Holdout Sample Research Articles

Individual risk of not responding to psychotherapy in Latin America: Bringing data-informed precision care to underresourced clinical settings.

Machine learning has a great potential for prospectively forecasting individual patient response to mental health care (MHC), thereby enabling treatment personalization. However, previous efforts have been limited to populations living in predominantly higher income, developed countries. This study aimed to extend the reach of precision MHC systems by developing and testing a feasible and readily implementable algorithm for identifying patients at risk of nonresponse to routinely delivered psychotherapy in Chile, a developing country in Latin America. Data were derived from a community-based, randomized trial that tested the effects of progress feedback on naturalistically delivered psychotherapy outcome. Patients were 547 adults who were consecutively admitted to an outpatient clinic in Santiago, Chile. Treatment response was defined using norms for reliable improvement on the Outcome Questionnaire-30. Based on 10 sociodemographic and seven clinical predictors, we trained elastic net and random forest algorithms on a randomly selected training set (70%; n = 384). The best performing algorithm was tested on a hold-out sample (30%; n = 163). Reliable improvement was achieved in 42% of the cases. A random forest algorithm demonstrated moderate performance in the hold-out sample (area under the curve = .74, Brier score = .21), correctly identifying 73% of the patients who did not respond. This study developed a predictive algorithm that demonstrated moderate accuracy in identifying patients at risk of nonresponse to naturalistic psychotherapy in Chile, using routinely assessed and easy-to-collect sociodemographic and clinical information. Using such tools may represent one step toward reducing the multilayered outcome disparities faced by individuals receiving MHC in socioeconomically disadvantaged contexts. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Novel Uncertainty Quantification Through Perturbation-Assisted Sample Synthesis.

This paper introduces a novel Perturbation-Assisted Inference (PAI) framework utilizing synthetic data generated by the Perturbation-Assisted Sample Synthesis (PASS) method. The framework focuses on uncertainty quantification in complex data scenarios, particularly involving unstructured data while utilizing deep learning models. On one hand, PASS employs a generative model to create synthetic data that closely mirrors raw data while preserving its rank properties through data perturbation, thereby enhancing data diversity and bolstering privacy. By incorporating knowledge transfer from large pre-trained generative models, PASS enhances estimation accuracy, yielding refined distributional estimates of various statistics via Monte Carlo experiments. On the other hand, PAI boasts its statistically guaranteed validity. In pivotal inference, it enables precise conclusions even without prior knowledge of the pivotal's distribution. In non-pivotal situations, we enhance the reliability of synthetic data generation by training it with an independent holdout sample. We demonstrate the effectiveness of PAI in advancing uncertainty quantification in complex, data-driven tasks by applying it to diverse areas such as image synthesis, sentiment word analysis, multimodal inference, and the construction of prediction intervals.

Analyzing and predicting short-term substance use behaviors of persons who use drugs in the great plains of the U.S.

Substance use induces large economic and societal costs in the U.S. Understanding the change in substance use behaviors of persons who use drugs (PWUDs) over time, therefore, is important in order to inform healthcare providers, policymakers, and other stakeholders toward more efficient allocation of limited resources to at-risk PWUDs. This study examines the short-term (within a year) behavioral changes in substance use of PWUDs at the population and individual levels. 237 PWUDs in the Great Plains of the U.S. were recruited by our team. The sample provides us longitudinal survey data regarding their individual attributes, including drug use behaviors, at two separate time periods spanning 4-12 months. At the population level, we analyze our data quantitatively for 18 illicit drugs; then, at the individual level, we build interpretable machine learning logistic regression and decision tree models for identifying relevant attributes to predict, for a given PWUD, (i) which drug(s) they would likely use and (ii) which drug(s) they would likely increase usage within the next 12 months. All predictive models were evaluated by computing the (averaged) Area under the Receiver Operating Characteristic curve (AUROC) and Area under the Precision-Recall curve (AUPR) on multiple distinct sets of hold-out sample. At the population level, the extent of usage change and the number of drugs exhibiting usage changes follow power-law distributions. At the individual level, AUROC's of the models for the top-4 prevalent drugs (marijuana, methamphetamines, amphetamines, and cocaine) range 0.756-0.829 (+2.88-7.66% improvement with respect to baseline models using only current usage of the respective drugs as input) for (i) and 0.670-0.765 (+4.34-18.0%) for (ii). The corresponding AUPR's of the said models range 0.729-0.947 (+2.49-13.6%) for (i) and 0.348-0.618 (+26.9-87.6%) for (ii). The observed qualitative changes in short-term substance usage and the trained predictive models for (i) and (ii) can potentially inform human decision-making toward efficient allocation of appropriate resources to PWUDs at highest risk.

