Cancer Detection Using Machine Learning

Abstract

Cancerhasbeenclassifiedintovarioussubtypes.So,itisessentialtodetectcancersymptomsearlyon. Artificial Intelligence(AI) and Machine Learning(ML) have been used to classify cancer categories. There are certain datasets and models capable of estimating key features. There are many methods used for the development of cancer detection techniques, some of the methods are artificial neural networks(ANNs), and decision trees(DTs).While considering these methods and understanding them, a valid procedure must be considered for their implementation. In this study, various ML and Deep Learning(DL) approaches are explained which can be used in cancer detection. Although there has been progress in the diagnosis andtreatingcancervictimswithpersonalization,itis difficult to provide cancer victims with data-driven care. To improve patient outcomes and medical efficiency, the application of Artificial Intelligence(AI) has become an effective means. Machine Learning provides an opportunity for systemstolearnbygainingknowledgefromlearning models, and this approach is very successful at forecasting different forms of cancer, among them relatedtoliver,lung,andothercancers.Professionals are not as precise in forecasting illness as machine learning and artificial intelligence are. The recent advancement in deep learning has transformedmedicalimaging,providingtoolsfor analyzing data automatically. Convolutional neural networks, also known as CNNs,havebeenemployedtoidentifyseveraltumor categoriessuchasMRIandCTscans.Inthisstudy, we shall learn about a structural framework that includes data preprocessing, model architecture design, and performance analysis. Deep learning models can achieve higher accuracy and sensitivity for the identification of symptoms relatedtocancerwithanefficientapproachcompared to traditional practices of treatment.

Introduction

Cancer is a global health concern with rising survival rates. It occurs when cells in the body begin to grow uncontrollably and spread, forming tumors. Tumors can be either malignant, spreading to other parts of the body (metastasis), or benign, which do not spread or recur once removed. There are over 100 types of cancer, classified based on the organ or tissue where they originate. Examples include carcinoma, sarcoma, and leukemia.

Key Cancer Types:

• Carcinoma: Most common, arises from epithelial cells (e.g., lung, breast, colon cancer).

• Sarcoma: Affects bone and soft tissues (e.g., osteosarcoma).

• Leukemia: Starts in the bone marrow, affecting blood cells.

Brain and spinal cord tumors are also significant and are classified based on the cell type they form. Lung cancer remains the leading cause of cancer-related deaths.

Causes of Cancer:

Cancer develops through genetic mutations and external agents like physical (radiation), chemical (tobacco), and biological (viruses, bacteria) carcinogens.

AI, Machine Learning (ML), and Deep Learning (DL) in Cancer Prediction:

AI refers to computer systems that analyze data to make predictions. ML, a subset of AI, improves with exposure to data and algorithms, making it useful for predicting and detecting cancer.

• Training Models: ML uses supervised (labeled data), unsupervised (unlabeled data), and reinforcement learning to train systems. Common algorithms include Artificial Neural Networks (ANN), Support Vector Machines (SVM), and Convolutional Neural Networks (CNN).

• AI/ML in Cancer: AI and ML help classify tumors, analyze medical imaging (e.g., MRI), and predict cancer outcomes such as survival rates and recurrence. SVM is especially useful in differentiating tumor types, while Natural Language Processing (NLP) analyzes patient records for early screening.

Applications of AI and ML:

AI improves early cancer detection, prognosis, and personalized treatment by identifying patterns in medical data. ML techniques like Decision Trees (DTs), ANNs, and SVM are employed to analyze tumor data, predict cancer survival, and detect recurrence. For example, ML models predict cancer susceptibility based on various factors like age, lifestyle, and environmental exposures.

Challenges and Opportunities:

• Early detection remains crucial but challenging, especially with small tumors.

• AI is being integrated into medical facilities to improve diagnosis, treatment, and patient monitoring.

• ML can automate diagnosis, making it faster and more accurate, as demonstrated in skin cancer and type 2 diabetes predictions.

Conclusion

ThestudyexplorestheconceptsofMLandits application in cancer-related estimation and prediction. Most studies were proposed focussing on creating predictive models to forecastdiseaseoutcomesutilizingsupervised ML methods and classification techniques. Applying different feature selection and classificationmethodstomultidimensional heterogeneous datasets can be enough to intervene in the cancer field.

References

[1] “WhatIsCancer?,”NationalCancerInstitute, Oct. 11, 2021. [2] B. Zhang, H. Shi, and H. Wang, “Machine LearningandAIinCancerPrognosis,Prediction,and Treatment Selection: A Critical Approach,” Machine Learning and AI in Cancer Prognosis, Prediction, and Treatment Selection: A Critical Approach, vol.Volume16,no.16,pp.1779–1791,Jun.2023. [3] WHO,“Cancer,”WorldHealthOrganization,Feb. 03, 2022. [4] K.Kourou,T.P.Exarchos,K.P.Exarchos,M.V. Karamouzis, and D. I. Fotiadis, “Machine learning applications in cancer prognosis and prediction,” Computational and Structural Biotechnology Journal, vol. 13, no. 13, pp. 8–17, 2015, doi: [5] Google,“WhatIsArtificialIntelligence(AI)?,” Google Cloud, 2023. [6] GeeksforGeeks, “Introduction to Convolution Neural Network,” GeeksforGeeks, Mar. 14, 2024. [7] S. Suman, “Brain Tumor classification and detectionfromMRIimagesusingCNNbasedon ResU-NetArchitecture,”Medium,Mar.03,2021. [8] S.Das,MohanKumarDey,RamDevireddy,and ManasRanjanGartia, “Biomarkers in Cancer Detection,Diagnosis,andPrognosis,”Sensors,vol. 24, no. 1, pp. 37–37, Dec. 2023. [9] “GoogleScholar,”Google.com,2024. . [10] “GoogleScholar,”Google.com,2024. [11] “GoogleScholar,”Google.com,2024. [12] “GoogleScholar,”Google.com,2024. [13] “GoogleScholar,”Google.com,2024. [14] D.V.Cicchetti,“NeuralNetworksandDiagnosis intheClinicalLaboratory:StateoftheArt,”Clinical Chemistry, vol. 38, no. 1, pp. 9–10, Jan. 1992. [15] “Google Scholar,” Google.com, 2024. [16] “Google Scholar,” Google.com, 2024. [17] “Google Scholar,” Google.com, 2024. [18] “Google Scholar,” Google.com, 2024. [19] “Google Scholar,” Google.com, 2024. [20] “Google Scholar,” Google.com, 2024. [21] “Google Scholar,” Google.com, 2024. [22] “GoogleScholar,”Google.com,2024. [23] “GoogleScholar,”Google.com,2024. [24] W.Li,Y.Zhang,andF.Chen,“ChatGPTinColorectal Surgery: A Promising Tool or a Passing Fad?,” Annals of Biomedical Engineering, May 2023. [25] S.Yang,Q.Li,W.Li,X.Li,andA.-A.Liu,“Dual-Level Representation Enhancement on CharacteristicandContextforImage-TextRetrieval,” IEEETransactionsonCircuitsandSystemsforvideo technology, vol. 32, no. 11, pp. 8037–8050, Nov.2022.

Copyright

Copyright © 2025 Ms. Poonam Chakravarty, Dhara Pandya , Mr. Jigar Pandya. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.