Neurodevelopmental trajectories of cerebellar grey matter associated with verbal abilities in males with autism spectrum disorder.

Autism spectrum disorder (ASD) is a neurodevelopmental condition frequently associated with structural cerebellar abnormalities. Whether cerebellar grey matter volumes (GMV) are linked to verbal impairments remains controversial. Here, the association between cerebellar GMV and verbal abilities in ASD was examined across the lifespan. Lobular segmentation of the cerebellum was performed on structural MRI scans from the ABIDE I dataset in male individuals with ASD (N=144, age: 8.5-64.0 years) and neurotypical controls (N=188; age: 8.0-56.2 years). Stepwise linear mixed effects modeling including group (ASD vs. neurotypical controls), lobule-wise GMV, and age was performed to identify cerebellar lobules which best predicted verbal abilities as measured by verbal IQ (VIQ). An age-specific association between VIQ and GMV of bilateral Crus II was found in ASD relative to neurotypical controls. In children with ASD, higher VIQ was associated with larger GMV of left Crus II but smaller GMV of right Crus II. By contrast, in adults with ASD, higher VIQ was associated with smaller GMV of left Crus II and larger GMV of right Crus II. These findings indicate that relative to the contralateral hemisphere, an initial reliance on the language-nonspecific left cerebellar hemisphere is offset by more typical right-lateralization in adulthood.

Residual graph transformer for autism spectrum disorder prediction.

Brain functional connectivity (FC) based on resting-state functional magnetic resonance imaging (rs-fMRI) has been in vogue to predict Autism Spectrum Disorder (ASD), which is a neuropsychiatric disease up the plight of locating latent biomarkers for clinical diagnosis. Albeit massive endeavors have been made, most studies are fed up with several chronic issues, such as the intractability of harnessing the interaction flourishing within brain regions, the astriction of representation due to vanishing gradient within deeper network architecture, and the poor interpretability leading to unpersuasive diagnosis. To ameliorate these issues, a FC-learned Residual Graph Transformer Network, namely RGTNet, is proposed. Specifically, we design a Graph Encoder to extract temporal-related features with long-range dependencies, from which interpretable FC matrices would be modeled. Besides, the residual trick is introduced to deepen the GCN architecture, thereby learning the higher-level information. Moreover, a novel Graph Sparse Fitting followed by weighted aggregation is proposed to ease dimensionality explosion. Empirically, the results on two types of ABIDE data sets demonstrate the meliority of RGTNet. Notably, the achieved ACC metric reaches 73.4%, overwhelming most competitors with merely 70.9% on the AAL atlas using a five-fold cross-validation policy. Moreover, the investigated biomarkers concord closely with the authoritative medical knowledge, paving a viable way for ASD-clinical diagnosis. Our code is available at https://github.com/CodeGoat24/RGTNet.

Identification of Autism Spectrum Disorder Using Topological Data Analysis.

Autism spectrum disorder (ASD) is a pervasive brain development disease. Recently, the incidence rate of ASD has increased year by year and posed a great threat to the lives and families of individuals with ASD. Therefore, the study of ASD has become very important. A suitable feature representation that preserves the data intrinsic information and also reduces data complexity is very vital to the performance of established models. Topological data analysis (TDA) is an emerging and powerful mathematical tool for characterizing shapes and describing intrinsic information in complex data. In TDA, persistence barcodes or diagrams are usually regarded as visual representations of topological features of data. In this paper, the Regional Homogeneity (ReHo) data of subjects obtained from Autism Brain Imaging Data Exchange (ABIDE) database were used to extract features by using TDA. The average accuracy of cross validation on ABIDE I database was 95.6% that was higher than any other existing methods (the highest accuracy among existing methods was 93.59%). The average accuracy for sampling with the same resolutions with the ABIDE I on the ABIDE II database was 96.5% that was also higher than any other existing methods (the highest accuracy among existing methods was 75.17%).

Deep learning based joint fusion approach to exploit anatomical and functional brain information in autism spectrum disorders.

