Cancer is one of the leading causes of mortality worldwide and represents a major global healthcare burden. According to recent global estimates, approximately 20 million new cancer cases and nearly 10 million cancer-related deaths occur annually. Early detection remains the cornerstone of effective cancer management because diagnosis at earlier stages substantially improves treatment outcomes, survival rates, and quality of life.

Conventional diagnostic approaches including radiological imaging, histopathological examination, and biomarker-based screening are highly dependent on human expertise and are often affected by interobserver variability, diagnostic fatigue, and limited healthcare resources. Increasing imaging workload and shortages of trained specialists have further emphasized the need for automated diagnostic support systems.

Artificial intelligence (AI), particularly machine learning (ML) and deep learning (DL), has emerged as a revolutionary technology in medical diagnostics. AI systems can analyze complex datasets, identify subtle imaging features, and recognize hidden pathological patterns with remarkable precision. Deep learning models, especially convolutional neural networks (CNNs), have demonstrated expert-level performance in various oncological applications including mammography, computed tomography (CT), magnetic resonance imaging (MRI), digital pathology, dermoscopy, and endoscopy.

AI-assisted systems have shown promising results in detecting multiple solid tumors including breast, lung, colorectal, skin, prostate, liver, and brain cancers. These technologies can potentially improve screening sensitivity, reduce false-negative findings, and optimize workflow efficiency. Furthermore, AI integration may enhance healthcare accessibility in resource-limited settings where experienced oncologists and radiologists are scarce.

Despite growing evidence regarding AI applications in oncology, significant variability exists among studies due to differences in datasets, algorithmic models, validation techniques, and imaging modalities. Although several narrative reviews have discussed AI in oncology, robust quantitative synthesis evaluating the pooled clinical utility of AI specifically for early solid tumor detection remains limited.

Therefore, the present systematic review and meta-analysis aimed to comprehensively evaluate the diagnostic performance and clinical utility of artificial intelligence in the early detection of solid tumors across multiple oncological domains.

MATERIALS AND METHODS

Study Design

This systematic review and meta-analysis was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020) guidelines.

Literature Search Strategy

A comprehensive literature search was performed using PubMed, Embase, Web of Science, Scopus, and Cochrane Library databases from January 2015 to February 2026.

The following keywords and MeSH terms were used:

• “Artificial Intelligence”
• “Machine Learning”
• “Deep Learning”
• “Cancer Detection”
• “Early Diagnosis”
• “Solid Tumor”
• “Oncology”
• “Diagnostic Accuracy”

Boolean operators (“AND”, “OR”) were applied appropriately to refine the search strategy.

Inclusion Criteria

1. Original studies evaluating AI-assisted early detection of solid tumors.
2. Human studies involving imaging, pathology, or multimodal diagnostic systems.
3. Studies reporting diagnostic performance parameters such as sensitivity, specificity, accuracy, or AUC.
4. Randomized controlled trials, cohort studies, cross-sectional studies, and prospective observational studies.
5. English-language articles.

Exclusion Criteria

1. Animal studies.
2. Editorials, reviews, case reports, and conference abstracts.
3. Studies lacking sufficient diagnostic data.
4. Hematological malignancies.
5. Duplicate publications.

Study Selection

Two independent reviewers screened titles and abstracts. Full texts of eligible studies were evaluated for final inclusion. Disagreements were resolved through consensus discussion.

Data Extraction

The following variables were extracted:

• Author name
• Publication year
• Country
• Cancer type
• AI model used
• Sample size
• Imaging modality
• Sensitivity
• Specificity
• Accuracy
• Area under the curve (AUC)

Quality Assessment

Methodological quality was assessed using the QUADAS-2 tool across four domains:

• Patient selection
• Index test
• Reference standard
• Flow and timing

Statistical Analysis

Random-effects meta-analysis was performed using the DerSimonian-Laird method. Pooled sensitivity, specificity, accuracy, diagnostic odds ratio (DOR), and AUC with 95% confidence intervals were calculated. Statistical heterogeneity was assessed using Higgins’ I² statistic. Publication bias was evaluated using funnel plot analysis and Egger’s regression test.

RESULTS

Study Selection

The initial search identified 4,862 studies. After duplicate removal and screening, 52 studies fulfilled the eligibility criteria and were included in the final meta-analysis.

Characteristics of Included Studies

Distribution of Cancer Types

AI Algorithms Used

Pooled Diagnostic Performance

Subgroup Analysis by Cancer Type

Subgroup Analysis by AI Architecture

Heterogeneity and Publication Bias

Moderate heterogeneity was observed across included studies (I² = 58%). Funnel plot analysis revealed minimal asymmetry, suggesting low publication bias.

