A shape context fully convolutional neural network for segmentation and classification of cervical nuclei in Pap smear images

Hussain, Elima; Mahanta, Lipi B.; Das, Chandana; Choudhury, Manjula; Chowdhury, Manish

doi:10.1016/j.artmed.2020.101897

Cited by 65 publications

(26 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…These datasets can be applied for automated machine learning approach for abnormality evaluation of stem cells. Hussain et al [136] has simultaneously conducted nuclei (nucleus) segmentation and classification from the cervical cancer morphological cell image using U-net architecture-based fully convolutional neural network (FCN). Figure 5 shows the nuclei image processing and machine learning classification architecture by Hussain et al [136].…”

Section: About Here]mentioning

confidence: 99%

“…Hussain et al [136] has simultaneously conducted nuclei (nucleus) segmentation and classification from the cervical cancer morphological cell image using U-net architecture-based fully convolutional neural network (FCN). Figure 5 shows the nuclei image processing and machine learning classification architecture by Hussain et al [136]. They adopted the shape representation model based on auto-encoders which act as a network regularizer to increase the strength and robustness of the FCN.…”

Section: About Here]mentioning

confidence: 99%

See 1 more Smart Citation

Machine Learning in Stem Cells Research: Application for Biosafety and Bioefficacy Assessment

et al. 2021

View full text Add to dashboard Cite

The applicability of machine learning-based analysis in the field of biomedical field has been very beneficial in determining the biological mechanism and validation for a wide range of biological scenarios. This approach is also gaining momentum in variou s stem cells research activities, specifically for stem cells characterization and differentiation pattern. The adoption of similar computational approaches to study and assess biosafety and bioefficacy risks of stem cells for clinical application is the n ext progression. In particular where tumorigenicity has been one of the major concerns in stem cells therapy. There are many factors influencing tumorigenicity in stem cells which may be difficult to capture under conventional laboratory settings. In addition, given the possible multifactorial etiology of tumorigenicity, defining a onesize-fits-all strategy to test such risk in stem cells might not be feasible and may compromise stem cells safety and effectiveness in therapy. Given the increase in biological datasets (which is no longer limited to genomic data) and the advancement of health informatics powered by state-of-the-art machine learning algorithms, there exists a potential for practical application in biosafety and bioefficacy of stem cells therap y. Here, we identified relevant machine learning approaches and suggested protocols intended for stem cells research focusing on the possibility of its usage for stem cells biosafety and bioefficacy assessment. Ultimately, generating models that may assist healthcare professionals to make a better-informed decision in stem cell therapy.

show abstract

Section: About Here]mentioning

confidence: 99%

Section: About Here]mentioning

confidence: 99%

Machine Learning in Stem Cells Research: Application for Biosafety and Bioefficacy Assessment

et al. 2021

View full text Add to dashboard Cite

show abstract

“…An enhanced Fuzzy C-Means algorithm was proposed by William et al [9] The Fully Convolutional Neural Network (FCN) was proposed by Hussain et al [11] for both segmentation and classification of Pap smear images. The feature reuse-ability property is ensured densely connected blocks that replace a number of convolutional layers in the standard Unit.…”

Section: Related Workmentioning

confidence: 99%

“…Confusion matrix received from MASO optimized DenseNet 121 modelThe state-of-art comparison results based on cervical cancer detection are depicted in Table7. We have selected five state-of-art methods such as the proposed MASO optimized DenseNet 121method with support vector machine[7], Enhanced Fuzzy C-Means (EFC-means) algorithm[9], Fully Convolutional Neural Network (FCN)[11], and CNN[12] methods. The proposed method achieved 98.38% accuracy,98.5%specificity,98.83% sensitivity,98.58% precision,99.3%recall and 98.25% F-score.…”

mentioning

confidence: 99%

Mutation based Atom Search Optimization Algorithm for Hyperparameter Optimization of the DenseNet121 Architecture for Cervical Cancer Classification

