Artificial intelligence classifies primary progressive aphasia from connected speech.

Neurodegenerative dementia syndromes, such as Primary Progressive Aphasias (PPA), have traditionally been diagnosed based in part on verbal and nonverbal cognitive profiles. Debate continues about whether PPA is best divided into three variants and also regarding the most distinctive linguistic features for classifying PPA variants. In this cross-sectional study, we first harnessed the capabilities of artificial intelligence (AI) and Natural Language Processing (NLP) to perform unsupervised classification of short, connected speech samples from 78 PPA patients. We then used NLP to identify linguistic features that best dissociate the three PPA variants. Large Language Models (LLMs) discerned three distinct PPA clusters, with 88.5% agreement with independent clinical diagnoses. Patterns of cortical atrophy of three data-driven clusters corresponded to the localization in the clinical diagnostic criteria. In the subsequent supervised classification, seventeen distinctive features emerged, including the observation that separating verbs into high and low-frequency types significantly improves classification accuracy. Using these linguistic features derived from the analysis of short, connected speech samples, we developed a classifier that achieved 97.9% accuracy in classifying the four groups (three PPA variants and healthy controls). The data-driven section of this study showcases the ability of LLMs to find natural partitioning in the speech of patients with PPA consistent with conventional variants. In addition, the work identifies a robust set of language features indicative of each PPA variant, emphasizing the significance of dividing verbs into high and low-frequency categories. Beyond improving diagnostic accuracy, these findings enhance our understanding of the neurobiology of language processing.

Natural language processing in psychiatry: the promises and perils of a transformative approach.

A person's everyday language can indicate patterns of thought and emotion predictive of mental illness. Here, we discuss how natural language processing methods can be used to extract indicators of mental health from language to help address long-standing problems in psychiatry, along with the potential hazards of this new technology.

A big data analysis of the relationship between future thinking and decision-making.

We use big data methods to investigate how decision-making might depend on future sightedness (that is, on how far into the future people's thoughts about the future extend). In study 1, we establish a link between future thinking and decision-making at the population level in showing that US states with citizens having relatively far future sightedness, as reflected in their tweets, take fewer risks than citizens in states having relatively near future sightedness. In study 2, we analyze people's tweets to confirm a connection between future sightedness and decision-making at the individual level in showing that people with long future sightedness are more likely to choose larger future rewards over smaller immediate rewards. In study 3, we show that risk taking decreases with increases in future sightedness as reflected in people's tweets. The ability of future sightedness to predict decisions suggests that future sightedness is a relatively stable cognitive characteristic. This implication was supported in an analysis of tweets by over 38,000 people that showed that future sightedness has both state and trait characteristics (study 4). In study 5, we provide evidence for a potential mechanism by which future sightedness can affect decisions in showing that far future sightedness can make the future seem more connected to the present, as reflected in how people refer to the present, past, and future in their tweets over the course of several minutes. Our studies show how big data methods can be applied to naturalistic data to reveal underlying psychological properties and processes.

Predicting future mental illness from social media: A big-data approach.

In the present research, we investigated whether people's everyday language contains sufficient signal to predict the future occurrence of mental illness. Language samples were collected from the social media website Reddit, drawing on posts to discussion groups focusing on different kinds of mental illness (clinical subreddits), as well as on posts to discussion groups focusing on nonmental health topics (nonclinical subreddits). As expected, words drawn from the clinical subreddits could be used to distinguish several kinds of mental illness (ADHD, anxiety, bipolar disorder, and depression). Interestingly, words drawn from the nonclinical subreddits (e.g., travel, cooking, cars) could also be used to distinguish different categories of mental illness, implying that the impact of mental illness spills over into topics unrelated to mental illness. Most importantly, words derived from the nonclinical subreddits predicted future postings to clinical subreddits, implying that everyday language contains signal about the likelihood of future mental illness, possibly before people are aware of their mental health condition. Finally, whereas models trained on clinical subreddits learned to focus on words indicating disorder-specific symptoms, models trained to predict future mental illness learned to focus on words indicating life stress, suggesting that kinds of features that are predictive of mental illness may change over time. Implications for the underlying causes of mental illness are discussed.

Using explainable artificial intelligence to identify linguistic biomarkers of amyloid pathology in primary progressive aphasia.

