Table 1.

Barriers to AI Implementation [53]

Barrier	Comment
Lack of standardized data	Heterogeneity of data sources used for training and validation
Data-sharing limitations	High-quality datasets needed to ensure geographic, technical, and patient diversity
Educational barriers and physician hesitancy	Physician distrust, technophobia, liability concerns, and a fear that AI may replace physicians
Regulatory hurdles	Evolving regulatory approval process for software as a medical device; concerns with labeling for AI/ML-based devices
Cost barriers	Substantial up-front investment may be required to incorporate AI into clinical practice; financial incentives provided through reimbursement fee codes will be needed

AI, artificial intelligence; ML, machine learning.