Predicting the toxicity of nanoparticles using artificial intelligence tools: a systematic review

Area of Science:

• * Nanotechnology and Environmental Science
• * Toxicology and Computational Biology
• * Materials Science and Risk Assessment

Background:

• * Nanoparticle applications necessitate rigorous safety evaluations due to potential environmental and biological risks.
• * Experimental toxicity assessments are often costly and time-consuming, creating a need for efficient alternatives.
• * Artificial intelligence (AI) presents a promising approach for predicting nanomaterial toxicity.

Approach:

• * A systematic literature search was conducted across PubMed, Web of Science, and Scopus.
• * Inclusion and exclusion criteria were applied to select relevant studies on AI for nanoparticle toxicity.
• * Twenty-six studies focusing on metal oxide and metallic nanoparticles were included in the review.

Key Points:

• * Random Forest (RF) and Support Vector Machine (SVM) were the most frequently utilized AI algorithms.
• * The majority of AI models reviewed demonstrated acceptable performance in predicting nanoparticle toxicity.
• * AI tools show potential for robust, rapid, and economical nanomaterial safety evaluations.

Conclusions:

• * AI provides a viable alternative to traditional methods for nanoparticle toxicity assessment.
• * The reviewed AI models offer a promising pathway for enhancing nanomaterial safety studies.
• * Further development and application of AI can significantly benefit the field of nanotoxicology.