BACKGROUND: The integration of the information encoded in multiparametric MRI images can enhance the performance of machine-learning classifiers. In this study, we investigate whether the combination of structural and functional MRI might improve the performances of a deep learning (DL) model trained to discriminate subjects with Autism Spectrum Disorders (ASD) with respect to typically developing controls (TD). MATERIAL AND METHODS: We analyzed both structural and functional MRI brain scans publicly available within the ABIDE I and II data collections. We considered 1383 male subjects with age between 5 and 40 years, including 680 subjects with ASD and 703 TD from 35 different acquisition sites. We extracted morphometric and functional brain features from MRI scans with the Freesurfer and the CPAC analysis packages, respectively. Then, due to the multisite nature of the dataset, we implemented a data harmonization protocol. The ASD vs. TD classification was carried out with a multiple-input DL model, consisting in a neural network which generates a fixed-length feature representation of the data of each modality (FR-NN), and a Dense Neural Network for classification (C-NN). Specifically, we implemented a joint fusion approach to multiple source data integration. The main advantage of the latter is that the loss is propagated back to the FR-NN during the training, thus creating informative feature representations for each data modality. Then, a C-NN, with a number of layers and neurons per layer to be optimized during the model training, performs the ASD-TD discrimination. The performance was evaluated by computing the Area under the Receiver Operating Characteristic curve within a nested 10-fold cross-validation. The brain features that drive the DL classification were identified by the SHAP explainability framework. RESULTS: The AUC values of 0.66±0.05 and of 0.76±0.04 were obtained in the ASD vs. TD discrimination when only structural or functional features are considered, respectively. The joint fusion approach led to an AUC of 0.78±0.04. The set of structural and functional connectivity features identified as the most important for the two-class discrimination supports the idea that brain changes tend to occur in individuals with ASD in regions belonging to the Default Mode Network and to the Social Brain. CONCLUSIONS: Our results demonstrate that the multimodal joint fusion approach outperforms the classification results obtained with data acquired by a single MRI modality as it efficiently exploits the complementarity of structural and functional brain information.

Diagnosis of Autism Spectrum Disorder (ASD) Using Recursive Feature Elimination-Graph Neural Network (RFE-GNN) and Phenotypic Feature Extractor (PFE).

Autism spectrum disorder (ASD) poses as a multifaceted neurodevelopmental condition, significantly impacting children's social, behavioral, and communicative capacities. Despite extensive research, the precise etiological origins of ASD remain elusive, with observable connections to brain activity. In this study, we propose a novel framework for ASD detection, extracting the characteristics of functional magnetic resonance imaging (fMRI) data and phenotypic data, respectively. Specifically, we employ recursive feature elimination (RFE) for feature selection of fMRI data and subsequently apply graph neural networks (GNN) to extract informative features from the chosen data. Moreover, we devise a phenotypic feature extractor (PFE) to extract phenotypic features effectively. We then, synergistically fuse the features and validate them on the ABIDE dataset, achieving 78.7% and 80.6% accuracy, respectively, thereby showcasing competitive performance compared to state-of-the-art methods. The proposed framework provides a promising direction for the development of effective diagnostic tools for ASD.

Effect of data harmonization of multicentric dataset in ASD/TD classification.

Machine Learning (ML) is nowadays an essential tool in the analysis of Magnetic Resonance Imaging (MRI) data, in particular in the identification of brain correlates in neurological and neurodevelopmental disorders. ML requires datasets of appropriate size for training, which in neuroimaging are typically obtained collecting data from multiple acquisition centers. However, analyzing large multicentric datasets can introduce bias due to differences between acquisition centers. ComBat harmonization is commonly used to address batch effects, but it can lead to data leakage when the entire dataset is used to estimate model parameters. In this study, structural and functional MRI data from the Autism Brain Imaging Data Exchange (ABIDE) collection were used to classify subjects with Autism Spectrum Disorders (ASD) compared to Typical Developing controls (TD). We compared the classical approach (external harmonization) in which harmonization is performed before train/test split, with an harmonization calculated only on the train set (internal harmonization), and with the dataset with no harmonization. The results showed that harmonization using the whole dataset achieved higher discrimination performance, while non-harmonized data and harmonization using only the train set showed similar results, for both structural and connectivity features. We also showed that the higher performances of the external harmonization are not due to larger size of the sample for the estimation of the model and hence these improved performance with the entire dataset may be ascribed to data leakage. In order to prevent this leakage, it is recommended to define the harmonization model solely using the train set.

