AI: The Effectiveness of Early Cancer Detection Programs Using AI Algorithms

Sara Pawlusek

Department of Science, College of West Anglia, King’s Lynn, UK.

DOI: 10.4236/abcr.2025.142004                

Abstract

Early detection is important for cancer control, improving patient survival and reducing unnecessary treatment. Methods for early detection of cancer include screening programs, risk stratification, and rapid diagnostic pathways. Nevertheless, limited diagnostic workforces, variability in the added value of screening efficiency, and delays in referral remain impediments to appropriate early detection, especially in post-pandemic healthcare environments. Machine learning (ML) and artificial intelligence (AI) have recently become transformative approaches that can help overcome these limitations for improved diagnostic accuracy, workflow efficiency, and patient selection for screening programs when applied to complex health data. Deep learning algorithms and convolutional neural networks are among the data-driven models that have yielded promising results in manipulating radiological images, pathology slides, and electronic health records for better risk stratification and early diagnosis. AI applications also help to identify cancer recurrence and automate clinical workflows where there are capacity limitations. Even considering these developments, sustainable research into ethical dilemmas, patient data security, biases in AI training datasets, and regulatory compliance represent key areas for ongoing investigations. This review discusses AI in the context of early cancer detection, its use in screening and diagnosis, and the barriers to broader clinical implementation.

Keywords

AI Algorithms, Early Cancer Detection

1. Introduction

Cancer is one of the leading health problems globally, with millions of new diagnoses each year. The earlier a malignancy is detected, the better the treatment success, significantly improving prognosis and decreasing overall healthcare expenditures [1]. Traditional methods of screening, such as mammography for breast cancer [2] and low-dose computed tomography (LDCT) for lung cancer [3], come with limitations, including variability in diagnosis, human error, and labor-intensive practices.

The explosive growth in artificial intelligence (AI)-specifically, machine-learning or deep-learning algorithms-has been viewed as a potential way to overcome these limitations [4]. AI systems are able to handle large datasets with high accuracy, potentially revolutionizing early detection methods. This systematic review aims to answer the following research question: “How do the diagnostic accuracy and efficacy of AI algorithms compare with traditional methods in breast and lung cancer screening?” Clarifying this point highlights the paper’s novel contribution to existing literature and informs potential policy implications and clinical adoption strategies.

2. Methodology

This review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines to promote transparency and reproducibility of the methods.

2.1. Search Strategy

A search on PubMed, IEEE Xplore, and Scopus databases was performed, focusing on journal articles published between January 2015 and December 2024. Keywords included: (“AI” OR “Artificial Intelligence” OR “machine learning” OR “deep learning”) AND (“breast cancer” OR “lung cancer”) AND (“early detection” OR “screening” OR “diagnosis”).

2.2. Inclusion Criteria

1) Studies that reported diagnostic accuracy, sensitivity, specificity, and clinical utility.

2) Studies discussing artificial intelligence in breast and lung cancer detection.

2.3. Exclusion Criteria

1) Studies primarily focusing on cancer treatment rather than detection.

2) Non-peer-reviewed publications, opinion papers, or purely theoretical studies.

2.4. Quality Assessment

The studies were systematically evaluated against standardized measures (e.g., area under the ROC curve, sensitivity, specificity). Quality appraisal focused on methodological transparency, reproducibility, representativeness of datasets, and generalizability of findings.

3. AI in Breast Cancer Detection

Improving Accuracy in Mammography

AI technologies can substantially improve mammographic performance and reduce the high false-positive and false-negative rates currently associated with screening [5]. Deep learning models may identify minute imaging anomalies that radiologists sometimes overlook, enabling earlier detection of breast cancer [6].

Critical Analysis and Case Studies

1) McKinney et al. demonstrated that a Google AI model significantly outperformed radiologists in terms of both sensitivity and specificity, greatly reducing false diagnoses [7].

2) Kim et al. reported increased early-stage detection and reduced unnecessary biopsies using a deep-learning approach [8]

3) Rodriguez-Ruiz et al. highlighted variability in AI performance among different demographic groups, underscoring the need for diverse validation datasets [9].

