Machine Learning Based Early Detection of Autism Spectrum Disorder

Abstract

The Autism Screening System is a web-based application designed to support early detection of Autism Spectrum Disorder (ASD) using machine learning techniques. Early identification is important for timely intervention and improving individuals\' quality of life. The system is developed using the Flask framework and provides an interactive interface for healthcare professionals. It analyzes various clinical parameters such as age, gender, family history, communication skills, social interaction, and behavioral patterns. Multiple machine learning algorithms, including Random Forest, Decision Tree, K-Nearest Neighbors, Naive Bayes, and Logistic Regression, are implemented and compared. The model with the highest accuracy is automatically selected for prediction. The system also includes graphical visualization to evaluate model performance. It generates predictions along with confidence levels to assist doctors in decision-making. This tool is intended to support, not replace, clinical diagnosis. Overall, it provides a fast and data-driven approach for preliminary ASD screening, with potential for future improvements using real-world data and healthcare deployment.

Introduction

The Autism Screening System is a Flask-based web application designed to support the early detection of Autism Spectrum Disorder (ASD) using machine learning. Autism affects communication, behavior, and social interaction, making early diagnosis essential for timely intervention and improved quality of life. Traditional diagnostic methods rely on clinical observation and expert evaluation, which can be time-consuming and may not always be readily available. The proposed system addresses these limitations by automating the screening process through data-driven machine learning techniques.

The application provides a simple and user-friendly interface where users enter patient information, including age, gender, family history, and behavioral responses. These inputs are processed by a trained machine learning model that predicts the likelihood of autism and provides a confidence score for the prediction. The system also stores patient history, supports data visualization through charts, and enables doctors and users to monitor screening trends and make informed decisions. By reducing manual effort and improving screening efficiency, the system can be effectively used in hospitals, clinics, and educational institutions.

The literature survey highlights significant advancements in applying machine learning to autism detection. Previous studies have employed algorithms such as Decision Trees, Support Vector Machines (SVM), Naïve Bayes, Artificial Neural Networks (ANNs), and ensemble learning methods to analyze behavioral and demographic data for ASD prediction. Researchers have also developed web-based and mobile screening applications to improve accessibility. However, many existing systems lack intuitive user interfaces, graphical visualization of results, and comprehensive patient history management, creating opportunities for further improvement.

Several challenges were encountered during system development. A major challenge is obtaining high-quality and balanced datasets, as incomplete or biased data can reduce prediction accuracy. Selecting an appropriate machine learning algorithm that achieves high accuracy while maintaining computational efficiency is another critical issue. Additional challenges include handling inconsistent user inputs, performing effective data preprocessing and feature selection, integrating machine learning models with the web application, securely managing patient records, and ensuring the privacy and confidentiality of sensitive medical information.

The proposed methodology follows a structured machine learning workflow. Initially, the dataset is preprocessed by handling missing values, encoding categorical variables, and selecting relevant features such as age, gender, and behavioral responses. Multiple classification algorithms—including Random Forest, Decision Tree, K-Nearest Neighbors (KNN), Naïve Bayes, and Logistic Regression—are trained and evaluated. The model with the highest prediction accuracy is selected and integrated into a Flask-based web application, where users submit patient information and receive ASD predictions along with confidence levels.

The system incorporates several key technologies and techniques:

• Machine Learning algorithms classify patient data and generate autism predictions with confidence scores.
• The Flask web framework manages routing, user sessions, request handling, and communication between the interface and the prediction model.
• Data preprocessing includes handling missing values and converting categorical variables into numerical values suitable for machine learning.
• SQLite serves as the backend database for secure user authentication, patient record storage, and history management.

The system follows a three-tier architecture consisting of:

1. Presentation Layer: Developed using HTML5, CSS3/Bootstrap, and JavaScript to provide an interactive and responsive user interface.
2. Logic Layer: Implemented using Flask, which processes user requests, executes the trained machine learning model, and generates prediction results.
3. Data Layer: Uses an SQLite database to store user accounts, screening records, and prediction history securely.

The application includes several functional modules:

• Home Page: Provides login and registration options through a simple and healthcare-oriented interface.
• Registration Page: Allows new users to create secure accounts with authentication features.
• Dashboard: Displays summary statistics such as total screenings conducted and detected autism cases, supported by graphical visualizations using Chart.js.
• Prediction Page: Collects patient demographics and responses to 10 behavioral screening questions (A1–A10) before submitting them to the machine learning model.
• Patient History: Maintains previous screening results, including patient names, predictions, confidence scores, and options for record management.
• Prediction Output: Displays the final autism prediction, confidence percentage, and a disclaimer stating that the screening result is not a substitute for professional medical diagnosis.

Conclusion

The Autism Screening System is an effective web-based application that uses machine learning to assist in the early detection of Autism Spectrum Disorder. It provides a simple interface for entering patient details and behavioral responses. The system processes the input data and generates predictions along with confidence levels. Features like dashboard, charts, and patient history improve usability and data management. It helps healthcare professionals make quicker preliminary assessments. However, it is not a substitute for expert medical diagnosis. The project demonstrates the power of integrating AI with healthcare systems. Overall, it contributes to faster screening and better awareness of autism.

