The lack of specificity is a frequent drawback of existing cancer treatments, impacting both healthy and malignant cells, resulting in unintended harm and side effects on normal tissues. Drug resistance and tolerance present major obstacles in cancer treatment, diminishing the effectiveness of therapies and restricting their long-term success. Identifying therapeutic standards to address these issues is essential for enhancing the effectiveness of cancer treatments.Nanoparticles have revolutionised targeted cancer therapy by enabling precise drug delivery, enhanced permeability and retention, stimuli-responsive release, and active tumor homing. The extreme complexity and variability of nanoparticle physicochemical properties, tumor biology, and patient-specific factors have made traditional empirical optimization inefficient. Artificial intelligence (AI) and machine learning (ML) are increasingly used to speed up nanoparticle design, predict therapeutic success, reduce off-target toxicity, and personalize nanomedicine treatments. A systematic review was conducted by searching reputable international databases, including Web of Science, PubMed, Scopus, Embase, and IEEE, from 2015 to January 2025. Studies were eligible if they used AI (ML and DL) techniques on nanoparticle-related datasets for any solid or hematological malignancy. Risk of bias and reporting quality were independently evaluated using an adapted PROBAST-AI checklist. SVM and CNN were the most common algorithms; these models achieved 100% accuracy when analyzing confocal images of TNBC cells treated with nanoparticles. The models also showed high accuracy in simulating nanoparticle behavior and predicting their therapeutic effects. Other studies reported AI techniques in nanomedicine with accuracies ranging from 90% to 99%. Improper handling of missing and outlier data and the use of data sets that are too small relative to the number of predictive features included create risk of bias. It was found that nanoparticle-based therapies, especially those using cell coatings, could help reduce cancer complications. Nanoparticles with features like targeted drug delivery and biological coatings have shown promising results, particularly in nine TNBC studies and ovarian cancer. AI algorithms' ability to analyze large data sets and detect hidden, complex patterns can improve the design of nanotechnologies for diagnosis and treatment. We believe AI-driven analysis reveals complex molecular interactions and pathways that are dysregulated in cancer, aiding in the discovery of therapeutic targets.
Rahmani et al. (Sun,) studied this question.