4) Schaffter et al. pointed out that institutional practices vary considerably, emphasizing the importance of standardized validation and implementation protocols [10].

5) Bahl stressed that AI should be integrated with clinical training to optimize real-world performance [11].

These results emphasize the critical necessity for validation in diverse patient cohorts. Patient perspectives, cost-effectiveness, and standardized clinical guidelines are also key for successful implementation [12].

4. AI in Lung Cancer Detection

With the development of AI technologies, the sensitivity of lung cancer detection using LDCT has significantly improved. AI reduces false-positive rates by accurately differentiating benign from malignant nodules, thereby lowering patient anxiety and avoiding unnecessary invasive procedures [13].

Lung cancer remains among the most aggressive cancer types worldwide, making early detection a pivotal factor in survival [3]. AI-based models that classify nodules as benign or malignant also incorporate patient risk factors to improve diagnostic precision [14] [15]. Research has shown that convolutional neural networks (CNNs) can outperform radiologists in detecting malignant nodules, leading to fewer false positives and unnecessary follow-ups [16] [17].

Critical Reviews and Case Studies:

1) Esteva et al. demonstrated that AI-driven CT diagnostics achieved higher sensitivity and specificity, delivering faster and more accurate diagnoses [18].

2) Ardila et al. found that validation in diverse populations is essential to ensure equitable outcomes [14].

3) Nam et al. reported improvements in patient comfort, diagnostic efficiency, and resource allocation [19].

4) Baldwin et al. highlighted ethical considerations such as algorithmic transparency, recommending guidelines that build both clinician and patient trust [15].

5) Christe et al. described variation in AI performance, advocating standardized training and validation to ensure safety in clinical use [16].

Ongoing challenges include the interpretability of AI models, dataset quality, and regulatory hurdles. Ensuring transparency, patient acceptance, and data security is paramount for broader AI adoption in lung cancer screening [20].

5. Issues and Ethical Considerations

The embedding of AI into cancer detection workflows presents several ethical and practical challenges. Data bias may result in unequal diagnostic accuracy across populations if models are trained on unrepresentative datasets [21]. Explainability is also crucial: clinicians must understand the rationale behind AI-driven recommendations to trust and effectively integrate these tools [22].

Additional concerns revolve around patient acceptance: some patients may be reluctant to trust automated systems for life-changing diagnoses. Privacy and data security are also paramount, as large patient datasets are required to train robust AI models [23]. Finally, financial and operational considerations such as initial development costs and integration into existing clinical pathways can pose barriers to widespread implementation [24].

6. Conclusion

AI algorithms represent a breakthrough in breast and lung cancer screening, enhancing diagnostic accuracy, minimizing errors, and streamlining clinical workflows [25]. Nevertheless, ethical, methodological, and practical barriers remain. Multi-center validations, efforts toward explainable AI, and robust policy frameworks will be necessary to ensure equitable and generalizable outcomes [26]. Collaborative efforts among clinicians, AI experts, ethicists, policymakers, and regulatory bodies are indispensable for realizing the full potential of AI in oncology [27]-[30].

Conflicts of Interest

The author declares no conflicts of interest regarding the publication of this paper.

