Early-Stage Detection of Autism Spectrum Disorder Using Optimized Feature Selection and Deep Long Short-Term Memory Modeling on Behavioral Screening for Electrochemical Sensor Data
Abstract
Autism Spectrum Disorder (ASD) is a neurodevelopmental condition that significantly affects social interaction, communication, and behavior. Early-stage detection is essential for timely intervention and improved long-term outcomes. However, developing robust machine learning models for ASD screening is challenging due to limited dataset size and high-dimensional feature spaces. This study proposes an optimized feature selection framework based on the Flamingo–Pelican Optimization Algorithm (FPOA) to identify the most informative attributes from publicly available ASD screening datasets. The selected features are classified using a Deep Weighted Long Short-Term Memory (DWLSTM) model. Experimental evaluation is conducted on two questionnaire-based datasets (toddlers and adults), EEG recordings obtained from Kaggle and UCI repositories. Performance is assessed using accuracy, precision, recall, F-measure, and Matthews Correlation Coefficient (MCC). The results demonstrate that the proposed optimization-based feature selection combined with DWLSTM improves classification performance while reducing dimensionality. The framework provides an efficient computational approach for early ASD risk prediction using structured screening data.
Read the paper
Related studies
- Nano Biosensors: Properties, applications and electrochemical techniques
- Electrochemically Synthesized MIP Sensors: Applications in Healthcare Diagnostics
- Toward a Harmonization for Using in situ Nutrient Sensors in the Marine Environment
- Electrochemical determination of rutin by using NiFe2O4 nanoparticles-loaded reduced graphene oxide
- Application of Polyoxometalate-based composites for sensor systems: A review
- Sensitive Electrochemical and Thermal Detection of Human Noroviruses Using Molecularly Imprinted Polymer Nanoparticles Generated against a Viral Target
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.