Dataset Information

Using machine-learning methods to identify early-life predictors of 11-year language outcome.

ABSTRACT:

Background

Language is foundational for neurodevelopment and quality of life, but an estimated 10% of children have a language disorder at age 5. Many children shift between classifications of typical and low language if assessed at multiple times in the early years, making it difficult to identify which children will have persisting difficulties and benefit most from support. This study aims to identify a parsimonious set of preschool indicators that predict language outcomes in late childhood, using data from the population-based Early Language in Victoria Study (n = 839).

Methods

Parents completed surveys about their children at ages 8, 12, 24, and 36 months. At 11 years, children were assessed using the Clinical Evaluation of Language Fundamentals 4th Edition (CELF-4). We used random forests to identify which of the 1990 parent-reported questions best predict children's 11-year language outcome (CELF-4 score ≤81 representing low language) and used SuperLearner to estimate the accuracy of the constrained sets of questions.

Results

At 24 months, seven predictors relating to vocabulary, symbolic play, pragmatics and behavior yielded 73% sensitivity (95% CI: 57, 85) and 77% specificity (95% CI: 74, 80) for predicting low language at 11 years. [Corrections made on 5 May 2023, after first online publication: In the preceding sentence 'motor skills' has been corrected to 'behavior' in this version.] At 36 months, 7 predictors relating to morphosyntax, vocabulary, parent-child interactions, and parental stress yielded 75% sensitivity (95% CI: 58, 88) and 85% specificity (95% CI: 81, 87). Measures at 8 and 12 months yielded unsatisfactory accuracy.

Conclusions

We identified two short sets of questions that predict language outcomes at age 11 with fair accuracy. Future research should seek to replicate results in a separate cohort.

SUBMITTER: Gasparini L

PROVIDER: S-EPMC10952842 | biostudies-literature | 2023 Aug

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Using machine-learning methods to identify early-life predictors of 11-year language outcome.

Gasparini Loretta L Shepherd Daisy A DA Bavin Edith L EL Eadie Patricia P Eadie Patricia P Reilly Sheena S Morgan Angela T AT Wake Melissa M

Journal of child psychology and psychiatry, and allied disciplines 20221207 8

<h4>Background</h4>Language is foundational for neurodevelopment and quality of life, but an estimated 10% of children have a language disorder at age 5. Many children shift between classifications of typical and low language if assessed at multiple times in the early years, making it difficult to identify which children will have persisting difficulties and benefit most from support. This study aims to identify a parsimonious set of preschool indicators that predict language outcomes in late ch ...[more]

PMID: 36478310

Similar Datasets

Project description:Objective: Congenital heart diseases (CHDs) are associated with an extremely heavy global disease burden as the most common category of birth defects. Genetic and environmental factors have been identified as risk factors of CHDs previously. However, high volume clinical indicators have never been considered when predicting CHDs. This study aimed to predict the occurrence of CHDs by considering thousands of variables from self-reported questionnaires and routinely collected clinical laboratory data using machine learning algorithms. Methods: We conducted a birth cohort study at one of the largest cardiac centers in China from 2011 to 2017. All fetuses were screened for CHDs using ultrasound and cases were confirmed by at least two pediatric cardiologists using echocardiogram. A total of 1,127 potential predictors were included to predict CHDs. We used the Explainable Boosting Machine (EBM) for prediction and evaluated the model performance using area under the Receive Operating Characteristics (ROC) curves (AUC). The top predictors were selected according to their contributions and predictive values. Thresholds were calculated for the most significant predictors. Results: Overall, 5,390 mother-child pairs were recruited. Our prediction model achieved an AUC of 76% (69-83%) from out-of-sample predictions. Among the top 35 predictors of CHDs we identified, 34 were from clinical laboratory tests and only one was from the questionnaire (abortion history). Total accuracy, sensitivity, and specificity were 0.65, 0.74, and 0.65, respectively. Maternal serum uric acid (UA), glucose, and coagulation levels were the most consistent and significant predictors of CHDs. According to the thresholds of the predictors identified in our study, which did not reach the current clinical diagnosis criteria, elevated UA (>4.38 mg/dl), shortened activated partial thromboplastin time (<33.33 s), and elevated glucose levels were the most important predictors and were associated with ranges of 1.17-1.54 relative risks of CHDs. We have developed an online predictive tool for CHDs based on our findings that may help screening and prevention of CHDs. Conclusions: Maternal UA, glucose, and coagulation levels were the most consistent and significant predictors of CHDs. Thresholds below the current clinical definition of "abnormal" for these predictors could be used to help develop CHD screening and prevention strategies.

Project description:Accurate, automated extraction of clinical stroke information from unstructured text has several important applications. ICD-9/10 codes can misclassify ischemic stroke events and do not distinguish acuity or location. Expeditious, accurate data extraction could provide considerable improvement in identifying stroke in large datasets, triaging critical clinical reports, and quality improvement efforts. In this study, we developed and report a comprehensive framework studying the performance of simple and complex stroke-specific Natural Language Processing (NLP) and Machine Learning (ML) methods to determine presence, location, and acuity of ischemic stroke from radiographic text. We collected 60,564 Computed Tomography and Magnetic Resonance Imaging Radiology reports from 17,864 patients from two large academic medical centers. We used standard techniques to featurize unstructured text and developed neurovascular specific word GloVe embeddings. We trained various binary classification algorithms to identify stroke presence, location, and acuity using 75% of 1,359 expert-labeled reports. We validated our methods internally on the remaining 25% of reports and externally on 500 radiology reports from an entirely separate academic institution. In our internal population, GloVe word embeddings paired with deep learning (Recurrent Neural Networks) had the best discrimination of all methods for our three tasks (AUCs of 0.96, 0.98, 0.93 respectively). Simpler NLP approaches (Bag of Words) performed best with interpretable algorithms (Logistic Regression) for identifying ischemic stroke (AUC of 0.95), MCA location (AUC 0.96), and acuity (AUC of 0.90). Similarly, GloVe and Recurrent Neural Networks (AUC 0.92, 0.89, 0.93) generalized better in our external test set than BOW and Logistic Regression for stroke presence, location and acuity, respectively (AUC 0.89, 0.86, 0.80). Our study demonstrates a comprehensive assessment of NLP techniques for unstructured radiographic text. Our findings are suggestive that NLP/ML methods can be used to discriminate stroke features from large data cohorts for both clinical and research-related investigations.