Advances in Corneal Diagnostics Using Machine Learning

This paper provides comprehensive insights into the cornea and its diseases, with a particular focus on keratoconus. This paper explores the cornea’s function in maintaining ocular health, detailing its anatomy, pathological conditions, and the latest developments in diagnostic techniques. Keratoconus is discussed extensively, covering its subtypes, etiology, clinical manifestations, and the application of the Q-value for quantification. Several diagnostic techniques, such as corneal topography, are crucial points of discussion. This paper also examines the use of machine learning models, specifically Decision Tree and Nearest Neighbor Analysis, which enhance the accuracy of diagnosing based on topographical corneal parameters from corneal topography. These models provide valuable insights into disease progression and aid in clinical decision making. Integrating these technologies in medical research opens promising avenues for enhanced disease detection. Our findings demonstrate the effectiveness of Decision Tree and Nearest Neighbor Analysis in classifying and predicting conditions based on corneal parameters. The Decision Tree achieved classification accuracy of 62% for training and 65.7% for testing, while Nearest Neighbor Analysis yielded 65.4% for training and 62.6% for holdout samples. These models offer valuable insights into the progression and severity of keratoconus, aiding clinicians in treatment and management decisions.

Application of artificial intelligence in the Prediction of Complications in patients with Malaria

Introduction: This study aims to develop a neural network (NN) that can serve as a useful tool for early diagnosis of complicated malaria. Materials and methods: In this study, a feedforward NN was developed, incorporating 10 clinical variables in the input nodes, hidden layer, and output node. The data were included in the input layer. Various validation techniques such as V cross, Random V cross, Modified Holdout, and Proportional Percentage Sample were applied to train and validate the network using data from 412 patients. Results: The variables included in the analysis were mean arterial pressure, hemoglobin, leukocyte count, platelet count, total bilirubin, presence of dyspnea, vomiting, previous history of malaria, prior use of malaria medication, and persistent fever. The V cross technique, Random V cross Validation, Modified Holdout Validation, and Proportional Percentage Sample Validation were utilized to evaluate the performance of a NN in diagnosing malaria. Sensitivity values varied from 13% to 47%, with positive predictive value values ranging from 37% to 88%. Specificity remained consistently high, ranging from 79% to 90%. Discussion: Sensitivity, specificity, and positive predictive values varied across techniques: V cross and random V cross validation showed narrower sensitivity ranges with strong specificities, while modified holdout validation exhibited wider sensitivity variability.

Abstract 4119992: 12-lead electrocardiograms predict adverse cardiovascular outcomes of emergency department patients

