Older people with mild cognitive impairment, especially when characterized by episodic memory loss, are at increased risk for dementia due to Alzheimer's disease. Now a study by researchers from MIT, Cornell, and Massachusetts General Hospital has identified a key deficit unrelated to memory that may help reveal the condition early--when any available treatments are likely to be most effective. The issue has to do with a subtle aspect of language processing: people with amnestic mild cognitive impairment (aMCI) struggle with certain ambiguous sentences in which pronouns could refer to people not referenced in the sentences themselves.For instance, in "The electrician fixed the light switch when he visited the tenant," it is not clear without context whether "he" refers to the electrician or some other visitor. But in "He visited the tenant when the electrician repaired the light switch," "he" and "the electrician" cannot be the same person. And in "The babysitter emptied the bottle and prepared the formula," there is no reference to a person beyond the sentence.

"Just Accepted" papers have undergone full peer review and have been accepted for publication in Radiology: Artificial Intelligence. This article will undergo copyediting, layout, and proof review before it is published in its final version. Please note that during production of the final copyedited article, errors may be discovered which could affect the content. To determine whether a brain age prediction model could quantify individual deviations from a healthy brain-aging trajectory (predicted age difference [PAD]) in patients with amnestic mild cognitive impairment (aMCI) and to determine if PAD was associated with individual cognitive impairment. In this retrospective study, a machine learning approach was trained to determine brain age based on T1-weighted MRI.

Current approaches to amortizing Bayesian inference focus solely on approximating the posterior distribution. Typically, this approximation is, in turn, used to calculate expectations for one or more target functions - a computational pipeline which is inefficient when the target function(s) are known upfront. In this paper, we address this inefficiency by introducing AMCI, a method for amortizing Monte Carlo integration directly. AMCI operates similarly to amortized inference but produces three distinct amortized proposals, each tailored to a different component of the overall expectation calculation. At runtime, samples are produced separately from each amortized proposal, before being combined to an overall estimate of the expectation. We show that while existing approaches are fundamentally limited in the level of accuracy they can achieve, AMCI can theoretically produce arbitrarily small errors for any integrable target function using only a single sample from each proposal at runtime. We further show that it is able to empirically outperform the theoretically optimal self-normalized importance sampler on a number of example problems. Furthermore, AMCI allows not only for amortizing over datasets but also amortizing over target functions.