Integrating Artificial Intelligence and Big Data Analytics in Personalized Autism Treatment through Stem Cell Therapy

Md Kamruzzaman; Anseena Anees Sabeena; Arifa Ahmed; Muslima Begom Riipa; Adib Hossain; Rayhan Khan; Arafat Islam; Fahad Ahmed

doi:10.63332/joph.v5i6.2077

Authors

Md Kamruzzaman Department of Business Administration, Westcliff University, Irvine, CA 92614, USA
Anseena Anees Sabeena Department of Business Administration, Westcliff University, Irvine, CA 92614, USA
Arifa Ahmed Department of Business Administration, International American University, Los Angeles, CA 90010, USA
Muslima Begom Riipa Department of Business Administration, International American University, Los Angeles, CA 90010, USA
Adib Hossain Department of Business Analytics, Trine University, Angola, IN 46703, USA
Rayhan Khan Department of Special Education and Counseling, California State University, Los Angeles, CA 90032, USA
Arafat Islam Department of Business Analytics, Trine University, Angola, IN 46703, USA
Fahad Ahmed Department of Science, Trine University, Angola, IN 46703, USA

DOI:

https://doi.org/10.63332/joph.v5i6.2077

Keywords:

Artificial Intelligence, Data Analytics, Cell Therapy

Abstract

Background: Autism Spectrum Disorder (ASD) presents challenges in early diagnosis and personalized treatment due to the subjectivity of traditional screening methods and the lack of tailored therapeutic approaches. Artificial Intelligence (AI) and Big Data Analytics offer innovative solutions for enhancing diagnostic accuracy and optimizing treatment pathways. Objective: This study explores the application of AI in ASD diagnosis and predictive modeling for personalized stem cell therapy, aiming to improve early detection and treatment effectiveness. Methods: A dataset of 704 individuals was analyzed using machine learning models, including LightGBM, Random Forest, Neural Networks, XGBoost, and Stacking Ensemble. Performance was assessed using ROC-AUC, Accuracy, Precision, Recall, and F1-score, while AI models were further applied to predict therapy response patterns for stem cell treatment optimization.Results: The Stacking Ensemble model (ROC-AUC = 0.9989, F1 = 0.9125) demonstrated superior performance in ASD classification. Neural Networks exhibited the highest recall (95.76%), making them ideal for early screening. AI-driven insights facilitated the identification of key ASD biomarkers, enabling personalized treatment strategies.Conclusion: AI significantly enhances ASD detection and treatment planning, providing a data-driven, personalized approach. Future research should focus on real-world validation, integrating genetic biomarkers, and ensuring ethical AI deployment in clinical settings. These advancements pave the way for precision medicine in neurodevelopmental disorders.