References

[1]	McPhail, S., Johnson, S., Greenberg, D., Peake, M. and Rous, B. (2015) Stage at Diagnosis and Early Mortality from Cancer in England. British Journal of Cancer, 112, S108-S115. https://doi.org/10.1038/bjc.2015.49
[2]	Blandin Knight, S., Crosbie, P.A., Balata, H., Chudziak, J., Hussell, T. and Dive, C. (2017) Progress and Prospects of Early Detection in Lung Cancer. Open Biology, 7, Article ID: 170070. https://doi.org/10.1098/rsob.170070
[3]	National Cancer Registration and Analysis Service: Staging Data in England.
[4]	Sasieni, P. (2003) Evaluation of the UK Breast Screening Programmes. Annals of Oncology, 14, 1206-1208. https://doi.org/10.1093/annonc/mdg325
[5]	Maroni, R., Massat, N.J., Parmar, D., Dibden, A., Cuzick, J., Sasieni, P.D., et al. (2020) A Case-Control Study to Evaluate the Impact of the Breast Screening Programme on Mortality in England. British Journal of Cancer, 124, 736-743. https://doi.org/10.1038/s41416-020-01163-2
[6]	Esserman, L.J., Anton-Culver, H., Borowsky, A., Brain, S., Cink, T., Crawford, B., et al. (2017) The WISDOM Study: Breaking the Deadlock in the Breast Cancer Screening Debate. npj Breast Cancer, 3, Article No. 34. https://doi.org/10.1038/s41523-017-0035-5
[7]	McKinney, S.M., Sieniek, M., Godbole, V., Godwin, J., Antropova, N., Ashrafian, H., et al. (2020) International Evaluation of an AI System for Breast Cancer Screening. Nature, 577, 89-94. https://doi.org/10.1038/s41586-019-1799-6
[8]	Kim, H., Kim, H.H., Han, B., Kim, K.H., Han, K., Nam, H., et al. (2020) Changes in Cancer Detection and False-Positive Recall in Mammography Using Artificial Intelligence: A Retrospective, Multireader Study. The Lancet Digital Health, 2, e138-e148. https://doi.org/10.1016/s2589-7500(20)30003-0
[9]	Rodriguez-Ruiz, A., Lång, K., Gubern-Merida, A., Broeders, M., Gennaro, G., Clauser, P., et al. (2019) Stand-alone Artificial Intelligence for Breast Cancer Detection in Mammography: Comparison with 101 Radiologists. JNCI: Journal of the National Cancer Institute, 111, 916-922. https://doi.org/10.1093/jnci/djy222
[10]	Schaffter, T., Buist, D.S.M., Lee, C.I., Nikulin, Y., Ribli, D., Guan, Y., et al. (2020) Evaluation of Combined Artificial Intelligence and Radiologist Assessment to Interpret Screening Mammograms. JAMA Network Open, 3, e200265. https://doi.org/10.1001/jamanetworkopen.2020.0265
[11]	Bahl, M., Barzilay, R., Yedidia, A.B., Locascio, N.J., Yu, L. and Lehman, C.D. (2018) High-Risk Breast Lesions: A Machine Learning Model to Predict Pathologic Upgrade and Reduce Unnecessary Surgical Excision. Radiology, 286, 810-818. https://doi.org/10.1148/radiol.2017170549
[12]	Dembrower, K., Wåhlin, E., Liu, Y., Salim, M., Smith, K., Lindholm, P., et al. (2020) Effect of Artificial Intelligence-Based Triaging of Breast Cancer Screening Mammograms on Cancer Detection and Radiologist Workload: A Retrospective Simulation Study. The Lancet Digital Health, 2, e468-e474. https://doi.org/10.1016/s2589-7500(20)30185-0
[13]	National Health Service (NHS): NHS Long Term Plan: Cancer. https://www.longtermplan.nhs.uk/areas-of-work/cancer/
[14]	Ardila, D., Kiraly, A.P., Bharadwaj, S., Choi, B., Reicher, J.J., Peng, L., et al. (2019) End-to-End Lung Cancer Screening with Three-Dimensional Deep Learning on Low-Dose Chest Computed Tomography. Nature Medicine, 25, 954-961. https://doi.org/10.1038/s41591-019-0447-x
[15]	Baldwin, D.R., Gustafson, J., Pickup, L., Arteta, C., Novotny, P., Declerck, J., et al. (2020) External Validation of a Convolutional Neural Network Artificial Intelligence Tool to Predict Malignancy in Pulmonary Nodules. Thorax, 75, 306-312. https://doi.org/10.1136/thoraxjnl-2019-214104