Detection of sparse differential dependent functional brain connectivity.

Functional brain connectivity analysis is an increasingly important technique in neuroscience, psychiatry, and autism research. Functional connectivity can be measured by considering co-activation of brain regions in resting-state functional magnetic resonance imaging (rs-fMRI). We propose a novel Bayesian model to detect differential connections in cross-correlated functional connectivity between region of interest (ROI) pairs. The proposed sparse clustered neighborhood model induces a lower-dimensional sparsity and clustering based on a nonparametric Bayesian approach to model sparse differentially connected ROI pairs. Second, it induces a structured dependence model for modeling potential dependence among ROI pairs. We demonstrate Bayesian inference and performance of the proposed model in simulation studies and compare with a standard model. We utilize the proposed model to contrast functional connectivities between participants with autism spectrum disorder and neurotypical participants using cross-correlated rs-fMRI data from four sites of the Autism Brain Image Data Exchange.

Confounding Effects on the Performance of Machine Learning Analysis of Static Functional Connectivity Computed from rs-fMRI Multi-site Data.

Resting-state functional magnetic resonance imaging (rs-fMRI) is a non-invasive imaging technique widely used in neuroscience to understand the functional connectivity of the human brain. While rs-fMRI multi-site data can help to understand the inner working of the brain, the data acquisition and processing of this data has many challenges. One of the challenges is the variability of the data associated with different acquisitions sites, and different MRI machines vendors. Other factors such as population heterogeneity among different sites, with variables such as age and gender of the subjects, must also be considered. Given that most of the machine-learning models are developed using these rs-fMRI multi-site data sets, the intrinsic confounding effects can adversely affect the generalizability and reliability of these computational methods, as well as the imposition of upper limits on the classification scores. This work aims to identify the phenotypic and imaging variables producing the confounding effects, as well as to control these effects. Our goal is to maximize the classification scores obtained from the machine learning analysis of the Autism Brain Imaging Data Exchange (ABIDE) rs-fMRI multi-site data. To achieve this goal, we propose novel methods of stratification to produce homogeneous sub-samples of the 17 ABIDE sites, as well as the generation of new features from the static functional connectivity values, using multiple linear regression models, ComBat harmonization models, and normalization methods. The main results obtained with our statistical models and methods are an accuracy of 76.4%, sensitivity of 82.9%, and specificity of 77.0%, which are 8.8%, 20.5%, and 7.5% above the baseline classification scores obtained from the machine learning analysis of the static functional connectivity computed from the ABIDE rs-fMRI multi-site data.

Bloody noise: The impact of blood-flow artifacts on registration.

Blood-flow artifacts present a serious challenge for most, if not all, volumetric analytical approaches. We utilize T1-weighted data with prominent blood-flow artifacts from the Autism Brain Imaging Data Exchange (ABIDE) multisite agglomerative dataset to assess the impact that such blood-flow artifacts have on registration of T1-weighted data to a template. We use a heuristic approach to identify the blood-flow artifacts in these data; we use the resulting blood masks to turn the underlying voxels to the intensity of the cerebro-spinal fluid, thus mimicking the effect of blood suppression. We then register both the original data and the deblooded data to a common T1-weighted template, and compare the quality of those registrations to the template in terms of similarity to the template. The registrations to the template based on the deblooded data yield significantly higher similarity values compared with those based on the original data. Additionally, we measure the nonlinear deformations needed to transform the data from the position achieved by registering the original data to the template to the position achieved by registering the deblooded data to the template. The results indicate that blood-flow artifacts may seriously impact data processing that depends on registration to a template, that is, most all data processing.