Chitra

Kumar

2021

Preprint

View full text Add to dashboard Cite

The cervical cancer patient’s death rate can be minimized by accurate and early detection of cervical cancer (CC). One of the popular techniques called the Pap test or Pap smear is widely used for the early detection of CC. The manual analysis consumed more time in the case of CC detection. The existing techniques met few shortcomings in terms of poor accuracy, more computational complexity, higher feature dimensionality, poor reliability, and higher time-consumption with poor hyperparameters optimization. Hence, the computer-aided diagnostic model provides reliable and accurate CC detection at the initial stage. In this paper, we proposed MASO optimized DenseNet 121 architecture for the early detection of cervical cancer. At first, different kinds of augmentation techniques such as horizontal flip, vertical flip, zooming, shearing, height shift, width shift, rotation, and brightness to increase the number of training samples. The Mutation based Atom Search Optimization (MASO) algorithm is established to optimize the hyperparameters in DenseNet 121 architecture suchnumber of neurons in the dense layer, learning rate value, and the batch sizes. Different kinds of performance metrics such as accuracy, specificity, sensitivity, precisions, recall, F-score, and confusion matrix evaluate the performance of MASO optimized DenseNet 121 architecture for CC detection. A single normal class with three abnormal classes namely Carcinoma, Light dysplastic, and Sever dysplastic were selected from the Hervel dataset for experimental investigation. The proposed MASO optimized DenseNet 121 architecture achieves 98.38% accuracy, 98.5% specificity, 98.83% sensitivity, 98.58% precision, 99.3% recall and 98.25% F-score values than other existing methods.

show abstract

“…Without exception, these systems have increased the sensitivity of CC screening and reduced the false negative rate. However, many studies only focused on a certain part of the automatic screening diagnosis of cervical Pap smear, such as using six different CNNs to classify cervical lesions 37 , or using different algorithms to segment cervical cell and nuclei 38 and detect and classify images of PAP smears 39 . A variety of automatic screening systems are able to identify suspicious intraepithelial lesion areas 31 , 40 and allow doctors to focus on those suspicious areas.…”

Section: Introductionmentioning

confidence: 99%

Hybrid AI-assistive diagnostic model permits rapid TBS classification of cervical liquid-based thin-layer cell smears

Zhu

Ong

et al. 2021

Nat Commun

View full text Add to dashboard Cite

Technical advancements significantly improve earlier diagnosis of cervical cancer, but accurate diagnosis is still difficult due to various factors. We develop an artificial intelligence assistive diagnostic solution, AIATBS, to improve cervical liquid-based thin-layer cell smear diagnosis according to clinical TBS criteria. We train AIATBS with >81,000 retrospective samples. It integrates YOLOv3 for target detection, Xception and Patch-based models to boost target classification, and U-net for nucleus segmentation. We integrate XGBoost and a logical decision tree with these models to optimize the parameters given by the learning process, and we develop a complete cervical liquid-based cytology smear TBS diagnostic system which also includes a quality control solution. We validate the optimized system with >34,000 multicenter prospective samples and achieve better sensitivity compared to senior cytologists, yet retain high specificity while achieving a speed of <180s/slide. Our system is adaptive to sample preparation using different standards, staining protocols and scanners.

show abstract

A shape context fully convolutional neural network for segmentation and classification of cervical nuclei in Pap smear images

Cited by 65 publications

References 32 publications

Machine Learning in Stem Cells Research: Application for Biosafety and Bioefficacy Assessment

Machine Learning in Stem Cells Research: Application for Biosafety and Bioefficacy Assessment

Mutation based Atom Search Optimization Algorithm for Hyperparameter Optimization of the DenseNet121 Architecture for Cervical Cancer Classification

Hybrid AI-assistive diagnostic model permits rapid TBS classification of cervical liquid-based thin-layer cell smears

Contact Info

Product

Resources

About