Integrating Artificial Intelligence with Plant Electrophysiology for Early Stress Diagnosis: A Scoping Review

This scoping review explores the ways in which plants respond on a constant basis to environmental and biological stresses like drought, nutrient deficiencies, heat, salinity, and pathogen attacks through their internal bioelectric processes. Electrophysiology is used to analyse plant internal bioelectrical processes, enabling direct measurement and delivering faster, more precise, and cost-effective diagnostic information compared to conventional methods. However, these processes have not been properly explored for the diagnosis of stress faced by plants until recent advances that used artificial intelligence. This book review compiles results of the 16 peer-reviewed studies carried out over the years, focusing on the integration of electrophysiological signal measurement and machine learning, deep learning, regression, and ensemble learning approaches like XGBoost, Random Forest, convolutional neural network(CNN), and long short-term memory(LSTM).The Studies involved several types of plants, including tomatoes, wheat, grapevine, soybeans, and peppers, grown both under controlled and field conditions. The results show that artificial intelligence (AI)-based approaches can accurately identify drought, nitrogen, heat, salinity, and pathogen stresses using the time-series electrophysiological signal of plants. Some studies have also shown the increasing use of wearable plant biosensors and real-time electrome monitoring systems. Despite the progress, several gaps remain. Most of the literature on the topic is based on small datasets, have experimental variability, and use non-standardized signal processing pipelines. This limits their use on a large scale. Noise sensitivity, the lack of open electrophysiology datasets, and model generalizability across species are among the challenges. Future studies should emphasize the following: the creation of universal bio signal databases, multi-species validation, the design of biosensor systems using the Internet of Things, and the development of Explainable AI models. Among the various applications of biosensors, plant electrophysiology using AI holds enormous potential for the efficient early diagnosis of stress patterns using biosensors.