Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 9s (2025) 2324 https://internationalpubls.com Review on Advancements and Challenges in Cervical Cancer Diagnosis S. Athinarayanan1, Dr. K. Dhanakodi2, Dr. A.T. Rajamanickam3, Dr. A. Sanjaygandhi4, Dr. S. Athinarayanan5, Dr. Kavitha Ramachandran6 1Research Scholar (Part time), 2,4Assistant Professor, 3Associate Professor Department of Electronics, Sri Ramakrishna Mission Vidhyalaya College of Arts and Science, Coimbatore, Tamilnadu-641 020 5Associate Professor, Department of CSE, Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology, Avadi, Chennai, Tamilnadu-600 062. 6Assistant Professor, Department of CSE, Vel Tech Hight Tech Dr. Rangarajan Dr. Sakunthala Engineering College, Avadi, Chennai, Tamilnadu-600 062. Article History: Received: 12-01-2025 Revised: 15-02-2025 Accepted: 01-03-2025 Abstract: Introduction: Cervical cancer is an important cancer in women. If it is detected in the early stages, proper treatment can provide relief. Various medical and technological advances are evolving to improve cancer diagnosis. Currently technologies such as Pap Smear, HPV Testing, MRI and Artificial Intelligence (AI) are used to diagnose the disease. In particular, modern methods such as machine learning and computer vision make it easier to diagnose cancer accurately. Despite these technological advances, various challenges remain. The main hurdles are the high cost of testing, lack of access to medical facilities, and lack of awareness in rural areas. Also, there are deficiencies in the accuracy and reliability of some diagnostic methods. There is a need to improve methods for faster and less effective diagnosis of cervical cancer. Soon, technological advancements will be important to develop ways to detect the disease at an early stage and make treatment available to all. Objectives: The primary objective of this study is to study and evaluate advanced methods of detecting cervical cancer. It is also important to analyse the benefits of new technologies and methods in the medical field. Methods: This research paper explores in detail the methods of diagnosing and studying susceptibility cancer. Various linguistic and modern machine learning techniques, use of clinical imaging and Pap smear test studies are mentioned in this article. Information on the accuracy, effectiveness and use of each method has been compiled. In addition, important recent research and developments also included. Results: The results of the study on the methods of diagnosis of Cervical Cancer show the effectiveness and accuracy of various techniques. A comparative assessment has been made of the effectiveness of machine learning models, and Pap Smear Image analysis methods. In particular, the results indicate that the use of artificial intelligence provides important improvements in diagnostic processes. In addition, technological improvements, shorter testing times, and more accurate diagnostic methods could improve ways to prevent the spread of infection, the study noted. Conclusions: This study on the methods of diagnosis and analysis of Cervical Cancer shows that modern technologies with high accuracy play a very important role in the medical field. Medical images, artificial intelligence, and machine learning techniques are helping to develop advanced diagnostic methods. Moreover, the studies carried out confirm the need to adopt new methods that reduce clinical trial time and increase diagnostic efficiency. The important conclusion is that in future, new research in this field will enhance clinical capacity and strengthen lifesaving efforts for early detection of diseases. Keywords Cervical Cancer, Pap Smear Test, Machine Learning, Deep Learning, Histology & Cytology. Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 9s (2025) 2325 https://internationalpubls.com 1. Introduction More reliable and advanced testing methods have gained increasing demand to boost both sensitivity and specificity in cervical cancer detection. New advances in artificial intelligence (AI) and machine learning (ML) demonstrate exceptional potential for transforming cervical cancer diagnosis and discrimination procedures. The use of AI platforms aided by deep learning methods shows exceptional performance in detecting normal from abnormal cells in cervical cytology samples and histopathological images. The new technologies outperform traditional methods by processing and interpreting large datasets at significant speeds hence providing improved diagnoses and lower occurrence of errors. Low-resource environments face challenges with AI solution scalability because they lack access both to sophisticated