Project description:Background: Recent evidence suggests that integration of multi-modal data improves performance in machine learning prediction of depression treatment outcomes. Here, we compared the predictive performance of three machine learning classifiers using differing combinations of sociodemographic characteristics, baseline clinical self-reports, cognitive tests, and structural magnetic resonance imaging (MRI) features to predict treatment outcomes in late-life depression (LLD). Methods: Data were combined from two clinical trials conducted with depressed adults aged 60 and older, including response to escitalopram (N = 32, NCT01902004) and Tai Chi (N = 35, NCT02460666). Remission was defined as a score of 6 or less on the 24-item Hamilton Rating Scale for Depression (HAMD) at the end of 24 weeks of treatment. Features subsets were constructed from baseline sociodemographic and clinical features, gray matter volumes (GMVs), or both. Three classification algorithms were compared: (1) Support Vector Machine-Radial Bias Function (SVMRBF), (2) Random Forest (RF), and (3) Logistic Regression (LR). A repeated 5-fold cross-validation approach with a wrapper-based feature selection method was used for model fitting. Model performance metrics included Area under the ROC Curve (AUC) and Matthews correlation coefficient (MCC). Cross-validated performance significance was tested by permutation analysis. Classifiers were compared by Cochran's Q and post-hoc pairwise comparisons using McNemar's Chi-Square test with Bonferroni correction. Results: For the RF and SVMRBF algorithms, the combined feature set outperformed the clinical and GMV feature sets with a final cross-validated AUC of 0.83 ± 0.11 and 0.80 ± 0.11, respectively. Both classifiers passed permutation analysis. The LR algorithm performed best using GMV features alone (AUC 0.79 ± 0.14) but failed to pass permutation analysis using any feature set. Performance of the three classifiers differed significantly for all three features sets. Important predictive features of treatment response included anterior and posterior cingulate volumes, depression characteristics, and self-reported health-related quality scores. Conclusion: This preliminary exploration into the use of ML and multi-modal data to identify predictors of general treatment response in LLD indicates that integration of clinical and structural MRI features significantly increases predictive capability. Identified features are among those previously implicated in geriatric depression, encouraging future work in this arena.

Project description:BackgroundEarly illness course correlates with long-term outcome in psychosis. Accurate prediction could allow more focused intervention. Earlier intervention corresponds to significantly better symptomatic and functional outcomes. Our study objective is to use routinely collected baseline demographic and clinical characteristics to predict employment, education or training (EET) status, and symptom remission in patients with first episode psychosis (FEP) at one-year.Methods and findings83 FEP patients were recruited from National Health Service (NHS) Glasgow between 2011 and 2014 to a 24-month prospective cohort study with regular assessment of demographic and psychometric measures. An external independent cohort of 79 FEP patients were recruited from NHS Glasgow and Edinburgh during a 12-month study between 2006 and 2009. Elastic net regularised logistic regression models were built to predict binary EET status, period and point remission outcomes at one-year on 83 Glasgow patients (training dataset). Models were externally validated on an independent dataset of 79 patients from Glasgow and Edinburgh (validation dataset). Only baseline predictors shared across both cohorts were made available for model training and validation. After excluding participants with missing outcomes, models were built on the training dataset for EET status, period and point remission outcomes and externally validated on the validation dataset. Models predicted EET status, period and point remission with receiver operating curve (ROC) area under the curve (AUC) performances of 0.876 (95%CI: 0.864, 0.887), 0.630 (95%CI: 0.612, 0.647) and 0.652 (95%CI: 0.635, 0.670) respectively. Positive predictors of EET included baseline EET and living with spouse/children. Negative predictors included higher PANSS suspiciousness, hostility and delusions scores. Positive predictors for symptom remission included living with spouse/children, and affective symptoms on the Positive and Negative Syndrome Scale (PANSS). Negative predictors of remission included passive social withdrawal symptoms on PANSS. A key limitation of this study is the small sample size (n) relative to the number of predictors (p), whereby p approaches n. The use of elastic net regularised regression rather than ordinary least squares regression helped circumvent this difficulty. Further, we did not have information for biological and additional social variables, such as nicotine dependence, which observational studies have linked to outcomes in psychosis.Conclusions and relevanceUsing advanced statistical machine learning techniques, we provide the first externally validated evidence, in a temporally and geographically independent cohort, for the ability to predict one-year EET status and symptom remission in individual FEP patients.

Dataset Information

Using machine-learning methods to identify early-life predictors of 11-year language outcome.

Background

Methods

Results

Conclusions

Publications

Using machine-learning methods to identify early-life predictors of 11-year language outcome.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets