Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2023 Nov 10;8(7):538.
doi: 10.3390/biomimetics8070538.

Enhanced Pelican Optimization Algorithm with Deep Learning-Driven Mitotic Nuclei Classification on Breast Histopathology Images

Fadwa Alrowais  1 , Faiz Abdullah Alotaibi  2 , Abdulkhaleq Q A Hassan  3 , Radwa Marzouk  4 , Mrim M Alnfiai  5 , Ahmed Sayed  6
Affiliations

Enhanced Pelican Optimization Algorithm with Deep Learning-Driven Mitotic Nuclei Classification on Breast Histopathology Images

Fadwa Alrowais et al. Biomimetics (Basel). .

Abstract

Breast cancer (BC) is a prevalent disease worldwide, and accurate diagnoses are vital for successful treatment. Histopathological (HI) inspection, particularly the detection of mitotic nuclei, has played a pivotal function in the prognosis and diagnosis of BC. It includes the detection and classification of mitotic nuclei within breast tissue samples. Conventionally, the detection of mitotic nuclei has been a subjective task and is time-consuming for pathologists to perform manually. Automatic classification using computer algorithms, especially deep learning (DL) algorithms, has been developed as a beneficial alternative. DL and CNNs particularly have shown outstanding performance in different image classification tasks, including mitotic nuclei classification. CNNs can learn intricate hierarchical features from HI images, making them suitable for detecting subtle patterns related to the mitotic nuclei. In this article, we present an Enhanced Pelican Optimization Algorithm with a Deep Learning-Driven Mitotic Nuclei Classification (EPOADL-MNC) technique on Breast HI. This developed EPOADL-MNC system examines the histopathology images for the classification of mitotic and non-mitotic cells. In this presented EPOADL-MNC technique, the ShuffleNet model can be employed for the feature extraction method. In the hyperparameter tuning procedure, the EPOADL-MNC algorithm makes use of the EPOA system to alter the hyperparameters of the ShuffleNet model. Finally, we used an adaptive neuro-fuzzy inference system (ANFIS) for the classification and detection of mitotic cell nuclei on histopathology images. A series of simulations took place to validate the improved detection performance of the EPOADL-MNC technique. The comprehensive outcomes highlighted the better outcomes of the EPOADL-MNC algorithm compared to existing DL techniques with a maximum accuracy of 97.83%.

Keywords: artificial intelligence; bio-inspired algorithm; deep learning; medical imaging; mitotic nuclei classification.

PubMed Disclaimer

Conflict of interest statement

The authors declare that they have no conflict of interest. The manuscript was written with contributions from all authors. All authors have approved the final version of the manuscript.

Figures

Figure 1
Figure 1
Workflow of EPOADL-MNC system.
Figure 2
Figure 2
ANFIS structure.
Figure 3
Figure 3
(a,b) Confusion matrices with 80:20 of TR phase/TS phase and (c,d) 70:30 of TR phase/TS phase.
Figure 4
Figure 4
Average of EPOADL-MNC model with 80% of TR phase.
Figure 5
Figure 5
Average of EPOADL-MNC model with 20% of TS phase.
Figure 6
Figure 6
Average of EPOADL-MNC model in 70% of TR phase.
Figure 7
Figure 7
Average of EPOADL-MNC algorithm at 30% of TS phase.
Figure 8
Figure 8
Accuy curve of EPOADL-MNC model with 70:30 of TR phase/TS phase.
Figure 9
Figure 9
Loss curve of EPOADL-MNC algorithm in 70:30 of TR phase/TS phase.
Figure 10
Figure 10
PR curve of EPOADL-MNC system at 70:30 of TR phase/TS phase.
Figure 11
Figure 11
ROC curve of EPOADL-MNC model at 70:30 of TR phase/TS phase.
Figure 12
Figure 12
Accuy and precn outcome of EPOADL-MNC algorithm with recent approaches.
Figure 13
Figure 13
Recal and Fscore outcomes of EPOADL-MNC algorithm with recent approaches.
See this image and copyright information in PMC

References

    1. Syam S., Pushpalatha K.P. Machine Learning Based Classification and Grading of Breast Cancer; Proceedings of the 2023 International Conference on Innovations in Engineering and Technology (ICIET); Kerala, India. 13 July 2023; pp. 1–5.
    1. Zainudin Z., Shamsuddin S.M., Hasan S. Machine Learning and Big Data Analytics Paradigms: Analysis, Applications and Challenges. Springer; Cham, Switzerland: 2021. Deep layer convolutional neural network (CNN) Architecture for breast cancer classification using histopathological images; pp. 347–364.
    1. Razavi S., Khameneh F.D., Nouri H., Androutsos D., Done S.J., Khademi A. Minugan: Dual segmentation of mitoses and nuclei using conditional gans on multi-centre breast h&e images. J. Pathol. Inform. 2022;13:100002. - PMC - PubMed
    1. Mandair D., Reis-Filho J.S., Ashworth A. Biological insights and novel biomarker discovery through deep learning approaches in breast cancer histopathology. NPJ Breast Cancer. 2023;9:21. doi: 10.1038/s41523-023-00518-1. - DOI - PMC - PubMed
    1. Mehak S., Ashraf M.U., Zafar R., Alghamdi A.M., Alfakeeh A.S., Alassery F., Hamam H., Shafiq M. Automated Grading of Breast Cancer Histopathology Images Using Multilayered Autoencoder. Comput. Mater. Contin. 2022;71:3407–3423. doi: 10.32604/cmc.2022.022705. - DOI

LinkOut - more resources