A Systematic Survey of Artificial Intelligence in Disease Detection and Diagnosis

In contemporary healthcare, artificial intelligence (AI) has quickly emerged as a key technology, especially for precise and early disease detection. Machine learning (ML), deep learning (DL), and hybrid models are highlighted in this survey's thorough analysis of current AI approaches used in disease diagnosis across various medical specialties. While Recurrent Neural Networks (RNNs) and Natural Language Processing (NLP) techniques are investigated for sequential and textual data interpretation, such as Electronic Health Records (EHRs), Convolutional Neural Networks (CNNs) and Transformer-based architectures are thoroughly examined for image-based detection tasks. The study also classifies AI applications by disease types, such as infectious, cardiology, neurology, and oncology, and assesses them using metrics like sensitivity, specificity, F1-score, and AUC. Important issues include cross-domain generalization, model interpretability, dataset imbalance, and regulatory compliance. There is also discussion of important issues like model interpretability, cross-domain generalization, dataset imbalance, and regulatory compliance. The survey concludes by outlining potential avenues for future research to improve the clinical applicability, fairness, and robustness of the model. Researchers and practitioners working on next-generation AI disease detection systems can use this work as a technical resource.