PLSNet: Position-aware GCN-based autism spectrum disorder diagnosis via FC learning and ROIs sifting.

Brain function connectivity, derived from functional magnetic resonance imaging (fMRI), has enjoyed high popularity in the studies of Autism Spectrum Disorder (ASD) diagnosis. Albeit rapid progress has been made, most studies still suffer from several knotty issues: (1) the hardship of modeling the sophisticated brain neuronal connectivity; (2) the mismatch of identically graph node setup to the variations of different brain regions; (3) the dimensionality explosion resulted from excessive voxels in each fMRI sample; (4) the poor interpretability giving rise to unpersuasive diagnosis. To ameliorate these issues, we propose a position-aware graph-convolution-network-based model, namely PLSNet, with superior accuracy and compelling built-in interpretability for ASD diagnosis. Specifically, a time-series encoder is designed for context-rich feature extraction, followed by a function connectivity generator to model the correlation with long range dependencies. In addition, to discriminate the brain nodes with different locations, the position embedding technique is adopted, giving a unique identity to each graph region. We then embed a rarefying method to sift the salient nodes during message diffusion, which would also benefit the reduction of the dimensionality complexity. Extensive experiments conducted on Autism Brain Imaging Data Exchange demonstrate that our PLSNet achieves state-of-the-art performance. Notably, on CC200 atlas, PLSNet reaches an accuracy of 76.4% and a specificity of 78.6%, overwhelming the previous state-of-the-art with 2.5% and 6.5% under five-fold cross-validation policy. Moreover, the most salient brain regions predicted by PLSNet are closely consistent with the theoretical knowledge in the medical domain, providing potential biomarkers for ASD clinical diagnosis. Our code is available at https://github.com/CodeGoat24/PLSNet.

Identifying autism spectrum disorder using edge-centric functional connectivity.

Brain network analysis is an effective method to seek abnormalities in functional interactions for brain disorders such as autism spectrum disorder (ASD). Traditional studies of brain networks focus on the node-centric functional connectivity (nFC), ignoring interactions of edges to miss much information that facilitates diagnostic decisions. In this study, we present a protocol based on an edge-centric functional connectivity (eFC) approach, which significantly improves classification performance by utilizing the co-fluctuations information between the edges of brain regions compared with nFC to build the classification mode for ASD using the multi-site dataset Autism Brain Imaging Data Exchange I (ABIDE I). Our model results show that even using the traditional machine-learning classifier support vector machine (SVM) on the challenging ABIDE I dataset, relatively high performance is achieved: 96.41% of accuracy, 98.30% of sensitivity, and 94.25% of specificity. These promising results suggest that the eFC can be used to build a reliable machine-learning framework to diagnose mental disorders such as ASD and promote identifications of stable and effective biomarkers. This study provides an essential complementary perspective for understanding the neural mechanisms of ASD and may facilitate future investigations on early diagnosis of neuropsychiatric disorders.

DeepMNF: Deep Multimodal Neuroimaging Framework for Diagnosing Autism Spectrum Disorder.

The growing prevalence of neurological disorders, e.g., Autism Spectrum Disorder (ASD), demands robust computer-aided diagnosis (CAD) due to the diverse symptoms which require early intervention, particularly in young children. The absence of a benchmark neuroimaging diagnostics paves the way to study transitions in the brain's anatomical structure and neurological patterns associated with ASD. The existing CADs take advantage of the large-scale baseline dataset from the Autism Brain Imaging Data Exchange (ABIDE) repository to improve diagnostic performance, but the involvement of multisite data also amplifies the variabilities and heterogeneities that hinder satisfactory results. To resolve this problem, we propose a Deep Multimodal Neuroimaging Framework (DeepMNF) that employs Functional Magnetic Resonance Imaging (fMRI) and Structural Magnetic Resonance Imaging (sMRI) to integrate cross-modality spatiotemporal information by exploiting 2-dimensional time-series data along with 3-dimensional images. The purpose is to fuse complementary information that increases group differences and homogeneities. To the best of our knowledge, our DeepMNF achieves superior validation performance than the best reported result on the ABIDE-1 repository involving datasets from all available screening sites. In this work, we also demonstrate the performance of the studied modalities in a single model as well as their possible combinations to develop the multimodal framework.

