Smart Drug Delivery Systems Using Large Language Models for Real-Time Treatment Personalization
Chinnala Balakrishna, Ankit Kumar Yadav, Jagendra Singh, Masarath Saba, Shashikant, Vineet Shrivastava · 2024
This research explores the use of large language models, such as BERT and GPT, in developing a smart drug delivery system utilizing real-time personalized treatments. The research aims to utilize large datasets with advanced natural language processing to recommend the appropriate drug for a patient based on their health record with enhanced accuracy and efficiency. The research, which evaluates and compares BERT and GPT, achieves the goal of predicting a drug with high accuracy, and GPT delivers the best results compared to BERT. Specifically, GPT achieved an accuracy of 97.95%, while BERT's accuracy was 95.50%. Additionally, the research emphasizes the essential aspect of a model's time response since these are real-time clinical decision systems. GPT took 110 milliseconds to predict the drug while the BERT took 120 milliseconds. It is clear from the results of this work that LLM has the potential of changing personalized medicine's approach by recommending drugs in real-time and according to the patient's health record within no time. The proposed system for smart drug delivery is promising to improve healthcare services, patient outcomes, and reduce drug administration errors. Apart from predicting the drug, these research findings can be simulated to the health sectors and integrated with AI technologies to improve decision support systems.