Introduction: Recent success has been achieved in Alzheimer's disease (AD) clinical trials targeting amyloid beta (ß), demonstrating a reduction in the rate of cognitive decline. However, testing methods for amyloid-ß positivity are currently costly or invasive, motivating the development of accessible screening approaches to steer patients toward appropriate diagnostic tests. Here, we employ a pre-trained language model (Distil-RoBERTa) to identify amyloid-ß positivity from a short, connected speech sample. We further use explainable AI (XAI) methods to extract interpretable linguistic features that can be employed in clinical practice. Methods: We obtained language samples from 74 patients with primary progressive aphasia (PPA) across its three variants. Amyloid-ß positivity was established through the analysis of cerebrospinal fluid, amyloid PET, or autopsy. 51% of the sample was amyloid-positive. We trained Distil-RoBERTa for 16 epochs with a batch size of 6 and a learning rate of 5e-5, and used the LIME algorithm to train interpretation models to interpret the trained classifier's inference conditions. Results: Over ten runs of 10-fold cross-validation, the classifier achieved a mean accuracy of 92%, SD = 0.01. Interpretation models were able to capture the classifier's behavior well, achieving an accuracy of 97% against classifier predictions, and uncovering several novel speech patterns that may characterize amyloid-ß positivity. Discussion: Our work improves previous research which indicates connected speech is a useful diagnostic input for prediction of the presence of amyloid-ß in patients with PPA. Further, we leverage XAI techniques to reveal novel linguistic features that can be tested in clinical practice in the appropriate subspecialty setting. Computational linguistic analysis of connected speech shows great promise as a novel assessment method in patients with AD and related disorders.

Spatial language as a window on representations of three-dimensional space.

Recent research investigating the language­thought interface in the spatial domain points to representations of the horizontal and vertical dimensions that closely resemble those posited by Jeffery et al. However, the findings suggest that such representations, rather than being tied to navigation, may instead reflect more general properties of the perception of space.

Spatial language and the psychological reality of schematization.

Although the representations underlying spatial language are often assumed to be schematic in nature, empirical evidence for a schematic format of representation is lacking. In this research, we investigate the psychological reality of such a format, using simulated motion during scene processing--previously linked to schematization--as a diagnostic. One group of participants wrote a verbal description of a scene and then completed a change detection task assessing simulated motion, while another group completed only the latter task. We expected that effects of simulated motion would be stronger following language use than not, and specifically following the use of spatial, relative to non-spatial, language. Both predictions were supported. Further, the effect of language was scene independent, suggesting that language may encourage a general mode of schematic construal. The study and its findings illustrate a novel approach to examining the perceptual properties of mental representations.

Using Generative Artificial Intelligence to Classify Primary Progressive Aphasia from Connected Speech.

Neurodegenerative dementia syndromes, such as Primary Progressive Aphasias (PPA), have traditionally been diagnosed based in part on verbal and nonverbal cognitive profiles. Debate continues about whether PPA is best subdivided into three variants and also regarding the most distinctive linguistic features for classifying PPA variants. In this study, we harnessed the capabilities of artificial intelligence (AI) and natural language processing (NLP) to first perform unsupervised classification of concise, connected speech samples from 78 PPA patients. Large Language Models discerned three distinct PPA clusters, with 88.5% agreement with independent clinical diagnoses. Patterns of cortical atrophy of three data-driven clusters corresponded to the localization in the clinical diagnostic criteria. We then used NLP to identify linguistic features that best dissociate the three PPA variants. Seventeen features emerged as most valuable for this purpose, including the observation that separating verbs into high and low-frequency types significantly improves classification accuracy. Using these linguistic features derived from the analysis of brief connected speech samples, we developed a classifier that achieved 97.9% accuracy in predicting PPA subtypes and healthy controls. Our findings provide pivotal insights for refining early-stage dementia diagnosis, deepening our understanding of the characteristics of these neurodegenerative phenotypes and the neurobiology of language processing, and enhancing diagnostic evaluation accuracy.

Language as a biomarker for psychosis: A natural language processing approach.

Human ratings of conceptual disorganization, poverty of content, referential cohesion and illogical thinking have been shown to predict psychosis onset in prospective clinical high risk (CHR) cohort studies. The potential value of linguistic biomarkers has been significantly magnified, however, by recent advances in natural language processing (NLP) and machine learning (ML). Such methodologies allow for the rapid and objective measurement of language features, many of which are not easily recognized by human raters. Here we review the key findings on language production disturbance in psychosis. We also describe recent advances in the computational methods used to analyze language data, including methods for the automatic measurement of discourse coherence, syntactic complexity, poverty of content, referential coherence, and metaphorical language. Linguistic biomarkers of psychosis risk are now undergoing cross-validation, with attention to harmonization of methods. Future directions in extended CHR networks include studies of sources of variance, and combination with other promising biomarkers of psychosis risk, such as cognitive and sensory processing impairments likely to be related to language. Implications for the broader study of social communication, including reciprocal prosody, face expression and gesture, are discussed.

A machine learning approach to predicting psychosis using semantic density and latent content analysis.

Subtle features in people's everyday language may harbor the signs of future mental illness. Machine learning offers an approach for the rapid and accurate extraction of these signs. Here we investigate two potential linguistic indicators of psychosis in 40 participants of the North American Prodrome Longitudinal Study. We demonstrate how the linguistic marker of semantic density can be obtained using the mathematical method of vector unpacking, a technique that decomposes the meaning of a sentence into its core ideas. We also demonstrate how the latent semantic content of an individual's speech can be extracted by contrasting it with the contents of conversations generated on social media, here 30,000 contributors to Reddit. The results revealed that conversion to psychosis is signaled by low semantic density and talk about voices and sounds. When combined, these two variables were able to predict the conversion with 93% accuracy in the training and 90% accuracy in the holdout datasets. The results point to a larger project in which automated analyses of language are used to forecast a broad range of mental disorders well in advance of their emergence.

Representing causation.

The dynamics model, which is based on L. Talmy's (1988) theory of force dynamics, characterizes causation as a pattern of forces and a position vector. In contrast to counterfactual and probabilistic models, the dynamics model naturally distinguishes between different cause-related concepts and explains the induction of causal relationships from single observations. Support for the model is provided in experiments in which participants categorized 3-D animations of realistically rendered objects with trajectories that were wholly determined by the force vectors entered into a physics simulator. Experiments 1-3 showed that causal judgments are based on several forces, not just one. Experiment 4 demonstrated that people compute the resultant of forces using a qualitative decision rule. Experiments 5 and 6 showed that a dynamics approach extends to the representation of social causation. Implications for the relationship between causation and time are discussed.

Causal reasoning with forces.

Causal composition allows people to generate new causal relations by combining existing causal knowledge. We introduce a new computational model of such reasoning, the force theory, which holds that people compose causal relations by simulating the processes that join forces in the world, and compare this theory with the mental model theory (Khemlani et al., 2014) and the causal model theory (Sloman et al., 2009), which explain causal composition on the basis of mental models and structural equations, respectively. In one experiment, the force theory was uniquely able to account for people's ability to compose causal relationships from complex animations of real-world events. In three additional experiments, the force theory did as well as or better than the other two theories in explaining the causal compositions people generated from linguistically presented causal relations. Implications for causal learning and the hierarchical structure of causal knowledge are discussed.

Direct causation in the linguistic coding and individuation of causal events.

This research proposes a new theory of direct causation and examines how this concept plays a key role in the linguistic coding and individuation of causal events. According to the no-intervening-cause hypothesis, a causal chain can be described by a single-clause sentence and construed as a single event if there are no intervening causers between the initial causer and the final causee. Consistent with this hypothesis, participants used single-clause sentences (lexical causatives) more often than two-clause sentences (e.g. periphrastic causatives) for causal chains in which (1) the causer and causee touched (Experiments 1 and 2), and (2) an intervening entity could be construed as an enabling condition rather than another cause (Experiments 2-4). In addition, event judgments paralleled linguistic descriptions: chains that could be described with single-clause expressions were more often construed as single events than chains that could not (Experiments 1-3). Implications for languages other than English, for the linguistic coding of accidental outcomes and for the relationship between cognition and language in general are discussed.

Does categorical perception in the left hemisphere depend on language?

Categorical perception (CP) refers to the influence of category knowledge on perception and is revealed by a superior ability to discriminate items across categories relative to items within a category. In recent years, the finding that CP is lateralized to the left hemisphere in adults has been interpreted as evidence for a kind of CP driven by language. The present research challenges this conclusion. In 2 experiments, we found that CP for novel object categories was stronger in the left hemisphere than in the right, consistent with a role for language. However, both labeled and unlabeled categories gave rise to such effects, and to comparable degrees. These results suggest that left-lateralized CP does not depend on language but rather may reflect the left hemisphere's more general propensity for categorical processing. Our findings raise implications for research on linguistic relativity.

Structure-mapping in metaphor comprehension.

Metaphor has a double life. It can be described as a directional process in which a stable, familiar base domain provides inferential structure to a less clearly specified target. But metaphor is also described as a process of finding commonalities, an inherently symmetric process. In this second view, both concepts may be altered by the metaphorical comparison. Whereas most theories of metaphor capture one of these aspects, we offer a model based on structure-mapping that captures both sides of metaphor processing. This predicts (a) an initial processing stage of symmetric alignment; and (b) a later directional phase in which inferences are projected to the target. To test these claims, we collected comprehensibility judgments for forward (e.g., "A rumor is a virus") and reversed ("A virus is a rumor") metaphors at early and late stages of processing, using a deadline procedure. We found an advantage for the forward direction late in processing, but no directional preference early in processing. Implications for metaphor theory are discussed.