References

[1] Akter, T., Satu, M. S., Khan, M. I., Ali, M. H., Uddin, S., Lio, P., Quinn, J. M. W., & Moni, M. A. (2019). Machine learning-based models for early stage detection of autism spectrum disorders. IEEE Access, 7 166509–166527. https://doi.org/10.1109/ACCESS.2019.2953968 [2] Abbas, H., Garberson, F., Glover, E., & Wall, D. P. (2018). Machine learning approach for early detection of autism by combining questionnaire and home video screening. Journal of the American Medical Informatics Association, 25(8) 1000–1007. https://doi.org/10.1093/jamia/ocy039 [3] Thabtah, F. (2019). Machine learning in autistic spectrum disorder behavioral research: A review and ways forward. Informatics for Health and Social Care, 44(3) 278–297. https://doi.org/10.1080/17538157.2017.1399132 [4] Simeoli, R., Rega, A., & Cerasuolo, M. (2024). Using machine learning for motion analysis to early detect autism spectrum disorder: A systematic review. Review Journal of Autism and Developmental Disorders https://doi.org/10.1007/s40489-024-00435-4 [5] Rajagopalan, S. S., Zhang, Y., Yahia, A., & Tammimies, K. (2024). Machine learning prediction of autism spectrum disorder from a minimal set of medical and background information. JAMA Network Open, 7(8) e2429229. https://doi.org/10.1001/jamanetworkopen.2024.29229 [6] Farooq, M. S., & Abid, A. (2023). Detection of autism spectrum disorder (ASD) in children and adults using machine learning. Scientific Reports, 13 9483. https://doi.org/10.1038/s41598-023-35910-1 [7] Abbas, R. T., Sultan, K., Sheraz, M., & Chuah, T. C. (2024). A comparative analysis of automated machine learning tools: A use case for autism spectrum disorder detection. Information, 15(1) 1–20. https://doi.org/10.3390/info15010020 [8] Dcouto, J., & Pradeepkandhasamy, R. (2023). Multimodal approaches for early autism spectrum disorder detection using machine learning. Sensors, 23(5) 1228. https://doi.org/10.3390/s23051228 [9] Crippa, A., Salvatore, C., Perego, P., Forti, S., Nobile, M., Molteni, M., & Castiglioni, I. (2015). Use of machine learning to identify children with autism and their motor abnormalities. Journal of Autism and Developmental Disorders, 45(7) 2146–2156. https://doi.org/10.1007/s10803-015-2379-8 [10] Baldi, P., Bruschi, M., Dellantonio, S., & Pastore, L. (2021). Use of machine learning to investigate the quantitative checklist for autism in toddlers (Q-CHAT) towards early autism screening. Diagnostics, 11(3) 574. https://doi.org/10.3390/diagnostics11030574 [11] Baron-Cohen, S., Wheelwright, S., Skinner, R., Martin, J., & Clubley, E. (2001). The autism-spectrum quotient (AQ): Evidence from Asperger syndrome/high-functioning autism, males and females, scientists and mathematicians. Journal of Autism and Developmental Disorders, 31(1) 5–17. https://doi.org/10.1023/A:1005653411471 [12] Thabtah, F. (2017). Autistic Spectrum Disorder Screening Data for Children [Dataset]. UCI Machine Learning Repository. https://doi.org/10.24432/C5659W [13] Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., ... & Duchesnay, E. (2011). Scikit-learn: Machine learning in Python. Journal of Machine Learning Research, 12 2825–2830. https://jmlr.org/papers/v12/pedregosa11a.html [14] Breiman, L. (2001). Random forests. Machine Learning, 45(1) 5–32. https://doi.org/10.1023/A:1010933404324 [15] Cortes, C., & Vapnik, V. (1995). Support-vector networks. Machine Learning, 20(3) 273–297. https://doi.org/10.1007/BF00994018 [16] Chen, T., & Guestrin, C. (2016). XGBoost: A scalable tree boosting system. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining 785–794. https://doi.org/10.1145/2939672.2939785 [17] Grinberg, M. (2018). Flask Web Development: Developing Web Applications with Python, (2nd ed.) O\'Reilly Media. [18] SQLite Consortium. (2024). SQLite: A self-contained, serverless, zero-configuration, transactional SQL database engine. https://www.sqlite.org [19] World Health Organization. (2023). Autism spectrum disorders. WHO Fact Sheet. https://www.who.int/news-room/fact-sheets/detail/autism-spectrum-disorders [20] American Psychiatric Association. (2013). Diagnostic and Statistical Manual of Mental Disorders, (5th ed.). APA Publishing. https://doi.org/10.1176/appi.books.9780890425596

Copyright

Copyright © 2026 P. Bindhu Priya , P. Manikantha. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.