Altered global modular organization of intrinsic functional connectivity in autism arises from atypical node-level processing.

Autism spectrum disorder (ASD) is a neurodevelopmental condition characterized by restricted interests and repetitive behaviors as well as social-communication deficits. These traits are associated with atypicality of functional brain networks. Modular organization in the brain plays a crucial role in network stability and adaptability for neurodevelopment. Previous neuroimaging research demonstrates discrepancies in studies of functional brain modular organization in ASD. These discrepancies result from the examination of mixed age groups. Furthermore, recent findings suggest that while much attention has been given to deriving atlases and measuring the connections between nodes, within node information may also be crucial in determining altered modular organization in ASD compared with typical development (TD). However, altered modular organization originating from systematic nodal changes are yet to be explored in younger children with ASD. Here, we used graph-theoretical measures to fill this knowledge gap. To this end, we utilized multicenter resting-state fMRI data collected from 5 to 10-year-old children-34 ASD and 40 TD obtained from the Autism Brain Image Data Exchange (ABIDE) I and II. We demonstrate that alterations in topological roles and modular cohesiveness are the two key properties of brain regions anchored in default mode, sensorimotor, and salience networks, and primarily relate to social and sensory deficits in children with ASD. These results demonstrate that atypical global network organization in children with ASD arises from nodal role changes, and contribute to the growing body of literature suggesting that there is interesting information within nodes providing critical markers of functional brain networks in autistic children.

Classifying ASD based on time-series fMRI using spatial-temporal transformer.

As the prevalence of autism spectrum disorder (ASD) increases globally, more and more patients need to receive timely diagnosis and treatment to alleviate their suffering. However, the current diagnosis method of ASD still adopts the subjective symptom-based criteria through clinical observation, which is time-consuming and costly. In recent years, functional magnetic resonance imaging (fMRI) neuroimaging techniques have emerged to facilitate the identification of potential biomarkers for diagnosing ASD. In this study, we developed a deep learning framework named spatial-temporal Transformer (ST-Transformer) to distinguish ASD subjects from typical controls based on fMRI data. Specifically, a linear spatial-temporal multi-headed attention unit is proposed to obtain the spatial and temporal representation of fMRI data. Moreover, a Gaussian GAN-based data balancing method is introduced to solve the data unbalance problem in real-world ASD datasets for subtype ASD diagnosis. Our proposed ST-Transformer is evaluated on a large cohort of subjects from two independent datasets (ABIDE I and ABIDE II) and achieves robust accuracies of 71.0% and 70.6%, respectively. Compared with state-of-the-art methods, our results demonstrate competitive performance in ASD diagnosis.

Exclusion of females in autism research: Empirical evidence for a "leaky" recruitment-to-research pipeline.