[16]	Christe, A., Peters, A., Drakopoulos, D., Heverhagen, J.T., Geiger, A., Bartholet, C., Woller, S., Kraemer, A. and Poellinger, A. (2020) Computer-Aided Diagnosis of Pulmonary Nodules in Chest CT: Effect of Nodule Characteristics on Detection Performance. European Radiology, 30, 369-383.
[17]	Farjah, F., Halgrim, S., Buist, D.S.M., Gould, M.K., Zeliadt, S.B., Loggers, E.T., et al. (2016) An Automated Method for Identifying Individuals with a Lung Nodule Can Be Feasibly Implemented across Health Systems. eGEMs (Generating Evidence & Methods to Improve Patient Outcomes), 4, 15. https://doi.org/10.13063/2327-9214.1254
[18]	Esteva, A., Kuprel, B., Novoa, R.A., Ko, J., Swetter, S.M., Blau, H.M., et al. (2017) Dermatologist-Level Classification of Skin Cancer with Deep Neural Networks. Nature, 542, 115-118. https://doi.org/10.1038/nature21056
[19]	Nam, J.G., Park, S., Hwang, E.J., Lee, J.H., Jin, K.N., Lim, K.Y., Vu, T.H., Sohn, J.H., Hwang, S., Goo, J.M., et al. (2019) Development and Validation of a Deep Learning Model for Screening Pneumonia on Chest Radiograph. Radiology, 292, 211-220.
[20]	Baldwin, D.R., Breathnach, O., Clifton, J., Chavaditiklu, M., Gleeson, F. and Jenkins, D. (2021) The Impact of Artificial Intelligence on Improving the Evidence-Based Use of Imaging: An Expert Opinion. British Journal of Radiology, 94, Article ID: 20210536.
[21]	Cirillo, D., Catuara-Solarz, S., Morey, C., Guney, E., Subirats, L., Mellino, S., et al. (2020) Sex and Gender Differences and Biases in Artificial Intelligence for Biomedicine and Healthcare. npj Digital Medicine, 3, Article No. 81. https://doi.org/10.1038/s41746-020-0288-5
[22]	Morley, J., Machado, C.C.V., Burr, C., Cowls, J., Joshi, I., Taddeo, M., et al. (2020) The Ethics of AI in Health Care: A Mapping Review. Social Science & Medicine, 260, Article ID: 113172. https://doi.org/10.1016/j.socscimed.2020.113172
[23]	Hindocha, S. and Badea, C. (2021) Moral Exemplars for the Virtuous Machine: The Clinician’s Role in Ethical Artificial Intelligence for Healthcare. AI and Ethics, 2, 167-175. https://doi.org/10.1007/s43681-021-00089-6
[24]	Anderson, M., O’Neill, C., Macleod Clark, J., Street, A., Woods, M., Johnston-Webber, C., et al. (2021) Securing a Sustainable and Fit-for-Purpose UK Health and Care Workforce. The Lancet, 397, 1992-2011. https://doi.org/10.1016/s0140-6736(21)00231-2
[25]	National Health Service (NHS): Digital Transformation of Screening-NHSX. https://www.nhsx.nhs.uk/key-tools-and-info/digital-transformation-of-screening/
[26]	Shaheen, N.J., Falk, G.W., Iyer, P.G. and Gerson, L.B. (2016) ACG Clinical Guideline: Diagnosis and Management of Barrett’s Esophagus. American Journal of Gastroenterology, 111, 30-50. https://doi.org/10.1038/ajg.2015.322
[27]	World Health Organization (WHO) (2021) Ethics and Governance of Artificial Intelligence for Health: WHO Guidance. 1-148.
[28]	Moore, C.R., Farrag, A. and Ashkin, E. (2017) Using Natural Language Processing to Extract Abnormal Results from Cancer Screening Reports. Journal of Patient Safety, 13, 138-143. https://doi.org/10.1097/pts.0000000000000127
[29]	Nayor, J., Borges, L.F., Goryachev, S., Gainer, V.S. and Saltzman, J.R. (2018) Natural Language Processing Accurately Calculates Adenoma and Sessile Serrated Polyp Detection Rates. Digestive Diseases and Sciences, 63, 1794-1800. https://doi.org/10.1007/s10620-018-5078-4
[30]	Glaser, A.P., Jordan, B.J., Cohen, J., Desai, A., Silberman, P. and Meeks, J.J. (2018) Automated Extraction of Grade, Stage, and Quality Information from Transurethral Resection of Bladder Tumor Pathology Reports Using Natural Language Processing. JCO Clinical Cancer Informatics, 2, 1-8. https://doi.org/10.1200/cci.17.00128