Introduction: Chest pain is a common cause for presentation to the emergency department (ED). Identifying patients at increased risk of adverse cardiovascular outcomes may help guide testing. Aims: We sought to determine if deep learning enabled 12-lead electrocardiogram (ECG) analysis may improve prediction of adverse cardiovascular outcomes of ED patients presenting with chest pain compared to conventional approaches. Methods: A pretrained convolutional neural network was applied to derive numerical representations of the index ECG waveforms from patients presenting to the Brigham and Women’s Hospital (BWH) ED with chest pain. We trained a neural network model (‘CP-AI’) to predict 30-day major adverse cardiovascular events (MACE) defined as myocardial infarction (MI), pulmonary embolism (PE), aortic dissection (AD), or mortality using age, sex, cardiac biomarkers (troponin and d-dimer), and the numerical ECG representations as inputs ( Figure 1 ). In a holdout sample of Massachusetts General Hospital (MGH) ED patients, we compared the performance of the CP-AI model to alternative models fit on (1) demographics (‘Age and Sex’) (2) demographics + cardiac biomarkers (‘Biomarker Model’), and (3) demographics + numerical ECG representations (‘ECG Model’). Results: The BWH derivation sample included 18,811 individuals (51% female, mean age 65 ± 16 years). The MGH validation sample included 14,476 individuals (45% female, mean age 65 ± 15 years). The CP-AI model significantly outperformed the comparison models for prediction of 30-day MACE with an area under the receiver operating characteristic curve (AUROC) of 0.82 [95% CI 0.81-0.83] ( Figure 1 ). CP-AI also significantly outperformed the comparison models for classification of the 30-day MACE components including MI (AUROC 0.91 [95% CI 0.90-0.92]), PE (AUROC 0.74 [95% CI 0.72-0.76]), and mortality (AUROC 0.86 [95% CI 0.85-0.87]), but not aortic dissection (0.71 [95% CI 0.68-0.75]). Conclusions: The integration of deep learning ECG analysis with conventional assessment improved prediction of adverse cardiovascular of ED patients presenting with chest pain.

Predicting maintenance lithium response for bipolar disorder from electronic health records-a retrospective study.

Optimising maintenance drug treatment selection for people with bipolar disorder is challenging. There is some evidence that clinical and demographic features may predict response to lithium. However, attempts to personalise treatment choice have been limited. We aimed to determine if machine learning methods applied to electronic health records could predict differential response to lithium or olanzapine. From electronic United Kingdom primary care records, we extracted a cohort of individuals prescribed either lithium (19,106 individuals) or olanzapine (12,412) monotherapy. Machine learning models were used to predict successful monotherapy maintenance treatment, using 113 clinical and demographic variables, 8,017 (41.96%) lithium responders and 3,831 (30.87%) olanzapine responders. We found a quantitative structural difference in that lithium maintenance responders were weakly predictable in our holdout sample, consisting of the 5% of patients with the most recent exposure. Age at first diagnosis, age at first treatment and the time between these were the most important variables in all models. Even if we failed to predict successful monotherapy olanzapine treatment, and so to definitively separate lithium vs. olanzapine responders, the characterization of the two groups may be used for classification by proxy. This can, in turn, be useful for establishing maintenance therapy. The further exploration of machine learning methods on EHR data for drug treatment selection could in the future play a role for clinical decision support. Signals in the data encourage further experiments with larger datasets to definitively separate lithium vs. olanzapine responders.

Machine learning and Bayesian network analyses identifies associations with insomnia in a national sample of 31,285 treatment-seeking college students

BackgroundA better understanding of the relationships between insomnia and anxiety, mood, eating, and alcohol-use disorders is needed given its prevalence among young adults. Supervised machine learning provides the ability to evaluate which mental disorder is most associated with heightened insomnia among U.S. college students. Combined with Bayesian network analysis, probable directional relationships between insomnia and interacting symptoms may be illuminated.MethodsThe current exploratory analyses utilized a national sample of college students across 26 U.S. colleges and universities collected during population-level screening before entering a randomized controlled trial. We used a 4-step statistical approach: (1) at the disorder level, an elastic net regularization model examined the relative importance of the association between insomnia and 7 mental disorders (major depressive disorder, generalized anxiety disorder, social anxiety disorder, panic disorder, post-traumatic stress disorder, anorexia nervosa, and alcohol use disorder); (2) This model was evaluated within a hold-out sample. (3) at the symptom level, a completed partially directed acyclic graph (CPDAG) was computed via a Bayesian hill-climbing algorithm to estimate potential directionality among insomnia and its most associated disorder [based on SHAP (SHapley Additive exPlanations) values)]; (4) the CPDAG was then tested for generalizability by assessing (in)equality within a hold-out sample using structural hamming distance (SHD).ResultsOf 31,285 participants, 20,597 were women (65.8%); mean (standard deviation) age was 22.96 (4.52) years. The elastic net model demonstrated clinical significance in predicting insomnia severity in the training sample [R2 = .44 (.01); RMSE = 5.00 (0.08)], with comparable performance in the hold-out sample (R2 = .33; RMSE = 5.47). SHAP values indicated that the presence of any mental disorder was associated with higher insomnia scores, with major depressive disorder as the most important disorder associated with heightened insomnia (mean |SHAP|= 3.18). The training CPDAG and hold-out CPDAG (SHD = 7) suggested depression symptoms presupposed insomnia with depressed mood, fatigue, and self-esteem as key parent nodes.ConclusionThese findings provide insights into the associations between insomnia and mental disorders among college students and warrant further investigation into the potential direction of causality between insomnia and depression.Trial registrationTrial was registered on the National Institute of Health RePORTER website (R01MH115128 || 23/08/2018).

