Handwritten Math Expression recognition is an active area of research that aims to enable machines to automatically interpret handwritten mathematical notation. This technology has many potential applications in education, office automation, and assisting people with visual impairments. In this review, we provide a comprehensive overview of the current state-of-the-art in HME recognition. Specifically, we summarize key preprocessing techniques, feature extraction methods, classification models, benchmark datasets, and performance metrics that have been explored over the past decade. Preprocessing techniques covered include binarization, normalization, smoothing, and outlier removal. Both handcrafted features like gradients and angles as well as automatically learned features from neural networks are discussed. We review the use of Hidden Markov Models, Support Vector Machines, Convolutional Neural Networks, Recurrent Neural Networks like LSTM, and more recently Transformer models for this task. Key benchmark datasets like CROHME 2014 and 2016 are analyzed. Performance metrics like symbol recognition rates, expression recognition rates, and expression structure prediction are compared across different techniques. We highlight strengths and limitations of the reviewed approaches. Finally, we outline important open challenges and promising directions for advancing HME recognition, such as larger datasets, multi-modal learning, and incorporation of syntactic and semantic context. This comprehensive review will serve as a valuable reference for researchers and developers working on interpreting handwritten mathematical expressions.