The disease schizophrenia is a highly complex mental disorder, making it difficult to diagnose reliably due to heterogeneity and subjective features of its symptoms, many overlapping with other disorders, which makes misdiagnosis possible. Traditional diagnosis includes methods such as the use of EEG signal, where regular methods often do not succeed in capturing subtle abnormalities and to handle data variability in the process. This paper hence endeavors to overcome the mentioned challenges by introducing a Distributed Activation Function-Based Statistical Attention Bidirectional Long Short-Term Memory model for schizophrenia detection using an analysis of EEG signals. The techniques used in detecting the incidence of schizophrenia, level of complexity, and comparison of the techniques are identified as well as research gaps present in relation to the study have also been discussed. It sheds light on the DA-SA-BiLSTM model. It also talks about the strengths of the model and provides a future perspective in the area of schizophrenia diagnosis.