The implementation of a database of digitised mammograms is discussed. The digitised images were collected beginning in 1999 by a community of physicists in collaboration with radiologists in several Italian
Investigation of functional magnetic resonance imaging (fMRI) data with machine learning (ML) techniques, including also deep learning (DL) methods, have been widely used to study Autism Spectrum Disorder (ASD). This disorder is characterized by symptoms that affect the individual’s behavioral aspects and social relationships. Early diagnosis is crucial for intervention, but the complexity of ASD poses challenges for treatment development. This study compares traditional ML techniques with deep learning (DL) methods in the analysis of functional connectivity measures obtained from the time series of multicentric ABIDE dataset. Specifically, Support Vector Machines (SVM) classifiers, with both linear and Radial Basis Function (RBF) kernels, as well as eXtreme Gradient Boosting (XGBoost) classifiers, are compared against the TabNet classifier, which is a DL architecture customized for tabular data analysis and a Multi Layer Perceptron (MLP). The findings suggest that DL classifiers may not be optimal for the type of data analyzed, as their performance trails behind that of standard classifiers. SVMs achieve performances, in terms of AUC, around 75%, compared to the best TabNet and MLP results, which are 65% and 71%, respectively. Additionally, this work investigates the brain regions that contribute most to the classification task, which are found to be those primarily responsible for sensory and spatial perception, as well as attention modulation, known to be altered in ASD.
Background: The brainstem is a neglected topic in autism research, despite major lines of evidence indicating its active involvement in sensory, motor, affect, arousal, and social regulation (Dadalko & Travers, 2018). It is the substrate of what affective neuroscience identifies as the ‘Core Self’ (Alcaro, Carta, & Panksepp, 2017), and disruption to its growth and function appears to disturb core conscious experience in autism (Delafield-Butt & Trevarthen, 2017; Trevarthen & Delafield-Butt, 2013). Yet, although evidence indicates brainstem growth is disrupted in early childhood (Bosco et al., 2018), how these growth differences compare to closely related neurodevelopmental disorders, such a Developmental Coordination Disorder (DCD), is not yet understood. Objectives: To determine brainstem morphometric differences between children with ASD, DCD, and those typically developing (TD). Methods: Study participants were 87 youths ages 8 to 17 assigned to the ASD (n = 30, 7 female), DCD (n =24, 12 female) or TD (n = 33, 12 female) group. Exclusion criteria for all groups included IQ <80. TD were excluded if they had any neuropsychological or psychopathological disorder. DCD eligibility additionally included performance 16th percentile on the MABC-2 and no concern about an ASD diagnosis. ASD participants had a previous clinical diagnosis confirmed by ADOS-2 and ADI-R. Individuals were excluded if they had another neuropsychological disorder, except attention deficit or anxiety disorder. T1-weighted MPRAGE (1mm isotropic resolution) MRI data were acquired on a 3T MAGNETOM Prisma (Siemens). Brainstem morphology was analysed using SPHARM-MAT (http://lishenlab.com/spharm/), a 3D Fourier surface representation method¬¬¬. A typical surface was calculated for the TD group, and distances from this norm computed for each vertex. Mean distances at each vertex were computed for each group (ASD, DCD, TD) and compared, taking into account age, gender and supratentorial volume as covariates. Results: Significant brainstem morphological differences were identified between all three (TD, ASD and DCD; Figure 1). Significant differences between TD and ASD (p<0.01) were identified in a large region of the anterior-most surface, extending caudally along the right posterior surface. Differences between TD and DCD groups were similar with reduced significance (p0.01), and the pattern diverged with more inclusion of the anterior ventricular surface and less pronouncement at the right anterior border. Finally, significant differences were found between ASD and DCD groups (p<0.01), specifically at the anterior midline either side of the ventricular surface, and especially in two long anteroposterior columns on the left side adjacent and parallel to the fourth ventricle. Conclusions: Surface morphology differences indicate alterations in local nuclei and/or tract growth within the brainstem, especially approaching the anterior surface in ASD and DCD children, and differentially between them at the ventricular surface. The former may relate to specific nerve growth of the pons, and the latter to cerebellar peduncle connectivity differences, superficial nuclei growth such as the hypoglossal, intercalatus, or vagus and associated tracts, or deeper nuclei such as the inferior olivary nucleus. Brainstem structural differences likely disturbs the integrative function of the Core Self. Higher resolution 7T MRI is required to resolve the underlying differential composition.
Local specific absorption rate (SAR) evaluation in ultra high field (UHF) magnetic resonance (MR) systems is a major concern. In fact, at UHF, radiofrequency (RF) field inhomogeneity generates hot-spots that could cause localized tissue heating. Unfortunately, local SAR measurements are not available in present MR systems; thus, electromagnetic simulations must be performed for RF fields and SAR analysis. In this study, we used three-dimensional full-wave numerical electromagnetic simulations to investigate the dependence of local SAR at 7.0 T with respect to subject size in two different scenarios: surface coil loaded by adult and child calves and quadrature volume coil loaded by adult and child heads. In the surface coil scenario, maximum local SAR decreased with decreasing load size, provided that the RF magnetic fields for the different load sizes were scaled to achieve the same slice average value. On the contrary, in the volume coil scenario, maximum local SAR was up to 15% higher in children than in adults.
The purpose of the work here presented consists in the evaluation of the performance of CAD (Computer Aided Detection) systems for automated lung nodule identification on multislice CT examinations based on different analysis approaches and on their combination. Three different CADe systems, the CAM CAD (Channeler Ant Model), the RGVP CAD (Region Growing Volume Plateau) and the VBNA CAD (Voxel Based Neural Approach) were tested on public research datasets and evaluated in terms of FROC (Free-response Receiver Operating Characteristics) curves both individually and combined.