equipment and qualified staff. Cervical cancer detection research advances reveal the critical demand for better diagnostic methods going forward. The combination of AI solutions and advanced imaging technology presents a major opportunity to enhance cervical cancer screening operations worldwide through better efficiency and higher accuracy and broader accessibility. Research funding and technological developments will need focused attention because equitable and scalable solutions must become standard practice in low- resource setting environments to achieve measurable global health outcomes. Cervical cancer is one of the most common cancers in women and offers better treatment options when diagnosed at an early stage. With the advancement in medical technology, the methods of diagnosing this disease have been improved to operate with greater accuracy and speed. Traditional method of Smear tests and machine learning and artificial intelligence-based language systems improve diagnostic skills. This study examines current diagnostic methods, their benefits and frontiers aimed at cervical cancer, and talks about new proposed techniques for future clinical development. The Herlev Pap Smear Dataset provides an extensive database with 917 cervical cell images used for analysis in diagnosis studies. The dataset categorizes samples into seven groups: three normal (superficial squamous, intermediate squamous, and columnar epithelial cells) and four abnormal (mild dysplasia, moderate dysplasia, severe dysplasia, and carcinoma in situ). Machine learning and deep learning researchers find value in this dataset because it has detailed labels and high-resolution imagery. The dataset empowers cell classification research together with segmentation work and feature extraction thus advancing automated screening of cervical cancer. Data from the Herlev Dataset illustrating category images appears in Figure 1. (a) Normal Cell (b) Mild Dysplasia (c) Moderate Dysplasia (d)Severe Dysplasia (e) Carcinoma in Situ (CIS) Figure 1. Sample Herlev Dataset images Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 9s (2025) 2326 https://internationalpubls.com Objectives The main purpose of this article is to examine temporary and innovative methods used to accurately and quickly detect cervical cancer. By comparatively evaluating the effectiveness of methods such as medical image processing, biopsy analysis, smear testing and artificial intelligence-based linguistic techniques, it can be found which reference methods work with greater accuracy. In addition, the paper includes studies on preparations, preventive measures, and low-cost diagnostic methods to control disease. If cancer can be detected and treated at an early stage, the impact on life can be minimized. Through this, the purpose of this paper is to understand the impact of recent medical research and help find out how new technological solutions are useful to patients. 2. Methods A workflow method shows the stages of diagnosing Cervical cancer through Pap smear image assessment shown in Figure 2. Figure 2. Cervical cancer diagnosis using Pap smear images High-resolution microscopy allows researchers to visually confirm the presence of abnormalities yet human interpretation limits the process' consistency [1]. Digital imaging methods transform microscopic visual content into digital patterns through which AI systems function effectively. Applied preprocessing techniques which combine image enhancement with noise reduction methods help maintain clear images that will optimize segmentation and feature extraction. Advanced segmentation methods including U-Net and Mask R-CNN enable the separation of cellular entities and nuclei which nucleus segmentation stands as essential for detecting particular morphological characteristics linked to cancer development. Systems of feature extraction examine morphological features together with texture dynamics along with intensity measurements that Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 9s (2025) 2327 https://internationalpubls.com function as biomarkers for cell pathology identification [2]. The diagnostic accuracy of traditional manual methods significantly improves when cells are grouped as normal, precancerous, or cancerous through classification techniques that employ support vector machines (SVMs) along with convolutional neural networks (CNNs) and other machine learning and deep learning models [3]. 