Autonomous Exploration System of Electrical Stimulus Condition for Maturating in vitro Skeletal Muscle Tissue with Bayesian Optimization

Abstract

This paper describes an autonomous exploration system to optimize electrical stimulation parameters for the maturation of in vitro skeletal muscle tissues with Bayesian optimization. We established a cyclic experimental procedures, where (i) C2C12 myoblasts were cultured under multiple electrical stimulus conditions in parallel to promote the maturation, (ii) the cell maturation levels were scored by microscopic image analysis with immunofluorescence staining, and (iii) each maturation score with electrical stimulus condition were then applied to Bayesian optimization to suggest new experimental stimulus conditions for the next cycle of the C2C12 culture to expect higher maturation scores. We performed this experimental cycle repeatedly and derived the trend of electrical stimulus conditions to obtain high maturation scores. The C2C12 myoblasts cultured under the derived condition were also verified by the increase of myogenic gene expression. We believe that our proposed method could reveal optimal application method for tissue maturation, and would contribute to regenerative medicine or other medical technologies using tissue culture.