Autism spectrum disorder (ASD) is characterized by challenges in social communication and the presence of repetitive behaviors or restricted interests. Notably, males are four times as likely as females to be diagnosed with autism. Despite efforts to increase representation and characterization of autistic females, research studies consistently enroll small samples of females, or exclude females altogether. Importantly, researchers often rely on standardized measures to confirm diagnosis prior to enrollment in research studies. We retrospectively analyzed the effects of one such measure (Autism Diagnostic Observation Schedule, ADOS) on research inclusion/exclusion rates by sex in autistic adults, all of whom had a preexisting community diagnosis of autism (n = 145, 95 male, 50 female). Using the ADOS as a confirmatory diagnostic measure resulted in the exclusion of autistic females at a rate over 2.5 times higher than that of autistic males. We compared sex ratios in our sample to those in other large, publically available datasets that rely either on community diagnosis (6 datasets, total n = 42,209) or standardized assessments (2 datasets, total n = 214) to determine eligibility of participants for research. Reliance on community diagnosis rather than confirmatory diagnostic assessments resulted in significantly more equal sex ratios. These results provide evidence for a "leaky" recruitment-to-research pipeline for females in autism research. LAY SUMMARY: Despite efforts to increase the representation of autistic females in research, studies consistently enroll small samples of females or exclude females altogether. We find that despite making up almost 50% of the initially recruited sample based upon self-report of community diagnosis, autistic females are disproportonately excluded from research participation as a result of commonly used autism diagnostic measures. In our sample, and several other publically available datasets, reliance on community diagnosis resulted in significantly more equal sex ratios.

Sparse Hierarchical Representation Learning on Functional Brain Networks for Prediction of Autism Severity Levels.

Machine learning algorithms have been widely applied in diagnostic tools for autism spectrum disorder (ASD), revealing an altered brain connectivity. However, little is known about whether an magnetic resonance imaging (MRI)-based brain network is related to the severity of ASD symptoms in a large-scale cohort. We propose a graph convolution neural network-based framework that can generate sparse hierarchical graph representations for functional brain connectivity. Instead of assigning initial features for each node, we utilized a feature extractor to derive node features and the extracted representations can be fed to a hierarchical graph self-attention framework to effectively represent the entire graph. By incorporating connectivity embeddings in the feature extractor, we propose adjacency embedding networks to characterize the heterogeneous representations of the brain connectivity. Our proposed model variants outperform the benchmarking model with different configurations of adjacency embedding networks and types of functional connectivity matrices. Using this approach with the best configuration (SHEN atlas for node definition, Tikhonov correlation for connectivity estimation, and identity-adjacency embedding), we were able to predict individual ASD severity levels with a meaningful accuracy: the mean absolute error (MAE) and correlation between predicted and observed ASD severity scores resulted in 0.96, and r = 0.61 (P < 0.0001), respectively. To obtain a better understanding on how to generate better representations, we investigate the relationships between the extracted feature embeddings and the graph theory-based nodal measurements using canonical correlation analysis. Finally, we visualized the model to identify the most contributive functional connections for predicting ASD severity scores.

CNNG: A Convolutional Neural Networks With Gated Recurrent Units for Autism Spectrum Disorder Classification.

As a neurodevelopmental disorder, autism spectrum disorder (ASD) severely affects the living conditions of patients and their families. Early diagnosis of ASD can enable the disease to be effectively intervened in the early stage of development. In this paper, we present an ASD classification network defined as CNNG by combining of convolutional neural network (CNN) and gate recurrent unit (GRU). First, CNNG extracts the 3D spatial features of functional magnetic resonance imaging (fMRI) data by using the convolutional layer of the 3D CNN. Second, CNNG extracts the temporal features by using the GRU and finally classifies them by using the Sigmoid function. The performance of CNNG was validated on the international public data-autism brain imaging data exchange (ABIDE) dataset. According to the experiments, CNNG can be highly effective in extracting the spatio-temporal features of fMRI and achieving a classification accuracy of 72.46%.

Identification of Autism spectrum disorder based on a novel feature selection method and Variational Autoencoder.