2.1 Commonly Available Pap Smear Datasets 2.1.1. Herlev Pap Smear Dataset Herlev provides a widely used dataset for Pap smear image analysis which includes 917 cervical cell images grouped as normal, mild dysplasia, moderate dysplasia and severe dysplasia types. The main purpose of this dataset lies in its ability to train and test machine learning algorithms when classifying cervical defects present in Pap smear slides. As a freely accessible repository for research use institutions in low-resource settings can benefit from its economic accessibility when developing AI- assisted diagnostic tools for academic and non-commercial applications. 2.1.2 Cervical Smear Dataset (from UCI Machine Learning Repository Researchers extract cell shape features from Pap smear images to establish patterns for cervical cancer screening diagnosis through the UCI dataset. The dataset has 858 records with twenty attributes which reveal clinical background and cervical screening test outputs making it appropriate for forecasting abnormalities. 2.1.3 Cervical Cancer Screening Dataset (SEER Database) The National Cancer Institute (NCI) of the United States created the SEER (Surveillance, Epidemiology, and End Results) Database, a clinical database for cancer. Important data including the risk of cancer, mortality rate, quality of life, risk factors, and preventative actions are stored there. This information is being used by physicians, researchers, and governments to enhance cancer treatment choices and prevalence. Through the SEER database, new cancer research will be conducted and public awareness will be increased. 2.1.4 Kaggle’s "Cervical Cancer Dataset" One medical resource that aids in analysing the causes of ovarian cancer is the Cervical Cancer (Risk Factors) dataset, which is accessible on Kaggle. This contains information about smoking, age, sexual preferences, HPV (human papillomavirus) infection, and other health conditions. This database is being used by researchers and specialists in computer displays to train machine learning algorithms to forecast the risk of ovarian cancer. This aids medical professionals in making early diagnoses and offering short-term fixes. 2.2 Herlev Dataset: Affordable Cervical Cancer Screening for All Cervical cancer may be detected using the Herlev Dataset, a medical database. Herlev Hospital in Denmark gave the microscopic photos of Papanicolaou (Pap) smear cells. This data is used by clinicians and machine learning researchers to develop extensive diagnostic models and information analysis, which aids in the creation of affordable diagnoses. Clinical research aiming at guaranteeing that women can get checked on time, especially in low-income countries, benefits greatly from this database. Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 9s (2025) 2328 https://internationalpubls.com 2.2.1 Herlev Dataset: A Solution for Low-Income Communities The Herlev Dataset is a low-cost medical database that facilitates the easy detection of cervical cancer. It aids in the separation of cancerous and healthy cells and includes delicate pictures of Pap smear cells from Herlev Hospital in Denmark. Given the lack of testing facilities and medical resources in low- income nations, this database uses AI and machine learning technology to provide accurate diagnoses more quickly and affordably. This will make it possible for all women to undergo routine screening for cancer in order to detect it early and take preventative action. 2.3 Advancements in Cervical Cancer Diagnosis Using Machine and Deep Learning Cervical cancer diagnosis has greatly advanced thanks to machine learning (ML) and deep learning (DL) technology. AI-based models carry out tasks including picture categorization, cell innocence analysis, and quicker, more accurate diagnosis in place of conventional Pap smear and HPV testing procedures. Doctors can identify ovarian cancer more quickly, more accurately, and at a reduced cost thanks to advanced handling technologies like Convolutional Neural Networks (CNN), Support Vector Machines (SVM), and Decision Trees. In low-resource hospitals, this has led to early cancer detection and a means of saving the lives of women. Deep learning approaches enhance cervical cancer examination through automated recognition of both pre-cancerous and cancerous cellular abnormalities within tissue studies. Deep learning through Convolutional Neural Networks (CNNs) achieved prominence over recent years because the systems can extract image-based features autonomously from unprocessed images. Convolutional Neural Networks process visual data to produce highly effective diagnostic tools primarily used in image-based diagnostics including Pap smear testing as well as histopathology image assessment. Pixel-level network analysis combined with detection capabilities allows these systems to find cell structure changes which standard human examination cannot identify. The large volume of annotated medical data enables these models to establish generalized patterns that deliver highly precise results thereby opening possibilities for modern diagnostic techniques to replace traditional approaches. Ensemble learning techniques have recently seen innovation to boost diagnostic accuracy by uniting multiple models into robust systems. Additionally, there have been innovations in combining ensemble learning techniques, which combine multiple models to increase diagnostic robustness and accuracy. Furthermore, transfer learning, where pre-trained models are fine-tuned with specific datasets, has helped reduce the need for vast amounts of labelled data and accelerated the development of diagnostic tools. Results Although there are many results in this review with respect to Herlev Pap Smear Dataset. Especially the researchers [4] and [5] found combined techniques of many researchers’ method based produced a new was very convincing to find this cervical cancer. They combined PCA feature extraction with SVM classification Arora et al. reached 95% achievement rate. The researchers at [6] used Relief to optimize the Random Forest classifier which achieved 94.4% accuracy in diagnosis. The findings confirm that selecting strong features leads to improved accuracy ratings in diagnostic systems. Cervical cancer diagnostic evaluation experienced a revolution because Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 9s (2025) 2329 https://internationalpubls.com deep learning technologies can identify medical image features automatically. The research of [7-8] and [9] proves deep neural networks effective for medical imaging classification tasks. The combination of applying fine-tuning to VGG16 and CaffeNet models along with pre-trained features delivered 99.5% accurate cervical cancer diagnoses accuracy reached 99.4% when Kamparia et al. applied CNN in combination with the hybrid DFCNN network structure to extract spatial and deep features. [10] approached delivers superior outcomes yet demands significant computational power alongside substantial training datasets for low-resource medical environments. The combination between various learning approaches through hybrid methods successfully enhances the prediction capabilities in classification processes. The ensemble methodology integrates multiple classifiers for improved data volatility defense through their inherent strengths. Fuzzy-based techniques adopted [11] resolved classification uncertainties through fuzzy networks which resulted in 92.03% accuracy. These contemporary approaches demonstrate high effectiveness yet their extensive computational expense hinders their deployment for wide-scale implementation. Feature transformation remains a leading approach to improve cervical cancer diagnostic accuracy. Image contrast enhancement through Histogram equalization achieved clustering success levels of 99.1% as reported by [12]. These methods reveal counting inaccuracy when exposed to noise because they need exact parameter adjustments to produce dependable outcomes. The researchers' effort produced great results: According to [13] their combination of DFCNN and MLP resulted in 99.32% accuracy and independently [13] achieved 94.5% accuracy with GoogleNet. The research team of [14] extracted features through DCT and Haar transformations leading to an 81.11% classification accuracy with RF analysis. 3.1 Summary and Analysis of Existing Review Studies The summary and analysis of existing review studies given in all referenced paper results [2-27] are presented in Table I as well as their results comparison illustrated in Figure3. Table I. Summary and Analysis of existing review studies Paper Findings Dataset Total No of Images Image Sizes Methodology Accu racy Limitations Arora et.al, 2021 Achieved high accuracy by using PCA for dimensionality reduction, improving classification Herlev Pap Smear Dataset 917 All images are different sizes SVM classifier with PCA-based feature extraction 95% PCA may fail to capture non- linear relationship is limited to features extracted. Sun et al., 2017 ReliefF helped select relevant features, improving RF performance Herlev Pap Smear Dataset 917 All images are different sizes RF with ReliefF 94.4% May require extensive parameter tuning; not ideal for very high- dimensional datasets. Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 9s (2025) 2330 https://internationalpubls.com Kalbhor et al., 2022 Leveraged DCT and Haar transforms for feature extraction, leading to moderate accuracy Herlev Pap Smear Dataset 917 All images are different sizes RF with DCT and Haar Transform 81.11% Lower accuracy compared to deep learning methods; performance depends on chosen transforms. Plissiti et al., 2018 Successfully localized and classified nuclei with Mask R-CNN Herlev Pap Smear Dataset 917 All images are different sizes Mask R-CNN 89.8 Computation ally intensive; requires large annotated training data for effective learning. Sompavon g et al., 2019 Enhanced Fuzzy C-Means provided robust clustering for accurate classification Herlev Pap Smear Dataset 917 All images are different sizes Enhanced Fuzzy C-Means 95.00 Sensitive to initialization ; performance drops with noisy data. Adem et al., 2019 GoogleNet achieved high accuracy leveraging pre- trained features Herlev Pap Smear Dataset 917 All images are different sizes GoogleNet 94.5% High computation al cost; requires significant memory and processing power for training. Lin et al., 2019 CNN showed poor performance due to small dataset size and lack of transfer learning Herlev Pap Smear Dataset 917 All images are different sizes CNN 68.0% Poor performance due to lack of data augmentatio n and advanced training techniques. Gorantla et al., 2019 Combining KNN, SVM, and DT improved classification accuracy Herlev Pap Smear Dataset 917 All images are different sizes KNN,SVM & DT 93% High computation al cost during inference due to ensemble; lacks scalability for larger datasets. Ghoneim et al., 2020 Fine-tuned VGG16 and CaffeNet achieved superior accuracy Herlev Pap Smear Dataset 917 All images are VCG16, CaffeNet 99.05% Overfitting risk with small datasets; Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 9s (2025) 2331 https://internationalpubls.com different sizes requires significant resources for fine-tuning. Xu et al., 2020 Combined classifiers delivered near- perfect accuracy with ensemble learning Herlev Pap Smear Dataset 917 All images are different sizes KNN,SVM & DT 99.27% Ensemble increases complexity; performance highly dependent on individual classifier quality. Kavitha et al., 2023 Histogram equalization enhanced image contrast, improving clustering accuracy Herlev Pap Smear Dataset 917 All images are different sizes Histogram Equivalization & Clustering 99.10% Effectivenes s depends on proper histogram adjustment; sensitive to noise in images. Fekri- Ershal et al., 2023 DFCNN enhanced the feature extraction for improved MLP performance Herlev Pap Smear Dataset 917 All images are different sizes MLP with DFCNN 99.32% Computation ally expensive; may overfit on small datasets without proper regularizatio n. Kamparia et al., 2021 Combined CNN and DFCNN achieved near- perfect accuracy Herlev Pap Smear Dataset 917 All images are different sizes CNN with DFCNN 99.40% Requires high computation al resources; training time increases significantly with complexity. Dongyova et.al 2020 GLCM features with SVM provided high classification accuracy Herlev Pap Smear Dataset 917 All images are different sizes SVM with GLCM 99.3% GLCM is sensitive to texture variation; may not generalize well across different datasets. Kalbhor et al., 2023 Achieved moderate accuracy; DFCNN combined with CNN improved classification Herlev Pap Smear Dataset 917 All images are different sizes CNN with DFCNN 87.3% Lower accuracy compared to other methods; needs further optimization for better results. https://arxiv.org/abs/2408.10498 https://arxiv.org/abs/2408.10498 Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 9s (2025) 2332 https://internationalpubls.com Chitra et al., 2024 CNN with DFCNN provided competitive accuracy Herlev Pap Smear Dataset 917 All images are different sizes CNN with DFCNN 98.38% Faces challenges with class imbalance; requires better data augmentatio n. Kallbhor & Sinde et al., 2023 Fuzzy Network improved accuracy by leveraging uncertainty in classification Herlev Pap Smear Dataset 917 All images are different sizes Fuzzy Network with DFCNN 92.03% Performance is sensitive to membership function parameters; computation al overhead is high. Kaur et al., 2022 Normalization and transformation improved CNN performance and classification accuracy Herlev Pap Smear Dataset 917 All images are different sizes CNN with Normalization and Transformation 99.36% Effectivenes s depends on proper normalizatio n; computation ally expensive for larger datasets. Ali et al., 2019 HMNN achieved moderate accuracy but struggled with complex feature relationships Herlev Pap Smear Dataset 917 All images are different sizes HMNN 88.41% Limited ability to handle non- linear features; performance highly dependent on data quality. Rahaman et al., 2021 HDFFN with DFCNN delivered high accuracy and robust feature extraction Herlev Pap Smear Dataset 917 All