Deep Learning and Neural Network Modeling to Optimize Tissue Culture of Medicinal Plants

Medicinal plants face significant threats of extinction due to habitat destruction, overexploitation, and environmental degradation. Effective conservation and sustainable utilization of these species require efficient propagation techniques. Plant tissue culture offers a promising method by allowing the regulated and swift increase of medicinal plants. As a key biotechnological tool, it possesses significant potential for progressing herbal medicine and therapeutic medical practices. This chapter aims to explore the application of deep learning (DL) methods in the in vitro development of medicinal plants. DL methods, especially convolutional neural networks (CNNs), have transformed image-based analysis in the processes of plant tissue culture. This chapter evaluates the processes involved in plant tissue induction and emphasizes the incorporation of DL in tissue culture techniques. Specifically, we discuss the development of predictive models that relate factors such as callus type, plant growth regulator (PGR) type, and PGR concentration to the success of indirect shoot regeneration and secondary metabolite production. The goal of these insights is to improve tissue culture techniques, which will, in turn, increase both the availability and yield of natural metabolites derived from medicinal plants. Despite extensive research on DL applications in plant biotechnology, a lack of empirical studies offering meaningful comparisons persists, primarily due to the use of diverse datasets in evaluations. This chapter addresses these challenges and proposes solutions to advance the automation of plant tissue culture systems. It also provides a comprehensive resource for plant biotechnologists and tissue culturists, fostering innovation in the conservation and sustainable production of endangered medicinal plant species.