Exploring the Relationship Between COVID-19 Transmission and Population Mobility over Time

This study explores the dynamic relationship between COVID-19 transmission and transportation mobility, with an emphasis on understanding the time-varying bidirectional interplay across the different phases of the pandemic. To gain insight into this relationship, we analyzed county-level data on transmission and mobility patterns from the United States over a 74-week period using a comprehensive list of factors including: temporal factors, socio-demographics, health indicators, health care infrastructure attributes, and spatial factors. For our analysis, we proposed a simultaneous econometric model system that explicitly accounts for the bidirectional relationship between COVID-19 transmission and mobility patterns while also accounting for the influence of common unobserved factors on the two variables. The model results strongly support our hypothesis that COVID-19 transmission and mobility patterns are interconnected. Further, our findings show distinct phases of the bidirectional relationship influenced by behavior changes, vaccine availability, and the emergence of new variants. Additionally, we conducted a validation exercise on a hold-out sample to assess the robustness of our model. The results confirm the superiority of the simultaneous model system with enhanced interpretability and prediction capability. By analyzing data from several weeks for the COVID-19 pandemic, our study provides valuable insights into the evolving dynamics and potential strategies for future pandemics.

Updating and calibrating the Real-World Progression In Diabetes (RAPIDS) model in a non-Veterans Affairs population.

To present the Real-World Progression In Diabetes (RAPIDS) 2.0 Risk Engine, the only simulation model to study the long-term trajectories of outcomes arising from dynamic sequences of glucose-lowering treatments in type 2 diabetes (T2DM). The RAPIDS model's risk equations were re-estimated using a Least Absolute Shrinkage and Selection Operator (LASSO)-based regularization of features that spanned baseline data from the last two quarters of current time and interactions with age. These equations were supplemented with estimates for the impact of dipeptidyl peptidase-4 inhibitors, glucagon-like peptide-1 receptor agonists, and sodium-glucose cotransporter-2 inhibitor classes of drugs as monotherapies and their combinations with metformin based on newer trial data and comprehensive meta-analyses. The probabilistic RAPIDS 2.0 model was calibrated (N = 25 000) and validated (N = 263 816) using electronic medical records (EMR) data between 2008 and 2021 from a national network of US healthcare organizations. The EMR-based cohort had a mean age of 61 years at baseline, with 50% women, 70% non-Hispanic White individuals and 20% non-Hispanic Black individuals, and was followed for 17.5 quarters (range: 3-50). The final RAPIDS 2.0 risk engine accurately predicted the long-term trajectories of all nine biomarkers and nine outcomes in the hold-out validation sample. Similar accuracies in predictions were observed in each of the 14 subgroups studied. The RAPIDS 2.0 model demonstrated valid long-term predictions of outcomes in individuals with T2DM in the United States as a function of dynamic sequences of treatment use patterns. This highlights its potential to project long-term comparative effectiveness between alternative sequences of glucose-lowering treatment uses in the United States.