images are different sizes HDFFN with DFCNN 99.12% High resource requirements ; risk of overfitting on small or imbalanced datasets. Ghoneim et.al 2020 ELM with DFCNN achieved superior classification accuracy Herlev Pap Smear Dataset 917 All images are different sizes ELM with DFCNN 99.32% Computation ally intensive; requires careful tuning for optimal performance . Fekri et.al 2022 GSV and VCG enhanced classification, leading to Herlev Pap Smear Dataset 917 All images are VCG with GSV 88.47% Lower accuracy compared to deep Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 9s (2025) 2333 https://internationalpubls.com competitive accuracy different sizes learning approaches; performance limited by the GSV features. Chen et.al 2021 VGV with DFCNN provided good accuracy Herlev Pap Smear Dataset 917 All images are different sizes VGV with DFCNN 94.81 Performance depends on the quality of the VGV features; may struggle with large datasets. The Herlev Pap Smear Dataset remains essential for improving cervical cancer detection through research that tests deep learning methods and machine learning approaches. The implementation of CNNs together with ensemble learning and feature extraction methods reveals high precision yet the challenges for deployment in limited-resource environments require method optimization to overcome processing and expansion problems. Future research must establish effective lightweight models that utilize both data augmentation and transfer learning techniques for expanding equitable delivery of cervical cancer screening tools. Through these efforts low-income communities worldwide will gain access to both precise and cost-effective diagnostic solutions. Figure.3. Analysis of Accuracy Comparison by Researchers Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 9s (2025) 2334 https://internationalpubls.com 3.2 Extensive Index of Articles Published Across Various Databases on Cervical Cancer The Figure.4 illustrated the number of articles published on cervical cancer a how much research articles have been published in various popular databases in the last 10 years. Google Scholar has the maximum number of articles on this map around 375, followed by Scopus and IEEE Xplore with 300- 350 articles published, respectively. The Web of Science has about 275 articles, with about 200 articles featured in the MDPI database. There are very few about 150 articles in the Science Direct database. With this data, we can see that large databases like Google Scholar and Scopus have seen a large number of studies related to ovarian cancer. Moreover, it can be expected that research related to the diagnosis and treatment of this disease is also happening in technology-based databases like IEEE Xplore. Studies are less recorded in databases such as Science Direct and MDPI, but these are both clinical and scientific based research articles. Figure.4. Articles Published Across Various Databases 4.Conclusion Current diagnostic systems integrating sophisticated technologies with AI-powered methods create transformative changes to cervical cancer diagnosis and disease classification. The cervical cancer is a very dangerous cancer for women in today's time, so women know before that it is better to protect them from it. No matter how long this evolving science has been developing, to address this problem, to condemn and finding the solution for this cancer, to recover the huge strength of the women. In this paper, we have seen the results of many of our researchers results, each of them has found some techniques to detect cervical cancer in high-cost testing techniques, but a much better than of this, we have to find a good way to come up and these surveys will be useful to find a disease in very basic and less cost test of called Pap smear test. 0 50 100 150 200 250 300 350 400 Web of Science MDPI Google Scholar IEEE Xplore Science Direct Scopus No of Publications D at ab as e ARTICLES PUBLISHED IN LAST 1O YEARS Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 9s (2025) 2335 https://internationalpubls.com References [1] Sun G, Li S, Cao Y, Lang F. Cervical cancer diagnosis based on random forest. Int J Perform Eng. 2017;13(4):446. [2] Kalbhor M, Shinde SV, Jude H. Cervical cancer diagnosis based on cytology pap smear image classification using fractional coefficient and machine learning classifiers. TELKOMNIKA (Telecommun Comput Elect Control). 2022;20(5):1091–1102. [3] Lin H, Hu Y, Chen S, Yao J, Zhang L. Fine-grained classification of cervical cells using morphological and appearance-based convolutional neural networks. IEEE Access. 