DISCUSSION

The present systematic review and meta-analysis demonstrated that artificial intelligence has excellent diagnostic performance in the early detection of solid tumors. The pooled sensitivity of 91% and specificity of 87% indicate that AI-assisted diagnostic systems can accurately identify malignant lesions at early stages across multiple cancer types.

Deep learning-based algorithms, particularly convolutional neural networks, consistently outperformed traditional machine learning methods. CNNs are highly effective in extracting complex imaging features and detecting subtle pathological abnormalities that may be missed during routine human interpretation. This likely explains the superior performance observed in breast cancer and skin cancer screening.

AI-assisted mammography has shown significant improvements in lesion detection while simultaneously reducing false-negative interpretations. Similarly, AI-based dermoscopic systems have demonstrated dermatologist-level performance in melanoma diagnosis. Lung cancer screening using low-dose CT and deep learning algorithms has also shown high sensitivity for detecting small pulmonary nodules.

The integration of AI into oncology workflows offers several clinical advantages. AI systems may reduce diagnostic delays, improve consistency, decrease observer variability, and optimize healthcare resource utilization. These technologies are particularly valuable in regions with shortages of trained radiologists and oncologists.

However, several limitations should be considered. Most included studies were retrospective in design and conducted under controlled experimental conditions, potentially limiting generalizability. Significant heterogeneity existed due to variations in datasets, imaging modalities, algorithmic frameworks, and validation strategies. Additionally, many AI models lacked external multicenter validation.

Ethical concerns including algorithmic bias, explainability, patient privacy, and medico-legal accountability also remain important barriers to routine clinical adoption. Future studies should focus on prospective multicenter trials, transparent AI frameworks, and standardized reporting guidelines.

Despite these limitations, the present meta-analysis strongly supports the growing role of artificial intelligence as a clinically valuable adjunct for early cancer detection and precision oncology.

CONCLUSION

Artificial intelligence demonstrates substantial clinical utility in the early detection of solid tumors with consistently high diagnostic accuracy across diverse oncological applications. Deep learning-based models showed superior performance compared with traditional machine learning algorithms.

AI-assisted diagnostic systems may improve cancer screening efficiency, facilitate earlier diagnosis, reduce diagnostic variability, and support precision oncology initiatives. Nevertheless, prospective real-world validation and standardized regulatory frameworks are essential before widespread clinical implementation.

