image classification, object detection, and semantic segmentation. However, neural networks are commonly trained supervised on large-scale, labeled datasets. These prerequisites raise issues in biomedical image recognition, as datasets are generally small-scale, challenging to obtain, expensive to label, and frequently heterogeneously labeled. Furthermore, heterogeneous labels are a challenge for supervised methods. If not all classes are labeled for an individual sample, supervised deep learning approaches can only learn on a subset of the dataset with common labels for each individual sample; consequently, biomedical image recognition engineers need to be frugal concerning their label and ground truth requirements. This paper discusses the effects of frugal labeling and proposes to train neural networks for multi-class semantic segmentation on heterogeneously labeled data based on a novel objective function. The objective function combines a class asymmetric loss with the Dice loss. The approach is demonstrated for training on the sparse ground truth of a heterogeneous labeled dataset, training within a transfer learning setting, and the use-case of merging multiple heterogeneously labeled datasets. For this purpose, a biomedical small-scale, multi-class semantic segmentation dataset is utilized. The heartSeg dataset is based on the medaka fish’s position as a cardiac model system. Automating image recognition and semantic segmentation enables high-throughput experiments and is essential for biomedical research. Our approach and analysis show competitive results in supervised training regimes and encourage frugal labeling within biomedical image recognition.

show abstract

“…To further improve cardiac measurement endpoint, there are some other parameters such as stroke volume and cardiac output [12] which were not mentioned in previous studies. Currently, there are also several methods using manual counting [10,13,14], MATLAB [15][16][17][18], commercialized software [19][20][21], deep learning [22], and free software [11,23,24] already published for cardiac performance detection; however, these methods might be expensive, require some coding skills, or have limited measurement endpoints (summarized in Table 1).…”

Section: Introductionmentioning

confidence: 99%

Measurement of Multiple Cardiac Performance Endpoints in Daphnia and Zebrafish by Kymograph

et al. 2021

View full text Add to dashboard Cite

Cardiovascular disease (CVD) is the number one cause of death worldwide. This condition resulted in huge research on CVD increasing the need for animal models suitable for in vivo research. Daphnia and zebrafish are good animal models for cardiovascular research due to their relative body transparency and easy culture property. Several methods have been developed to conduct cardiac performance measurement in Daphnia and zebrafish. However, most of the methods are only able to obtain heartbeat rate. The other important cardiac endpoints like stroke volume, ejection fraction, fraction shortening, cardiac output, and heartbeat regularity must use other programs for measurement. To overcome this limitation, in this study, we successfully developed a one-stop ImageJ-based method using kymograph macros language that is able to obtain multiple cardiac performance endpoints simultaneously for the first time. To validate its utility, we incubated Daphnia magna at different ambient temperatures and exposed zebrafish with astemizole to detect the corresponding cardiac performance alterations. In summary, the kymograph method reported in this study provides a new, easy to use, and inexpensive one-stop method obtaining multiple cardiac performance endpoints with high accuracy and convenience.

show abstract

Machine Learning Methods for Automated Quantification of Ventricular Dimensions

Cited by 12 publications

References 9 publications

Cytotoxic Evaluation and Anti-Angiogenic Effects of Two Furano-Sesquiterpenoids from Commiphora myrrh Resin

Cytotoxic Evaluation and Anti-Angiogenic Effects of Two Furano-Sesquiterpenoids from Commiphora myrrh Resin

Methods for the frugal labeler: Multi-class semantic segmentation on heterogeneous labels

Measurement of Multiple Cardiac Performance Endpoints in Daphnia and Zebrafish by Kymograph

Contact Info

Product

Resources

About