2019;7:71541–71549. [4] Ghoneim A, Muhammad G, Hossain MS. Cervical cancer classification using convolutional neural networks and extreme learning machines. Future Gener Comput Syst. 2020;102:643– 649. [5] Kavitha R, et al. Ant colony optimization enabled CNN deep learning technique for accurate detection of cervical cancer. Biomed Res Int. 2023;2023:1–9. [6] Fekri-Ershad S, Alsaffar MF. Developing a tuned three-layer perceptron fed with trained deep convolutional neural networks for cervical cancer diagnosis. Diagnostics. 2023;13(4):1–18. [7] Khamparia A, Gupta D, Rodrigues JJ, de Albuquerque VHC. DCAVN: Cervical cancer prediction and classification using deep convolutional and variational autoencoder network. Multimed Tools Appl. 2021;80:30399–30415. [8] Jia D, Li Z, Zhang C. Detection of cervical cancer cells based on strong feature CNN-SVM network. Neurocomputing. 2020;411:112–127. [9] Kalbhor M, et al. Pap smear-based cervical cancer detection using hybrid deep learning and performance evaluation. Comput Methods Biomech Biomed Eng Imaging Vis. 2023;11:1–10. [10] Chitra B, Kumar SS, Subbulekshmi D. Prediction models applying convolutional neural network-based deep learning to cervical cancer outcomes. IETE J Res. 2024:1–12. doi:10.1080/03772063.2024.2305831. [11] Kalbhor M, Shinde S. ColpoClassifier: a hybrid framework for classification of the cervigrams. Diagnostics. 2023;13(6):1103. doi:10.3390/diagnostics13061103. [12] Ali M, et al. Artificial neural network-based screening of cervical cancer using a hierarchical modular neural network architecture (HMNNA) and novel benchmark uterine cervix cancer database. Neural Comput Appl. 2019;31:2979–2993. [13] Rahaman MM, et al. Deep Cervix: A deep learning-based framework for the classification of cervical cells using hybrid deep feature fusion techniques. Comput Biol Med. 2021;136:104649. [14] Fekri-Ershad S, Ramakrishnan S. Cervical cancer diagnosis based on modified uniform local ternary patterns and feedforward multilayer network optimized by genetic algorithm. Comput Biol Med. 2022;144:105392. [15] Chen X, et al. Application of EfficientNet-B0 and GRU-based deep learning on classifying the colposcopy diagnosis of precancerous cervical lesions. Cancer Med. 2023;12(7):8690–89. doi:10.1002/cam4.5581. [16] Khare SK, et al. Emotion recognition and artificial intelligence: A systematic review (2014– 2023) and research recommendations. Inf Fusion. 2024;102:102019. Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 9s (2025) 2336 https://internationalpubls.com [17] Chauhan R, et al. Predictive modeling and web-based tool for cervical cancer risk assessment: a comparative study of machine learning models. MethodsX. 2024;12:102653. [18] Alias NA, et al. Pap Smear Image Analysis Based on Nucleus Segmentation and Deep Learning – A Recent Review. J Adv Res Appl Sci Eng Technol. 2023;29(3):37–47. [19] Arora A, Tripathi A, Bhan A. Classification of cervical cancer detection using machine learning algorithms. In: 2021 6th International Conference on Inventive Computation Technologies (ICICT). IEEE; 2021. p. 827–835. [20] Plissiti ME, et al. Sipakmed: A new dataset for feature and image-based classification of normal and pathological cervical cells in pap smear images. In: 2018 25th IEEE International Conference on Image Processing (ICIP). IEEE; 2018. p. 3144–3148. doi:10.1109/ICIP.2018.8451588. [21] Sompawong N, et al. Automated pap smear cervical cancer screening using deep learning. In: 2019 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC). IEEE; 2019. p. 7044–7048. [22] Adem K, Kilicarslan S, Comert O. Classification and diagnosis of cervical cancer with softmax classification with stacked autoencoder. 2019. [23] Gorantla R, Singh RK, Pandey R, Jain M. Cervical cancer diagnosis using CervixNet—a deep learning approach. In: 2019 IEEE 19th International Conference on Bioinformatics and Bioengineering (BIBE). IEEE; 2019. p. 397–404. [24] Kaur H, Sharma R, Kaur J. Classification of cervical cancer from pap smear images using deep learning: a comparison of transfer learning models. Res Square Preprint. 2024. doi:10.21203/rs.3.rs-3977123/v1. [25] Xu Y, et al. Machine learning algorithms for predicting the recurrence of stage IV colorectal cancer after tumor resection. Sci Rep. 2020;10(1):2519. [26] Alias NA, et al. Pap Smear Image Analysis Based on Nucleus Segmentation and Deep Learning – A Recent Review. J Adv Res Appl Sci Eng Technol. 2023;29(3):37–47. [27] S. Athinarayanan, M. V. Srinath, and R. Kavitha, "Multi-class cervical cancer classification by using ERSTCM, EMSD & CFE methods-based texture features and fuzzy logic-based hybrid kernel support vector machine classifier," IOSR Journal of Computer Engineering, vol. 19, pp. 23-34.