DOI: https://doi.org/10.36347/sjams.2025.v13i02.040

Pages: 530-535

Esophageal and skin cancers are among the most challenging malignancies, with early detection critical for improving survival rates and reducing healthcare costs. This paper explores the role of artificial intelligence (AI) in the early detection of these cancers in the United States, synthesizing methodologies from two key studies. For esophageal cancer, advanced machine learning techniques like Random Forest and XGBoost are employed to analyze multimodal data, including medical imaging, electronic health records (EHRs), and genomic profiles, achieving 92% accuracy in detecting early-stage cancer. For skin cancer, convolutional neural networks (CNNs) are used to analyze dermoscopic images, achieving an 87% accuracy in identifying malignant lesions. The study highlights the design and implementation of AI-driven models, covering data preprocessing, feature engineering, and evaluation metrics while addressing challenges such as class imbalance and overfitting. The results demonstrate AI's potential to enhance diagnostic accuracy, scalability, and accessibility, particularly in underserved areas. However, data privacy, algorithm interpretability, and regulatory compliance must be addressed to integrate AI into healthcare systems fully. This paper asserts that AI-driven diagnostics hold immense promise for revolutionizing cancer detection and calls for further research to overcome existing limitations while ensuring equitable access to these transformative technologies, ultimately improving patient outcomes and reshaping the landscape of cancer care.