Mental issues, internet addiction and quality of life predict burnout among Hungarian teachers: a machine learning analysis

BackgroundBurnout is usually defined as a state of emotional, physical, and mental exhaustion that affects people in various professions (e.g. physicians, nurses, teachers). The consequences of burnout involve decreased motivation, productivity, and overall diminished well-being. The machine learning-based prediction of burnout has therefore become the focus of recent research. In this study, the aim was to detect burnout using machine learning and to identify its most important predictors in a sample of Hungarian high-school teachers.MethodsThe final sample consisted of 1,576 high-school teachers (522 male), who completed a survey including various sociodemographic and health-related questions and psychological questionnaires. Specifically, depression, insomnia, internet habits (e.g. when and why one uses the internet) and problematic internet usage were among the most important predictors tested in this study. Supervised classification algorithms were trained to detect burnout assessed by two well-known burnout questionnaires. Feature selection was conducted using recursive feature elimination. Hyperparameters were tuned via grid search with 10-fold cross-validation. Due to class imbalance, class weights (i.e. cost-sensitive learning), downsampling and a hybrid method (SMOTE-ENN) were applied in separate analyses. The final model evaluation was carried out on a previously unseen holdout test sample.ResultsBurnout was detected in 19.7% of the teachers included in the final dataset. The best predictive performance on the holdout test sample was achieved by random forest with class weigths (AUC = 0.811; balanced accuracy = 0.745, sensitivity = 0.765; specificity = 0.726). The best predictors of burnout were Beck’s Depression Inventory scores, Athen’s Insomnia Scale scores, subscales of the Problematic Internet Use Questionnaire and self-reported current health status.ConclusionsThe performances of the algorithms were comparable with previous studies; however, it is important to note that we tested our models on previously unseen holdout samples suggesting higher levels of generalizability. Another remarkable finding is that besides depression and insomnia, other variables such as problematic internet use and time spent online also turned out to be important predictors of burnout.

Analysis of team success based on technical match-play performance in the Australian Football League Women’s (AFLW) competition

ABSTRACT An understanding of the effect contextual data may have on key match-play technical performance indicators in the Australian Football League Women’s (AFLW) competition is warranted due to its rapid evolution. To address this, predictive models were fit to determine which technical match-play data, including new contextual information, more accurately predict AFLW match outcomes (win/loss, margin), and what are the most important contexts and technical predictors of team performance? Thirteen random forest models were fit, each with greater data contextual interaction including relative to opposition and harder-to-attain match-play variables, field location, and individual player contributions. Models were assessed by prediction performance on match outcome in a holdout sample and variable importance through Mean Decrease in Gini Index. Effective kicks and entries into attacking locations were important in models. Territory gained, contexts of relative performance to the opposition, and locational information around actions improved prediction. This methodology represents the most in-depth analysis of women’s Australian football technical match-play performance to date. Commentary presented surrounded issues of using aggregated datasets, prediction with match-play success as a dependent variable, and that detailed, process-oriented approaches are needed in future to avoid large assumptions.

Using natural language processing to facilitate the harmonisation of mental health questionnaires: a validation study using real-world data

BackgroundPooling data from different sources will advance mental health research by providing larger sample sizes and allowing cross-study comparisons; however, the heterogeneity in how variables are measured across studies poses a challenge to this process.MethodsThis study explored the potential of using natural language processing (NLP) to harmonise different mental health questionnaires by matching individual questions based on their semantic content. Using the Sentence-BERT model, we calculated the semantic similarity (cosine index) between 741 pairs of questions from five questionnaires. Drawing on data from a representative UK sample of adults (N = 2,058), we calculated a Spearman rank correlation for each of the same pairs of items, and then estimated the correlation between the cosine values and Spearman coefficients. We also used network analysis to explore the model’s ability to uncover structures within the data and metadata.ResultsWe found a moderate overall correlation (r = .48, p < .001) between the two indices. In a holdout sample, the cosine scores predicted the real-world correlations with a small degree of error (MAE = 0.05, MedAE = 0.04, RMSE = 0.064) suggesting the utility of NLP in identifying similar items for cross-study data pooling. Our NLP model could detect more complex patterns in our data, however it required manual rules to decide which edges to include in the network.ConclusionsThis research shows that it is possible to quantify the semantic similarity between pairs of questionnaire items from their meta-data, and these similarity indices correlate with how participants would answer the same two items. This highlights the potential of NLP to facilitate cross-study data pooling in mental health research. Nevertheless, researchers are cautioned to verify the psychometric equivalence of matched items.

