Comparative study of machine learning and deep learning methods on ASD classification
This work addresses autism spectrum disorder diagnosis, but it is incremental as it compares existing machine learning and deep learning methods without introducing new techniques.
The study tackled the problem of classifying autistic versus healthy individuals using resting-state fMRI connectivity patterns, achieving a classification accuracy of 71% on the ABIDE I dataset.
The autism dataset is studied to identify the differences between autistic and healthy groups. For this, the resting-state Functional Magnetic Resonance Imaging (rs-fMRI) data of the two groups are analyzed, and networks of connections between brain regions were created. Several classification frameworks are developed to distinguish the connectivity patterns between the groups. The best models for statistical inference and precision were compared, and the tradeoff between precision and model interpretability was analyzed. Finally, the classification accuracy measures were reported to justify the performance of our framework. Our best model can classify autistic and healthy patients on the multisite ABIDE I data with 71% accuracy.