Application of Machine Learning Techniques to Assess Alpha-Fetoprotein at Diagnosis of Hepatocellular Carcinoma.

Hepatocellular carcinoma (HCC) is the most common primary liver tumor and is associated with high mortality rates. Approximately 80% of cases occur in cirrhotic livers, posing a significant challenge for appropriate therapeutic management. Adequate screening programs in high-risk groups are essential for early-stage detection. The extent of extrahepatic tumor spread and hepatic functional reserve are recognized as two of the most influential prognostic factors. In this retrospective multicenter study, we utilized machine learning (ML) methods to analyze predictors of mortality at the time of diagnosis in a total of 208 patients. The eXtreme gradient boosting (XGB) method achieved the highest values in identifying key prognostic factors for HCC at diagnosis. The etiology of HCC was found to be the variable most strongly associated with a poorer prognosis. The widely used Barcelona Clinic Liver Cancer (BCLC) classification in our setting demonstrated superiority over the TNM classification. Although alpha-fetoprotein (AFP) remains the most commonly used biological marker, elevated levels did not correlate with reduced survival. Our findings suggest the need to explore new prognostic biomarkers for individualized management of these patients.

Prognostic Factors for Mortality in Hepatocellular Carcinoma at Diagnosis: Development of a Predictive Model Using Artificial Intelligence.

BACKGROUND: Hepatocellular carcinoma (HCC) accounts for 75% of primary liver tumors. Controlling risk factors associated with its development and implementing screenings in risk populations does not seem sufficient to improve the prognosis of these patients at diagnosis. The development of a predictive prognostic model for mortality at the diagnosis of HCC is proposed. METHODS: In this retrospective multicenter study, the analysis of data from 191 HCC patients was conducted using machine learning (ML) techniques to analyze the prognostic factors of mortality that are significant at the time of diagnosis. Clinical and analytical data of interest in patients with HCC were gathered. RESULTS: Meeting Milan criteria, Barcelona Clinic Liver Cancer (BCLC) classification and albumin levels were the variables with the greatest impact on the prognosis of HCC patients. The ML algorithm that achieved the best results was random forest (RF). CONCLUSIONS: The development of a predictive prognostic model at the diagnosis is a valuable tool for patients with HCC and for application in clinical practice. RF is useful and reliable in the analysis of prognostic factors in the diagnosis of HCC. The search for new prognostic factors is still necessary in patients with HCC.

Bayesian reasoning with emotional material in patients with schizophrenia.

Delusions are one of the most classical symptoms described in schizophrenia. However, despite delusions are often emotionally charged, they have been investigated using tasks involving non-affective material, such as the Beads task. In this study we compared 30 patients with schizophrenia experiencing delusions with 32 matched controls in their pattern of responses to two versions of the Beads task within a Bayesian framework. The two versions of the Beads task consisted of one emotional and one neutral, both with ratios of beads of 60:40 and 80:20, considered, respectively, as the "difficult" and "easy" variants of the task. Results indicate that patients showed a greater deviation from the normative model, especially in the 60:40 ratio, suggesting that more inaccurate probability estimations are more likely to occur under uncertainty conditions. Additionally, both patients and controls showed a greater deviation in the emotional version of the task, providing evidence of a reasoning bias modulated by the content of the stimuli. Finally, a positive correlation between patients' deviation and delusional symptomatology was found. Impairments in the 60:40 ratio with emotional content was related to the amount of disruption in life caused by delusions. These results contribute to the understanding of how cognitive mechanisms interact with characteristics of the task (i.e., ambiguity and content) in the context of delusional thinking. These findings might be used to inform improved intervention programs in the domain of inferential reasoning.

Effects of affective and emotional congruency on facial expression processing under different task demands.

Contextual influences on responses to facial expressions of emotion were studied using a context-target paradigm that allowed distinguishing the effects of affective congruency (context and target of same/different valence: positive or negative) and emotional congruency (context and target representing the same/different emotion: anger, fear, happiness). Sentences describing anger, fear or happiness-inducing events and faces expressing each of these emotions were used as contexts and targets, respectively. While between-valence comparisons (context and target of similar/different valence) revealed affective congruency effects, within-valence comparisons (context and target of similar valence and same/different emotion) revealed emotional congruency effects. In Experiment 1 no evidence of emotional congruency and limited evidence of affective congruency were found with an evaluative task. In Experiment 2 effects of both affective and emotional congruency were observed with an emotion recognition task. In this case, angry and fearful faces were recognized faster in emotionally congruent contexts. In Experiment 3 the participants were asked explicitly to judge the emotional congruency of the target faces. Emotional congruency effects were again found, with faster judgments of angry and fearful faces in the corresponding emotional contexts. Moreover, judgments of angry expressions were faster and more accurate in happy than in anger contexts. Thus, participants found easier to decide that angry faces did not match a happy context than to judge that they did match an anger context. These results suggest that there are differences in the way that facial expressions of positive and negative emotions are discriminated and integrated with their contexts. Specifically, compared to positive expressions, contextual integration of negative expressions seems to require a double check of the valence and the specific emotion category of the expression and the context.