The association of lumbar intervertebral disc degeneration with low back pain is modified by underlying genetic propensity to pain

BACKGROUND CONTEXTAssociations between magnetic resonance imaging (MRI)-detected lumbar intervertebral disc degeneration (LDD) and LBP are often of modest magnitude. This association may be larger in specific patient subgroups. PURPOSETo examine whether the association between LDD and LBP is modified by underlying genetic predispositions to pain. STUDY DESIGNCross-sectional study in UK Biobank (UKB) and Twins UK. PATIENT SAMPLESA genome-wide association study (GWAS) of the number of anatomical chronic pain locations was conducted in 347,538 UKB participants. The GWAS was used to develop a genome-wide polygenic risk score (PRS) in a holdout sample of 30,000 UKB participants. The PRS model was then used in analyses of 645 TwinsUK participants with standardized LDD MRI assessments. OUTCOME MEASURESEver having had LBP associated with disability lasting ≥1 month (LBP1). METHODSUsing the PRS as a proxy for “genetically-predicted propensity to pain”, we stratified TwinsUK participants into PRS quartiles. A “basic” model examined the association between an LDD summary score (LSUM) and LBP1, adjusting for covariates. A “fully-adjusted” model also adjusted for PRS quartile and LSUM x PRS quartile interaction terms. RESULTSIn the basic model, the odds ratio (OR) of LBP1 was 1.8 per standard deviation of LSUM (95% confidence interval [CI] 1.4–2.3). In the fully-adjusted model, there was a statistically significant LSUM-LBP1 association in quartile 4, the highest PRS quartile (OR=2.5 [95% CI 1.7–3.7], p=2.6×10−6), and in quartile 3 (OR=2.0, [95% CI 1.3–3.0]; p=.002), with small-magnitude and/or nonsignificant associations in the lowest 2 PRS quartiles. PRS quartile was a significant effect modifier of the LSUM-LBP1 association (interaction p≤.05). CONCLUSIONSGenetically-predicted propensity to pain modifies the LDD-LBP association, with the strongest association present in people with the highest genetic propensity to pain. Lumbar MRI findings may have stronger connections to LBP in specific subgroups of people.

Assessing calibration and bias of a deployed machine learning malnutrition prediction model within a large healthcare system

Malnutrition is a frequently underdiagnosed condition leading to increased morbidity, mortality, and healthcare costs. The Mount Sinai Health System (MSHS) deployed a machine learning model (MUST-Plus) to detect malnutrition upon hospital admission. However, in diverse patient groups, a poorly calibrated model may lead to misdiagnosis, exacerbating health care disparities. We explored the model’s calibration across different variables and methods to improve calibration. Data from adult patients admitted to five MSHS hospitals from January 1, 2021 - December 31, 2022, were analyzed. We compared MUST-Plus prediction to the registered dietitian’s formal assessment. Hierarchical calibration was assessed and compared between the recalibration sample (N = 49,562) of patients admitted between January 1, 2021 - December 31, 2022, and the hold-out sample (N = 17,278) of patients admitted between January 1, 2023 - September 30, 2023. Statistical differences in calibration metrics were tested using bootstrapping with replacement. Before recalibration, the overall model calibration intercept was −1.17 (95% CI: −1.20, −1.14), slope was 1.37 (95% CI: 1.34, 1.40), and Brier score was 0.26 (95% CI: 0.25, 0.26). Both weak and moderate measures of calibration were significantly different between White and Black patients and between male and female patients. Logistic recalibration significantly improved calibration of the model across race and gender in the hold-out sample. The original MUST-Plus model showed significant differences in calibration between White vs. Black patients. It also overestimated malnutrition in females compared to males. Logistic recalibration effectively reduced miscalibration across all patient subgroups. Continual monitoring and timely recalibration can improve model accuracy.