Linguistic relativity.

The central question in research on linguistic relativity, or the Whorfian hypothesis, is whether people who speak different languages think differently. The recent resurgence of research on this question can be attributed, in part, to new insights about the ways in which language might impact thought. We identify seven categories of hypotheses about the possible effects of language on thought across a wide range of domains, including motion, color, spatial relations, number, and false belief understanding. While we do not find support for the idea that language determines the basic categories of thought or that it overwrites preexisting conceptual distinctions, we do find support for the proposal that language can make some distinctions difficult to avoid, as well as for the proposal that language can augment certain types of thinking. Further, we highlight recent evidence suggesting that language may induce a relatively schematic mode of thinking. Although the literature on linguistic relativity remains contentious, there is growing support for the view that language has a profound effect on thought. WIREs Cogni Sci 2011 2 253-265 DOI: 10.1002/wcs.104 For further resources related to this article, please visit the WIREs website.

For want of a nail: How absences cause events.

Causation by omission is instantiated when an effect occurs from an absence, as in The absence of nicotine causes withdrawal or Not watering the plant caused it to wilt. The phenomenon has been viewed as an insurmountable problem for process theories of causation, which specify causation in terms of conserved quantities, like force, but not for theories that specify causation in terms of statistical or counterfactual dependencies. A new account of causation challenges these assumptions. According to the force theory, absences are causal when the removal of a force leads to an effect. Evidence in support of this account was found in 3 experiments in which people classified animations of complex causal chains involving force removal, as well as chains involving virtual forces, that is, forces that were anticipated but never realized. In a 4th experiment, the force theory's ability to predict synonymy relationships between different types of causal expressions provided further evidence for this theory over dependency theories. The findings show not only how causation by omission can be grounded in the physical world but also why only certain absences, among the potentially infinite number of absences, are causal.

Models of causation and the semantics of causal verbs.

This research examines the relationship between the concept of CAUSE as it is characterized in psychological models of causation and the meaning of causal verbs, such as the verb cause itself. According to focal set models of causation (; ), the concept of CAUSE should be more similar to the concepts of ENABLE and PREVENT than either is to each other. According to a model based on theory of force dynamics, the force dynamic model, the concepts of CAUSE, ENABLE, and PREVENT should be roughly equally similar to one another. The relationship between these predictions and the meaning of causal verbs was examined by having participants sort causal verbs and rate them with respect to the dimensions specified by the two models. The results from five experiments indicated that the force dynamic model provides a better account of the meaning of causal verbs than do focal set models of causation. Implications for causal inference and induction are discussed.

Fertilizaçäo in vitro com ciclos programados de baixo custo - Avaliaçäo de resultados iniciais de um Centro de Reproduçäo Humana de hospital de ensino / In vitro fertilization with low-cost programmed cycles - First outcome in a teaching hospital

Objetivos: avaliaçäo dos resultados iniciais de fertilizaçäo in vitro (FIV) em instituiçäo universitária, empregando ciclos programados de baixo custo. Métodos: entre maio e dezembro de 2002, foram iniciados 66 ciclos programados de FIV, utilizando acetato de noretisterona, citrato de clomifeno e gonadotrofina coriônica humana (hCG). A punçäo folicular guiada por ultra-sonografia foi realizada 34 a 36 horas após a administraçäo de hCG e a transferência, 48 horas após a punçäo. A gestaçäo foi considerada clínica após visualizaçäo de batimentos cardíacos à ultra-sonografia transvaginal. Resultados: a taxa de cancelamento foi de 21,2 por cento. Em média, 2,8 folículos e 1,7 oócitos foram obtidos por punçäo. Em 79,6 por cento dos ciclos puncionados recuperaram-se oócitos, que foram fertilizados em 69 por cento dos casos. O número de embriões por transferência foi de 1,5. Houve algum grau de dificuldade em 10,2 por cento das punções e 32,4 por cento das transferências realizadas. A taxa de gestaçäo obtida foi de 10,8 por cento por transferência, entretanto, o custo com medicaçäo por embriäo transferido foi de apenas R$ 96,00. Conclusäo: evidenciam-se as dificuldades de iniciar um programa de FIV em instituiçäo de ensino, sem fins lucrativos e voltada à populaçäo carente. Com a prática, a taxa cumulativa de gravidez tende a ser semelhante às dos centros de referência, porém com custo e incidência de complicações significativamente inferiores