yunidar, Arya Suyanda, Melinda Melinda, Lailatul Qadri Zakaria, Siti Rusdiana

Classification of Autism Spectrum Disorder (ASD) in Children Using the VGG19 CNN Model Based on Facial Landmarks of the Eye and Forehead Areas

Introduction

Classification of autism spectrum disorder (asd) in children using the vgg19 cnn model based on facial landmarks of the eye and forehead areas. Detect Autism Spectrum Disorder (ASD) early in children using a VGG19 CNN model analyzing facial landmarks from eye and forehead areas. Achieved 94.35% accuracy for objective, practical screening.

Abstract

Early detection of Autism Spectrum Disorder (ASD) is a crucial challenge in child development interventions because conventional screening methods are often subjective and prone to assessor bias. This study proposes an objective solution in the form of a deep learning approach for automatic ASD classification using facial landmark representations that focus exclusively on the eye and forehead areas. The selection of these areas is based on the eye avoidance hypothesis, which states that these regions contain very rich diagnostic information and behavioral biomarkers related to the ASD phenotype. The pre-processing stage involves isolating the eye and forehead areas using Dlib 68-landmark detection to eliminate background visual noise, followed by detailed topological visualization using MediaPipe Face Mesh with 478 landmark points as the model input. The Convolutional Neural Network (CNN) architecture used is the VGG19 model modified with transfer learning techniques and the addition of Dropout layers to improve efficiency and prevent overfitting. The model was trained on a primary dataset of 1,238 images collected under controlled conditions from children in Banda Aceh. The test results showed very promising performance with an overall accuracy of 94.35%. Specifically, the model achieved a recall (sensitivity) of 95.24%, a precision of 93.75%, and an AUC score of 0.9831. This high sensitivity is crucial in a medical context to minimize the risk of misdetection of positive cases. These results demonstrate that landmark visualization in the eye and forehead areas with the VGG19 model is a highly effective, accurate, and practical method for serving as an economical early screening tool for ASD.

Review

This study presents a compelling deep learning approach for the objective classification of Autism Spectrum Disorder (ASD) in children, addressing a critical need to move beyond subjective conventional screening methods. Utilizing a VGG19 Convolutional Neural Network (CNN) model, the research innovatively focuses on facial landmarks specifically within the eye and forehead regions. This targeted approach is theoretically grounded in the "eye avoidance hypothesis," positing these areas as rich sources of diagnostic behavioral biomarkers. The methodology employs advanced pre-processing with Dlib and MediaPipe to capture detailed topological features, culminating in a robust system with significant potential for early ASD detection. Methodologically, the paper demonstrates strong technical execution, integrating transfer learning and Dropout layers into the VGG19 architecture to enhance efficiency and prevent overfitting. The reported performance metrics are notably high, with an impressive overall accuracy of 94.35%, a recall (sensitivity) of 95.24%, and an AUC score of 0.9831. The emphasis on high sensitivity is particularly commendable, acknowledging its paramount importance in minimizing false negatives within a medical screening context. While the use of a primary dataset of 1,238 images collected under controlled conditions is a strength, a full appreciation of the model's generalizability would benefit from further detail on the dataset's demographic diversity, the rigorousness of "controlled conditions," and the clinical validation process for the ground truth labels. Such details would reinforce confidence in its broad applicability beyond the specific population of Banda Aceh. Overall, this research offers a highly promising and practical contribution towards developing an economical and objective early screening tool for ASD. The combination of targeted facial landmark analysis and a well-optimized deep learning model presents a significant step forward in leveraging computational vision for child development interventions. The demonstrated high accuracy and sensitivity strongly advocate for the method's potential to improve the timeliness and reliability of ASD diagnosis. Further validation across diverse populations and real-world clinical settings will be crucial, but the foundational work presented here establishes an excellent precedent for future research in this vital area, warranting strong consideration for publication given its innovation and potential impact.

Full Text

Comments

You need to be logged in to post a comment.