ABSTRACT
The research is focused on an AI-driven medical prescription system which integrates Named Entity Recognition , Particle Swarm Optimization , and Text Generation . The system uses cloud computing for operational reasons. Currently, the system is in its experimental stages and adheres to the goal of improving and optimizing treatment plans while also increasing the accuracy of prescriptions. Precision of the NER module is within the range from 0.85 to 0.92. The Recall of those modules remains within the range from 0.85 to 0.92. Finally, F1 on trials is from 0.88 to 0.89 uniformity. These results indicate that the NER can successfully extract medical entities from text and provide a strong basis for the development of automated medical entity extraction. The PSO also finds solutions at a rapid convergence rate of 88% to 96% with medium to high solution quality. This is a key indicator that PSO is a suitable method for generating optimal treatment plans that comply with patient-specific attributes and clinical guidelines. Moreover, Text Generation component demonstrates a stable level of coherence, as the system produced text characterized with coherence ratings in the range from 4 to 4.6, widely across a 5-points scale. We come to the conclusion that these features of the produced text show its readability and understand ability that could promote the use of the prescription system. To sum up, it is possible to say that integrating NER, PSO, and Text Generation in a cloud-based prescription system is believed to be beneficial from the perspective of healthcare delivery.