The development of noninvasive brain imaging such as resting-state functional magnetic resonance imaging (rs-fMRI) and its combination with AI algorithm provides a promising solution for the early diagnosis of Autism spectrum disorder (ASD). However, the performance of the current ASD classification based on rs-fMRI still needs to be improved. This paper introduces a classification framework to aid ASD diagnosis based on rs-fMRI. In the framework, we proposed a novel filter feature selection method based on the difference between step distribution curves (DSDC) to select remarkable functional connectivities (FCs) and utilized a multilayer perceptron (MLP) which was pretrained by a simplified Variational Autoencoder (VAE) for classification. We also designed a pipeline consisting of a normalization procedure and a modified hyperbolic tangent (tanh) activation function to replace the classical tanh function, further improving the model accuracy. Our model was evaluated by 10 times 10-fold cross-validation and achieved an average accuracy of 78.12%, outperforming the state-of-the-art methods reported on the same dataset. Given the importance of sensitivity and specificity in disease diagnosis, two constraints were designed in our model which can improve the model's sensitivity and specificity by up to 9.32% and 10.21%, respectively. The added constraints allow our model to handle different application scenarios and can be used broadly.

Classification of Autism Spectrum Disorder Using rs-fMRI data and Graph Convolutional Networks.

Autism spectrum disorder (ASD) affects large number of children and adults in the US, and worldwide. Early and quick diagnosis of ASD can improve the quality of life significantly both for patients and their families. Prior research provides strong evidence that structural and functional magnetic resonance imaging (MRI) data collected from individuals with ASD exhibit distinguishing characteristics that differ in local and global, spatial and temporal neural patterns of the brain - and therefore can be used for diagnostic purposes for various mental disorders. However, the data from MRI are high-dimensional and advanced methods are needed to make sense out of these datasets. In this paper, we present a novel model based on graph convolutional network (GCN) that can utilize resting state fMRI (rs-fMRI) data to classify ASD subjects from health controls (HC). In addition to using the graph from traditional correlation matrices, our proposed GCN model incorporates graphlet topological counting as one of the training features. Our results show that graphlets can preserve the topological information of the graphs obtained from fMRI data. Combined with our GCN, the graphlets retain enough topological information to differentiate between the ASD and HC. Our proposed model gives an average accuracy of 64.27% on the whole ABIDE-I data sets (1035 subjects) and highest site-specific accuracy of 75.9%, which is comparable to other state-of-the-art methods - while potentially open to being more interpretable.

The Role of Diffusion Tensor MR Imaging (DTI) of the Brain in Diagnosing Autism Spectrum Disorder: Promising Results.

Autism spectrum disorder (ASD) is a combination of developmental anomalies that causes social and behavioral impairments, affecting around 2% of US children. Common symptoms include difficulties in communications, interactions, and behavioral disabilities. The onset of symptoms can start in early childhood, yet repeated visits to a pediatric specialist are needed before reaching a diagnosis. Still, this diagnosis is usually subjective, and scores can vary from one specialist to another. Previous literature suggests differences in brain development, environmental, and/or genetic factors play a role in developing autism, yet scientists still do not know exactly the pathology of this disorder. Currently, the gold standard diagnosis of ASD is a set of diagnostic evaluations, such as the Autism Diagnostic Observation Schedule (ADOS) or Autism Diagnostic Interview-Revised (ADI-R) report. These gold standard diagnostic instruments are an intensive, lengthy, and subjective process that involves a set of behavioral and communications tests and clinical history information conducted by a team of qualified clinicians. Emerging advancements in neuroimaging and machine learning techniques can provide a fast and objective alternative to conventional repetitive observational assessments. This paper provides a thorough study of implementing feature engineering tools to find discriminant insights from brain imaging of white matter connectivity and using a machine learning framework for an accurate classification of autistic individuals. This work highlights important findings of impacted brain areas that contribute to an autism diagnosis and presents promising accuracy results. We verified our proposed framework on a large publicly available DTI dataset of 225 subjects from the Autism Brain Imaging Data Exchange-II (ABIDE-II) initiative, achieving a high global balanced accuracy over the 5 sites of up to 99% with 5-fold cross validation. The data used was slightly unbalanced, including 125 autistic subjects and 100 typically developed (TD) ones. The achieved balanced accuracy of the proposed technique is the highest in the literature, which elucidates the importance of feature engineering steps involved in extracting useful knowledge and the promising potentials of adopting neuroimaging for the diagnosis of autism.