When birds and sias fly: A neural indicator of inferring a word meaning in context.

Inference generation is a crucial skill in language comprehension. Recent research suggests that readers use both the contents from prior written text and their background knowledge, stored in long-term memory, to generate predictive inferences about what will come up next in a sentence. We recorded Event-Related Potentials (ERPs) to examine the reader's ability to make online inferences even in the presence of pseudowords (orthographically legal, but meaningless letter strings), that is, in the presence of referents with no a priori match to vocabulary stored knowledge. As expected, a large and sustained negativity (250-900ms) was elicited by the target word 'fly' when preceded by the pseudoword 'Sias' in the sentence 'Sias fly.' relative to when preceded by 'Birds' in the sentence 'Birds fly'. However, when readers were provided with an initial statement inviting to make an inference: 'Sias have wings', the word 'fly' in 'Sias fly' only elicited a negative voltage deflection over 100ms period (250-350), rapidly falling down to baseline. This result indicates that participants rapidly generated online inferences even with a hindered access to a referent's meaning (i.e. not knowing what 'Sias' are). Remarkably, brainwave traces to the access to a word's meaning in long-term memory (access to a well-known fact such as 'Birds fly') only diverged from ERPs for an inferred-from-reading knowledge ('Sias fly') for 100ms. We conclude that a fundamental search for across sentence coherence drives fast inference making processes in reading tasks. This pattern of brain response is critical to understand the rapid acquisition of new vocabulary when learning first and second languages.

Moved by words: Affective ratings for a set of 2,266 Spanish words in five discrete emotion categories.

The two main theoretical accounts of the human affective space are the dimensional perspective and the discrete-emotion approach. In recent years, several affective norms have been developed from a dimensional perspective, including ratings for valence and arousal. In contrast, the number of published datasets relying on the discrete-emotion approach is much lower. There is a need to fill this gap, considering that discrete emotions have an effect on word processing above and beyond those of valence and arousal. In the present study, we present ratings from 1,380 participants for a set of 2,266 Spanish words in five discrete emotion categories: happiness, anger, fear, disgust, and sadness. This will be the largest dataset published to date containing ratings for discrete emotions. We also present, for the first time, a fine-grained analysis of the distribution of words into the five emotion categories. This analysis reveals that happiness words are the most consistently related to a single, discrete emotion category. In contrast, there is a tendency for many negative words to belong to more than one discrete emotion. The only exception is disgust words, which overlap least with the other negative emotions. Normative valence and arousal data already exist for all of the words included in this corpus. Thus, the present database will allow researchers to design studies to contrast the predictions of the two most influential theoretical perspectives in this field. These studies will undoubtedly contribute to a deeper understanding of the effects of emotion on word processing.

The Madrid Affective Database for Spanish (MADS): Ratings of Dominance, Familiarity, Subjective Age of Acquisition and Sensory Experience.

The current study presents ratings by 540 Spanish native speakers for dominance, familiarity, subjective age of acquisition (AoA), and sensory experience (SER) for the 875 Spanish words included in the Madrid Affective Database for Spanish (MADS). The norms can be downloaded as supplementary materials for this manuscript from https://figshare.com/s/8e7b445b729527262c88 These ratings may be of potential relevance to researches who are interested in characterizing the interplay between language and emotion. Additionally, with the aim of investigating how the affective features interact with the lexicosemantic properties of words, we performed correlational analyses between norms for familiarity, subjective AoA and SER, and scores for those affective variables which are currently included in the MADs. A distinct pattern of significant correlations with affective features was found for different lexicosemantic variables. These results show that familiarity, subjective AoA and SERs may have independent effects on the processing of emotional words. They also suggest that these psycholinguistic variables should be fully considered when formulating theoretical approaches to the processing of affective language.