Systematic Literature Review: Research Ethics in Artificial Intelligence Development for Breast Cancer Histopathology Detection – Data Privacy, Informed Consent, and Accountability
Abstrak
Breast cancer remains a leading cause of mortality in women worldwide. The development of Artificial Intelligence (AI) models, particularly Convolutional Neural Networks (CNN) and adaptive thresholding methods, has demonstrated significant potential in analyzing breast cancer histopathology images. However, developing these AI models requires large datasets containing sensitive patient data, thereby creating comprehensive research ethics challenges related to data privacy protection, informed consent implementation, and accountability clarity. This systematic literature review aims to identify current ethical practices, ethical challenges, and best recommendations regarding data privacy protection, informed consent implementation, and research accountability in AI model development for breast cancer histopathology detection. Study analysis encompassed identification, selection, eligibility assessment, and results synthesis stages using biomedical ethics frameworks (Belmont Principles and Four Principles Approach). The study identified three significant ethical gaps in current research practices: (1) data privacy protection remains suboptimal due to limited de-identification techniques and privacy-preserving methods; (2) informed consent implementation is still restricted in terms of comprehensiveness, comprehensibility, and voluntariness for AI-based research; and (3) institutional accountability remains unclear regarding responsibility allocation and oversight mechanisms. This systematic literature review concludes that a structured ethical framework and clear implementation guidelines are essential to ensure that AI research in breast cancer histopathology adheres to biomedical ethics principles while supporting sustainable and beneficial technological innovation for patients.
Identifier
- DOI: https://doi.org/10.33474/infotron.v5i2.24677
- OpenAlex: https://openalex.org/W7203528285