Abstract
In response to the global need for efficient early diagnosis of AutismSpectrum Disorder (ASD), this paper bridges the gap between traditional,time-consuming diagnostic methods and potential automated solutions. We proposea multi-atlas deep ensemble network, MADE-for-ASD, that integrates multipleatlases of the brain's functional magnetic resonance imaging (fMRI) datathrough a weighted deep ensemble network. Our approach integrates demographicinformation into the prediction workflow, which enhances ASD diagnosisperformance and offers a more holistic perspective on patient profiling. Weexperiment with the well-known publicly available ABIDE (Autism Brain ImagingData Exchange) I dataset, consisting of resting state fMRI data from 17different laboratories around the globe. Our proposed system achieves 75.20%accuracy on the entire dataset and 96.40% on a specific subset $-$ bothsurpassing reported ASD diagnosis accuracy in ABIDE I fMRI studies.Specifically, our model improves by 4.4 percentage points over prior works onthe same amount of data. The model exhibits a sensitivity of 82.90% and aspecificity of 69.70% on the entire dataset, and 91.00% and 99.50%,respectively, on the specific subset. We leverage the F-score to pinpoint thetop 10 ROI in ASD diagnosis, such as \emph{precuneus} and anterior\emph{cingulate/ventromedial}. The proposed system can potentially pave the wayfor more cost-effective, efficient and scalable strategies in ASD diagnosis.Codes and evaluations are publicly available at TBA.