Given the large scale of public functional Magnetic Resonance Imaging (fMRI), e.g., UK Biobank (UKB) and Human Connectome Projects (HCP), brain foundation models are emerging. Although the amount of samples under rich environmental variables is unprecedented, existing brain foundation models learn from fMRI derived from a narrow range of cognitive states stimulated by similar environments, causing the limited robustness demonstrated in various applications and datasets acquired with different pipelines and limited sample size. By capitalizing on the variety of cognitive status as subjects performing explicit tasks, we present the mixture of brain experts, namely BrainMoE, pre-training on tasking fMRI with rich behavioral tasks in addition to resting fMRI for a robust brain foundation model. Brain experts are designed to produce embeddings for different behavioral tasks related to cognition. Afterward, these cognition embeddings are mixed by a cognition adapter via cross-attention so that BrainMoE can handle orthogonal embeddings and be robust on those boutique downstream datasets. We have pre-trained two existing self-regressive architectures and one new supervised architecture as brain experts on 68,251 fMRI scans among UKB and HCP, containing 12 different cognitive states. Then, BrainMoE is evaluated on a variety of applications, including sex, age prediction, human behavior recognition, disease early diagnosis of Autism, Parkinson's disease, Alzheimer's disease, and Schizophrenia, and fMRI-EEG multimodal applications, where promising results in eight datasets from three different pipelines indicate great potential to facilitate current neuroimaging applications in clinical routines.

Uta Frith seems remarkably cheerful and content for someone who's spent six decades trying and failing to get to grips with her life's obsession. "Very little has stood the test of time," she tells me as we sit down in her living room in a leafy estate in Harrow-on-the-Hill, London. Around us, high-ceilinged walls papered in a luxurious red print are barely visible between rammed bookshelves, several model brains and a collection of abstract art. Frith has been searching for the mechanisms that underpin the enigmatic condition of autism ever since she first met profoundly autistic children in the late 1960s. "We could identify them intuitively, but not really scientifically - and I have to say that this is, unfortunately, still the case." Still, Frith's influence on our ever-shifting understanding of autism has been monumental.

Learning a graph topology to reveal the underlying relationship between data entities plays an important role in various machine learning and data analysis tasks. Under the assumption that structured data vary smoothly over a graph, the problem can be formulated as a regularised convex optimisation over a positive semidefinite cone and solved by iterative algorithms. Classic methods require an explicit convex function to reflect generic topological priors, e.g. the `1 penalty for enforcing sparsity, which limits the flexibility and expressiveness in learning rich topological structures. We propose to learn a mapping from node data to the graph structure based on the idea of learning to optimise (L2O). Specifically, our model first unrolls an iterative primal-dual splitting algorithm into a neural network. The key structural proximal projection is replaced with a variational autoencoder that refines the estimated graph with enhanced topological properties. The model is trained in an end-to-end fashion with pairs of node data and graph samples. Experiments on both synthetic and real-world data demonstrate that our model is more efficient than classic iterative algorithms in learning a graph with specific topological properties.

In a place denied access to basic forensic technology--and where people disappear into Israeli detention--the fate of thousands remains unknown. One of them is an autistic teenager. In the early morning dark, Abeer Skaik turned to her husband, Ali Al-Qatta, and said that today would be the day they would find their son. Ali nodded in silence, and she handed him the stack of flyers. Each bore a photograph of 16-year-old Hassan smiling widely, his shoulders loose, wearing a plain red T-shirt. He is looking directly at the camera, unguarded. On top of the page, in large letters, Abeer had written a single word in bold red ink: --an appeal. Abeer watched as Ali stepped into a car with a few close friends and drove away. They started the 30-kilometer trip south, from al-Tuffah, east of Gaza City, to the European Hospital in Khan Younis. They had heard that a group of people detained by Israel, including children, would be released there. The gate was already crowded. Families stood shoulder to shoulder, wrapped in blankets against the cold, clutching photographs and ID cards. Ali distributed the flyers among his friends. When the buses of released detainees arrived, he and the others moved slowly through the narrow gaps between clusters of people. Some of those who had just been released were being pulled into embraces. Ali waited at the edge of each reunion. "Have you seen my son?" he asked. One after another, people shook their heads.

Alexandre Rocha, a French father in Iceland, claims he lost parental rights after opposing his autistic son's sex reassignment, alleging courts prioritized ideology over his rights.

In many biological and medical contexts, we construct a large labeled corpus by aggregating many sources to use in target prediction tasks. Unfortunately, many of the sources may be irrelevant to our target task, so ignoring the structure of the dataset is detrimental. This work proposes a novel approach, the Multiple Domain Matching Network (MDMN), to exploit this structure. MDMN embeds all data into a shared feature space while learning which domains share strong statistical relationships. These relationships are often insightful in their own right, and they allow domains to share strength without interference from irrelevant data. This methodology builds on existing distribution-matching approaches by assuming that source domains are varied and outcomes multi-factorial. Therefore, each domain should only match a relevant subset. Theoretical analysis shows that the proposed approach can have a tighter generalization bound than existing multiple-domain adaptation approaches. Empirically, we show that the proposed methodology handles higher numbers of source domains (up to 21 empirically), and provides state-of-the-art performance on image, text, and multi-channel time series classification, including clinically relevant data of a novel treatment of Autism Spectrum Disorder.

In this paper we propose a novel method that provides contrastive explanations justifying the classification of an input by a black box classifier such as a deep neuralnetwork.

We propose to learn a mapping from node data to the graph structure based on the idea of learning to optimise (L2O).

Among those Robert F. Kennedy Jr. recently named to a federal autism committee are people who tout dangerous treatments and say vaccine manufacturers are "poisoning children." US health secretary Robert F. Kennedy Jr. has filled an autism committee with friends, associates, and former colleagues who believe that autism is caused by vaccines. Autism advocates are now worried the group could pave the way for dangerous pseudoscientific treatments going mainstream. Last week, Kennedy announced an entirely new lineup for the Interagency Autism Coordinating Committee (IACC), a group that recommends what types of autism research the government should fund and provides guidance on the services the autism community requires. The group is typically composed of experts in the area of autism research, along with policy experts and autistic people advocating for their own community.