Autism spectrum disorder (ASD) is a developmental disorder, affecting about 1% of the global population. Currently, the only clinical method for diagnosing ASD are standardized ASD tests which require prolonged diagnostic time and increased medical costs. Our objective was to explore the predictive power of personal characteristic data (PCD) from a large well-characterized dataset to improve upon prior diagnostic models of ASD. We extracted six personal characteristics (age, sex, handedness, and three individual measures of IQ) from 851 subjects in the Autism Brain Imaging Data Exchange (ABIDE) database. ABIDE is an international collaborative project that collected data from a large number of ASD patients and typical non-ASD controls from 17 research and clinical institutes. We employed this publicly available database to test nine supervised machine learning models. We implemented a cross-validation strategy to train and test those machine learning models for classification between typical non-ASD controls and ASD patients. We assessed classification performance using accuracy, sensitivity, specificity, and area under the receiver operating characteristic curve (AUC). Of the nine models we tested using six personal characteristics, the neural network model performed the best with a mean AUC (SD) of 0.646 (0.005), followed by k-nearest neighbor with a mean AUC (SD) of 0.641 (0.004). This study established an optimal ASD classification performance with PCD as features. With additional discriminative features (e.g., neuroimaging), machine learning models may ultimately enable automated clinical diagnosis of autism.

BACKGROUND AND PURPOSE: Microstructural white matter abnormalities on DTI using Tract-Based Spatial Statistics at term-equivalent age are associated with cognitive and motor outcomes at 2 years of age or younger. However, neurodevelopmental tests administered at such early time points are insufficiently predictive of mild-moderate motor and cognitive impairment at school age. Our objective was to evaluate the microstructural antecedents of cognitive and motor outcomes at 3 years' corrected age in a cohort of very preterm infants. MATERIALS AND METHODS:We prospectively recruited 101 very preterm infants (,32 weeks' gestational age) and performed DTI at term-equivalent age. The Differential Ability Scales, 2nd ed, Verbal and Nonverbal subtests, and the Bayley Scales of Infant and Toddler Development, 3rd ed, Motor subtest, were administered at 3 years of age. We correlated DTI metrics from Tract-Based Spatial Statistics with the Bayley Scales of Infant and Toddler Development, 3rd ed, and the Differential Ability Scales, 2nd ed, scores with correction for multiple comparisons.RESULTS: Of the 101 subjects, 84 had high-quality DTI data, and of these, 69 returned for developmental testing (82%). Their mean (SD) gestational age was 28.4 (2.5) weeks, and birth weight was 1121.4 (394.1) g. DTI metrics were significantly associated with Nonverbal Ability in the corpus callosum, posterior thalamic radiations, fornix, and inferior longitudinal fasciculus and with Motor scores in the corpus callosum, internal and external capsules, posterior thalamic radiations, superior and inferior longitudinal fasciculi, cerebral peduncles, and corticospinal tracts. CONCLUSIONS:We identified widespread microstructural white matter abnormalities in very preterm infants at term that were significantly associated with cognitive and motor development at 3 years' corrected age. ABBREVIATIONS: Bayley-III ¼ Bayley Scales of Infant and Toddler Development, 3rd ed; CA ¼ corrected age; DAS-II ¼ Differential Ability Scales, 2nd ed; FA ¼ fractional anisotropy; IFOF ¼ inferior fronto-occipital fasciculi; ILF ¼ inferior longitudinal fasciculus; MD ¼ mean diffusivity; PTR ¼ posterior thalamic radiations P remature birth is associated with a significantly increased risk of brain abnormalities and long-term neurodevelopmental impairment. Injuries or maturational delays affecting the WM are observed in 50%-80% of very preterm infants. [1][2][3] These abnormalities are associated with serious neurodevelopmental impairment. 1,3,4 However, such abnormalities are challenging to detect using conventional MR imaging techniques alone. Fortunately, DTI, a specialized form of MR imaging that can sensitively query the brain's microstructure, offers a novel approach for identifying these WM injuries. In preterm brains, the evolution of fractional anisotropy (FA) and mean diffusivity (MD), 2 metrics derived from DTI, varies from that of normative populations, and underlying brain injury may lead to neurodevelopmental impairment later in life. [4][5][6] Functional MR i...

Diffuse white matter abnormality (DWMA), or diffuse excessive high signal intensity is observed in 50–80% of very preterm infants at term-equivalent age. It is subjectively defined as higher than normal signal intensity in periventricular and subcortical white matter in comparison to normal unmyelinated white matter on T 2 -weighted MRI images. Despite the well-documented presence of DWMA, it remains debatable whether DWMA represents pathological tissue injury or a transient developmental phenomenon. Manual tracing of DWMA exhibits poor reliability and reproducibility and unduly increases image processing time. Thus, objective and ideally automatic assessment is critical to accurately elucidate the biologic nature of DWMA. We propose a deep learning approach to automatically identify DWMA regions on T 2 -weighted MRI images. Specifically, we formulated DWMA detection as an image voxel classification task; that is, the voxels on T 2 -weighted images are treated as samples and exclusively assigned as DWMA or normal white matter voxel classes. To utilize the spatial information of individual voxels, small image patches centered on the given voxels are retrieved. A deep convolutional neural networks (CNN) model was developed to differentiate DWMA and normal voxels. We tested our deep CNN in multiple validation experiments. First, we examined DWMA detection accuracy of our CNN model using computer simulations. This was followed by in vivo assessments in a cohort of very preterm infants ( N = 95) using cross-validation and holdout validation. Finally, we tested our approach on an independent preterm cohort ( N = 28) to externally validate our model. Our deep CNN model achieved Dice similarity index values ranging from 0.85 to 0.99 for DWMA detection in the aforementioned validation experiments. Our proposed deep CNN model exhibited significantly better performance than other popular machine learning models. We present an objective and automated approach for accurately identifying DWMA that may facilitate the clinical diagnosis of DWMA in very preterm infants.

No abstract

No abstract