Automated free speech analysis reveals distinct markers of Alzheimer’s and frontotemporal dementia

Dementia can disrupt how people experience and describe events as well as their own role in them. Alzheimer’s disease (AD) compromises the processing of entities expressed by nouns, while behavioral variant frontotemporal dementia (bvFTD) entails a depersonalized perspective with increased third-person references. Yet, no study has examined whether these patterns can be captured in connected speech via natural language processing tools. To tackle such gaps, we asked 96 participants (32 AD patients, 32 bvFTD patients, 32 healthy controls) to narrate a typical day of their lives and calculated the proportion of nouns, verbs, and first- or third-person markers (via part-of-speech and morphological tagging). We also extracted objective properties (frequency, phonological neighborhood, length, semantic variability) from each content word. In our main study (with 21 AD patients, 21 bvFTD patients, and 21 healthy controls), we used inferential statistics and machine learning for group-level and subject-level discrimination. The above linguistic features were correlated with patients’ scores in tests of general cognitive status and executive functions. We found that, compared with HCs, (i) AD (but not bvFTD) patients produced significantly fewer nouns, (ii) bvFTD (but not AD) patients used significantly more third-person markers, and (iii) both patient groups produced more frequent words. Machine learning analyses showed that these features identified individuals with AD and bvFTD (AUC = 0.71). A generalizability test, with a model trained on the entire main study sample and tested on hold-out samples (11 AD patients, 11 bvFTD patients, 11 healthy controls), showed even better performance, with AUCs of 0.76 and 0.83 for AD and bvFTD, respectively. No linguistic feature was significantly correlated with cognitive test scores in either patient group. These results suggest that specific cognitive traits of each disorder can be captured automatically in connected speech, favoring interpretability for enhanced syndrome characterization, diagnosis, and monitoring.

Resolving spatial response heterogeneity in glioblastoma.

Spatial intratumoral heterogeneity poses a significant challenge for accurate response assessment in glioblastoma. Multimodal imaging coupled with advanced image analysis has the potential to unravel this response heterogeneity. Based on automated tumor segmentation and longitudinal registration with follow-up imaging, we categorized contrast-enhancing voxels of 61 patients with suspected recurrence of glioblastoma into either true tumor progression (TP) or pseudoprogression (PsP). To allow the unbiased analysis of semantically related image regions, adjacent voxels with similar values of cerebral blood volume (CBV), FET-PET, and contrast-enhanced T1w were automatically grouped into supervoxels. We then extracted first-order statistics as well as texture features from each supervoxel. With these features, a Random Forest classifier was trained and validated employing a 10-fold cross-validation scheme. For model evaluation, the area under the receiver operating curve, as well as classification performance metrics were calculated. Our image analysis pipeline enabled reliable spatial assessment of tumor response. The predictive model reached an accuracy of 80.0% and a macro-weighted AUC of 0.875, which takes class imbalance into account, in the hold-out samples from cross-validation on supervoxel level. Analysis of feature importances confirmed the significant role of FET-PET-derived features. Accordingly, TP- and PsP-labeled supervoxels differed significantly in their 10th and 90th percentile, as well as the median of tumor-to-background normalized FET-PET. However, CBV- and T1c-related features also relevantly contributed to the model's performance. Disentangling the intratumoral heterogeneity in glioblastoma holds immense promise for advancing precise local response evaluation and thereby also informing more personalized and localized treatment strategies in the future.

Prediction of Deleterious Single Amino Acid Polymorphisms with a Consensus Holdout Sampler.

