Comet Assay Classification for Buccal Mucosa’s DNA Damage Measurement with Super Tiny Dataset Using Transfer Learning

Afiahayati, FNU; Anarossi, Edgar; Yanuaryska, Ryna Dwi; Nuha, Fajar Ulin; Mulyana, Sri

doi:10.1007/978-3-030-14132-5_22

Cited by 4 publications

(4 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Since it is difficult to obtain external comet image datasets from other laboratories, each comet segmentation method might be optimized for their own comet images from a single laboratory. Previously reported comparison results of comet analysis programs 16 , 34 have shown a large gap in performance between their developed method and others when experimented with using their own dataset, which indicates the image variability issue. Given that there is no publicly available well-organized comet assay image dataset, we released our dataset to be utilized by researchers for fairer comparison between methods.…”

Section: Methodsmentioning

confidence: 94%

“…Next is the availability of large, well-organized, and publicly accessible image datasets, such as MNIST 29 , ImageNet 30 , and COCO 31 . Although there are a few reports applying deep learning on visual scoring automatically 32 – 34 , still none of the open-source software tools for comet assay analysis deployed DL methods to detect and segment comets so far. We believe that one of the primary reasons is the lack of suitable datasets to train a deep neural network.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Deep learning method for comet segmentation and comet assay image analysis

Hong¹,

Han

Lee

et al. 2020

Sci Rep

View full text Add to dashboard Cite

Comet assay is a widely used method, especially in the field of genotoxicity, to quantify and measure DNA damage visually at the level of individual cells with high sensitivity and efficiency. Generally, computer programs are used to analyze comet assay output images following two main steps. First, each comet region must be located and segmented, and next, it is scored using common metrics (e.g., tail length and tail moment). Currently, most studies on comet assay image analysis have adopted hand-crafted features rather than the recent and effective deep learning (DL) methods. In this paper, however, we propose a DL-based baseline method, called DeepComet, for comet segmentation. Furthermore, we created a trainable and testable comet assay image dataset that contains 1037 comet assay images with 8271 manually annotated comet objects. From the comet segmentation test results with the proposed dataset, the DeepComet achieves high average precision (AP), which is an essential metric in image segmentation and detection tasks. A comparative analysis was performed between the DeepComet and the state-of-the-arts automatic comet segmentation programs on the dataset. Besides, we found that the DeepComet records high correlations with a commercial comet analysis tool, which suggests that the DeepComet is suitable for practical application.

show abstract

Section: Methodsmentioning

confidence: 94%

Section: Introductionmentioning

confidence: 99%

Deep learning method for comet segmentation and comet assay image analysis

Hong¹,

Han

Lee

et al. 2020

Sci Rep

View full text Add to dashboard Cite

show abstract

“…A combination of CNN and SVM has been widely used to solve a classification task and showed a good performance in comparison with plain CNN [9][10][11]. However, most of the researches implemented 2-Dimensional CNN as they were dealing with image classification instead of sequence labelling.…”

Section: Introductionmentioning

confidence: 99%

Predicting Secondary Structure of Protein Using Hybrid of Convolutional Neural Network and Support Vector Machine

Sutanto¹,

Sukma²,

Afiahayati³

2021

IJIES

View full text Add to dashboard Cite

Protein secondary structure prediction is one of the problems in the Bioinformatics field, which conducted to find the function of proteins. Protein secondary structure prediction is done by classifying each sequence of protein primary structure into the sequence of protein secondary structure, which fall in sequence labelling problems and can be solved with the machine learning. Convolutional Neural Network (CNN) and Support Vector Machine (SVM) are 2 methods that often used to solve classification problems. In this research, we proposed a hybrid of 1-Dimensional CNN and SVM to predict the secondary structure of the protein. In this research, we used a novel hybrid 1-Dimensional CNN and SVM for sequence labelling, specifically to predict the secondary structure of the protein. Our hybrid model managed to outperform previous studies in term of Q3 and Q8 accuracy on CB513 dataset.

show abstract

“…These experimental conditions must be controlled adequately to get consistent results. 9 The percentage of DNA damage is directly proportional to the DNA migration in the comet assay. The main scoring approaches are visual scoring and computer-based image analysis.…”

mentioning

confidence: 99%

Automated Quantification of DNA Damage Using Deep Learning and Use of Synthetic Data Generated from Basic Geometric Shapes

Namuduri,

Mehta,

Barbe

et al. 2024

ECS Sens. Plus

View full text Add to dashboard Cite

Comet assays are used to assess the extent of Deoxyribonucleic acid (DNA) damage, in human cells, caused by substances such as novel drugs or nano materials. Deep learning is showing promising results in automating the process of quantifying the percentage of damage using the assay images. But the lack of large datasets and imbalanced data is a challenge. In this study, synthetic comet assay images generated from simple geometric shapes were used to augment the data for training the Convolutional Neural Network. The results from the model trained using the augmented data were compared with the results from a model trained exclusively on real images. It was observed that the use of synthetic data in training not only gave a significantly better coefficient of determination (R2), but also resulted in a more robust model i.e., with less variation in R2 compared to training without synthetic data. This approach can lead to improved training while using a smaller training dataset, saving cost and effort involved in capturing additional experimental images and annotating them. Additional benefits include addressing imbalanced datasets, and data privacy concerns. Similar approaches must be explored in other low data domains to extract the same benefits.

show abstract

Comet Assay Classification for Buccal Mucosa’s DNA Damage Measurement with Super Tiny Dataset Using Transfer Learning

Cited by 4 publications

References 9 publications

Deep learning method for comet segmentation and comet assay image analysis

Deep learning method for comet segmentation and comet assay image analysis

Predicting Secondary Structure of Protein Using Hybrid of Convolutional Neural Network and Support Vector Machine

Automated Quantification of DNA Damage Using Deep Learning and Use of Synthetic Data Generated from Basic Geometric Shapes

Contact Info

Product

Resources

About