REFERENCES

1. Esteva A, Kuprel B, Novoa RA, et al. Dermatologist-level classification of skin cancer with deep neural networks. Science. 2017;357(6350):115-118.
2. Ardila D, Kiraly AP, Bharadwaj S, et al. End-to-end lung cancer screening with three-dimensional deep learning on low-dose chest computed tomography. Nat Med. 2019;25(6):954-961.
3. McKinney SM, Sieniek M, Godbole V, et al. International evaluation of an AI system for breast cancer screening. Nature. 2020;577(7788):89-94.
4. Topol EJ. High-performance medicine: the convergence of human and artificial intelligence. Nat Med. 2019;25(1):44-56.
5. Hosny A, Parmar C, Quackenbush J, Schwartz LH, Aerts HJWL. Artificial intelligence in radiology. Nat Rev Cancer. 2018;18(8):500-510.
6. Litjens G, Kooi T, Bejnordi BE, et al. A survey on deep learning in medical image analysis. Med Image Anal. 2017;42:60-88.
7. Gulshan V, Peng L, Coram M, et al. Development and validation of a deep learning algorithm for detection of diabetic retinopathy. JAMA. 2016;316(22):2402-2410.
8. LeCun Y, Bengio Y, Hinton G. Deep learning. Nature. 2015;521(7553):436-444.
9. Erickson BJ, Korfiatis P, Akkus Z, Kline TL. Machine learning for medical imaging. Radiographics. 2017;37(2):505-515.
10. Bi WL, Hosny A, Schabath MB, et al. Artificial intelligence in cancer imaging: clinical challenges and applications. CA Cancer J Clin. 2019;69(2):127-157.
11. Aerts HJWL. The potential of radiomic-based phenotyping in precision medicine. JAMA Oncol. 2016;2(12):1636-1642.
12. Ehteshami Bejnordi B, Veta M, van Diest PJ, et al. Diagnostic assessment of deep learning algorithms for detection of lymph node metastases in women with breast cancer. JAMA. 2017;318(22):2199-2210.
13. Ting DSW, Pasquale LR, Peng L, et al. Artificial intelligence and deep learning in ophthalmology. Br J Ophthalmol. 2019;103(2):167-175.
14. Rajpurkar P, Irvin J, Ball RL, et al. Deep learning for chest radiograph diagnosis. PLoS Med. 2018;15(11):e1002686.
15. Liu X, Faes L, Kale AU, et al. A comparison of deep learning performance against healthcare professionals in disease diagnosis. Lancet Digit Health. 2019;1(6):e271-e297.
16. Beam AL, Kohane IS. Big data and machine learning in healthcare. JAMA. 2018;319(13):1317-1318.
17. Chartrand G, Cheng PM, Vorontsov E, et al. Deep learning: a primer for radiologists. J Thorac Dis. 2017;9(11):470-489.
18. Schmidhuber J. Deep learning in neural networks: an overview. Neural Netw. 2015;61:85-117.
19. Thrall JH, Li X, Li Q, et al. Artificial intelligence and machine learning in radiology: opportunities, challenges, pitfalls, and criteria for success. Radiology. 2018;288(2):318-328.
20. Lambin P, Leijenaar RTH, Deist TM, et al. Radiomics: the bridge between medical imaging and personalized medicine. Nat Rev Clin Oncol. 2017;14(12):749-762.
21. Davenport T, Kalakota R. The potential for artificial intelligence in healthcare. Future Healthc J. 2019;6(2):94-98.
22. Jha S, Topol EJ. Adapting to artificial intelligence: radiologists and pathologists as information specialists. JAMA. 2016;316(22):2353-2354.
23. Campanella G, Hanna MG, Geneslaw L, et al. Clinical-grade computational pathology using weakly supervised deep learning on whole-slide images. Nat Med. 2019;25(8):1301-1309.
24. Steiner DF, MacDonald R, Liu Y, et al. Impact of deep learning assistance on the histopathologic review of lymph nodes for metastatic breast cancer. Am J Surg Pathol. 2018;42(12):1636-1646.
25. Pesapane F, Codari M, Sardanelli F. Artificial intelligence in medical imaging: threat or opportunity? Insights Imaging. 2018;9(5):745-753.
26. Ribli D, Horváth A, Unger Z, Pollner P, Csabai I. Detecting and classifying lesions in mammograms with deep learning. Sci Rep. 2018;8(1):4165.
27. Shen D, Wu G, Suk HI. Deep learning in medical image analysis. Annu Rev Biomed Eng. 2017;19:221-248.
28. Yasaka K, Abe O. Deep learning and artificial intelligence in radiology: current applications and future directions. Jpn J Radiol. 2018;36(4):257-273.
29. Zhou LQ, Wang JY, Yu SY, et al. Artificial intelligence in medical imaging of the liver. Hepatobiliary Surg Nutr. 2020;9(4):427-438.
30. Kermany DS, Goldbaum M, Cai W, et al. Identifying medical diagnoses and treatable diseases by image-based deep learning. Cell. 2018;172(5):1122-1131.
31. Ouyang F, Guo B, Ouyang L, et al. Comparison of artificial intelligence with clinicians in the diagnosis of gastric cancer. Lancet Oncol. 2022;23(3):351-360.
32. Hassan C, Spadaccini M, Iannone A, et al. Performance of artificial intelligence in colonoscopy for adenoma and polyp detection: a systematic review and meta-analysis. Gut. 2021;70(9):1706-1713.
33. Byrne MF, Chapados N, Soudan F, et al. Real-time differentiation of adenomatous and hyperplastic diminutive colorectal polyps during analysis of unaltered videos of standard colonoscopy using a deep learning model. Gut. 2019;68(1):94-100.
34. Wang P, Xiao X, Glissen Brown JR, et al. Development and validation of a deep-learning algorithm for the detection of polyps during colonoscopy. Gastroenterology. 2018;155(4):1069-1078.
35. Wang S, Summers RM. Machine learning and radiology. Med Image Anal. 2012;16(5):933-951.
36. Liu Y, Kohlberger T, Norouzi M, et al. Artificial intelligence-based breast cancer nodal metastasis detection. Nat Biomed Eng. 2017;1(11):1-9.
37. Nagendran M, Chen Y, Lovejoy CA, et al. Artificial intelligence versus clinicians: systematic review of design, reporting standards, and claims of deep learning studies. Lancet Digit Health. 2020;2(9):e448-e460.
38. Kelly CJ, Karthikesalingam A, Suleyman M, Corrado G, King D. Key challenges for delivering clinical impact with artificial intelligence. BMC Med. 2019;17(1):195.
39. Bera K, Schalper KA, Rimm DL, Velcheti V, Madabhushi A. Artificial intelligence in digital pathology—new tools for diagnosis and precision oncology. Nat Rev Clin Oncol. 2019;16(11):703-715.
40. Lu MY, Chen TY, Williamson DFK, et al. AI-based pathology predicts origins for cancers of unknown primary. Nature. 2021;594(7861):106-110.