Single Amino Acid Polymorphisms (SAPs) or nonsynonymous Single Nucleotide Variants (nsSNVs) are the most common genetic variations. They result from missense mutations where a single base pair substitution changes the genetic code in such a way that the triplet of bases (codon) at a given position is coding a different amino acid. Since genetic mutations sometimes cause genetic diseases, it is important to comprehend and foresee which variations are harmful and which ones are neutral (not causing changes in the phenotype). This can be posed as a classification problem. Computational methods using machine intelligence are gradually replacing repetitive and exceedingly overpriced mutagenic tests. By and large, uneven quality, deficiencies, and irregularities of nsSNVs datasets debase the convenience of artificial intelligence-based methods. Subsequently, strong and more exact approaches are needed to address these problems. In the present work paper, we show a consensus classifier built on the holdout sampler, which appears strong and precise and outflanks all other popular methods. We produced 100 holdouts to test the structures and diverse classification variables of diverse classifiers during the training phase. The finest performing holdouts were chosen to develop a consensus classifier and tested using a k-fold (1 ≤ k ≤5) cross-validation method. We also examined which protein properties have the biggest impact on the precise prediction of the effects of nsSNVs. Our Consensus Holdout Sampler outflanks other popular algorithms, and gives excellent results, highly accurate with low standard deviation. The advantage of our method emerges from using a tree of holdouts, where diverse LM/AI-based programs are sampled in diverse ways.

Fusing Repeated Cross-Sectional Revealed Preference Datasets based on Rational Inattention Theory: Accounting for Changing Modal Preferences

To address the methodological limitation of cross-sectional studies and the data constraints of longitudinal/panel studies, this paper presents a model-based method to fuse repeated cross-sectional travel survey data based on the theory of rational inattention (RI) in discrete choice modeling. In the proposed framework, older cross-sectional data are used to model the prior probability of choice alternatives, and more recent cross-sectional data are used to capture conditional heterogeneous choices. The fusion method is theoretically more robust and computationally less burdensome than existing data pooling techniques. The method is empirically tested using data from two cycles of a large-sample post-secondary student travel survey in the Greater Toronto and Hamilton Area to investigate the commuting mode choices of post-secondary students. Parameter estimates of the RI-based multinomial logit (MNL) model indicate that the proposed method can generate behaviorally consistent results. Validation of the estimated model using a holdout sample indicates its improved forecasting performance compared with the classical random utility maximizing MNL model. The fusion method can be extended to more than two cycles of repeated cross-sectional data by updating the prior probabilities whenever new cross-sectional data become available. Thus, the study presents a continuous framework for fusing information from multiple time points using repeated cross-sectional datasets to capture preference evolution better and enhance the forecasting robustness of discrete choice models.

Assessing the reliability and accuracy of sex estimation models utilizing sternal morphometry derived from computed tomography in the Ghanaian population

Sex estimation models are specific to populations and cannot be generalized due to genetic and environmental variabilities. This retrospective cross-sectional study, conducted between January and September 2023 at the Tamale Teaching Hospital, included 119 (50.9 %) females and 115 (49.1 %) males aged 23–82 years. Measurements, including manubrium length (M), manubrium width (MW), sternal body length (B), combined manubrium and sternal body lengths (CL), corpus sterni width at first sternebrae (CSWS1), and corpus sterni width at third sternebrae (CSWS3), were obtained from Computerized Tomographic (CT) images of the sternum using DICOM Viewer, accurate to 0.1 cm. Subsequently, sternal area (SA) and sternal index (SI) were calculated. Univariable and stepwise multivariable discriminant function analysis (DFA) and logistic regression (LR) models were developed using a training sample (70 %), and cross-validation was performed on a holdout sample (30 %). Results showed that the linear measurements, excluding M, and sternal area were higher in males than females (P<0.001), while the sternal index was <50 % in males but >50 % in females (P<0.001). Univariable sex estimation accuracies, in cross-validation, ranged from 43.7 % to 92.9 % in DFA and 50.0–92.9 % in LR. For multivariable models, the accuracy ranges were 92.9–94.3 % in DFA and 91.6–93.0 % in LR. The sternal body length was the most accurate at 90.1 % in DFA and 90.2 % in LR, with lower sex bias (male-female) in LR than in DFA (-0.1 vs. 8.7). The sternum proves valuable for sex estimation, with sternal body length as the most accurate linear measurement. However, multivariable models, particularly LR, demonstrate higher accuracy compared to DFA.