Population-Based Chemotherapy Optimization Using Genetic Algorithm.

We can enhance the current medical technology by introducing engineering methods. The current chemotherapy treatments are designed for the average and are only taking into account a few of the patient parameters (e.g., weight, history of diagnoses). By creating personalized therapy, we can decrease the injected doses, thus reducing the harmful side effects and the risk of developing drug resistance. To personalize therapy, we used a mathematical model that contains the unique patient parameters to describe the tumor dynamics with respect to the injected doses. In this study, a previously developed genetic algorithm was improved by examining four different fitness function variations and three different mutation functions. The main goal was to generate one single therapy for a set of patients. The parameters of the patients were examined, in order to create a set with similar patients. One of the results shows that the generated therapy can produce a 100% survival rate in a simulated environment with slightly higher doses compared to the standard clinically used treatment. Also by administering slightly lower doses, we can still achieve better survival rates compared to the clinically used treatment.