RESUMO
Machine learning (ML) has revolutionized data processing in recent years. This study presents the results of the first prediction models based on a long-term monocentric data registry of patients with microsurgically treated unruptured intracranial aneurysms (UIAs) using a temporal train-test split. Temporal train-test splits allow to simulate prospective validation, and therefore provide more accurate estimations of a model's predictive quality when applied to future patients. ML models for the prediction of the Glasgow outcome scale, modified Rankin Scale (mRS), and new transient or permanent neurological deficits (output variables) were created from all UIA patients that underwent microsurgery at the Kepler University Hospital Linz (Austria) between 2002 and 2020 (n = 466), based on 18 patient- and 10 aneurysm-specific preoperative parameters (input variables). Train-test splitting was performed with a temporal split for outcome prediction in microsurgical therapy of UIA. Moreover, an external validation was conducted on an independent external data set (n = 256) of the Department of Neurosurgery, University Medical Centre Hamburg-Eppendorf. In total, 722 aneurysms were included in this study. A postoperative mRS > 2 was best predicted by a quadratic discriminant analysis (QDA) estimator in the internal test set, with an area under the receiver operating characteristic curve (ROC-AUC) of 0.87 ± 0.03 and a sensitivity and specificity of 0.83 ± 0.08 and 0.71 ± 0.07, respectively. A Multilayer Perceptron predicted the post- to preoperative mRS difference > 1 with a ROC-AUC of 0.70 ± 0.02 and a sensitivity and specificity of 0.74 ± 0.07 and 0.50 ± 0.04, respectively. The QDA was the best model for predicting a permanent new neurological deficit with a ROC-AUC of 0.71 ± 0.04 and a sensitivity and specificity of 0.65 ± 0.24 and 0.60 ± 0.12, respectively. Furthermore, these models performed significantly better than the classic logistic regression models (p < 0.0001). The present results showed good performance in predicting functional and clinical outcomes after microsurgical therapy of UIAs in the internal data set, especially for the main outcome parameters, mRS and permanent neurological deficit. The external validation showed poor discrimination with ROC-AUC values of 0.61, 0.53 and 0.58 respectively for predicting a postoperative mRS > 2, a pre- and postoperative difference in mRS > 1 point and a GOS < 5. Therefore, generalizability of the models could not be demonstrated in the external validation. A SHapley Additive exPlanations (SHAP) analysis revealed that this is due to the most important features being distributed quite differently in the internal and external data sets. The implementation of newly available data and the merging of larger databases to form more broad-based predictive models is imperative in the future.
Assuntos
Aneurisma Intracraniano , Humanos , Aneurisma Intracraniano/diagnóstico , Aneurisma Intracraniano/cirurgia , Prognóstico , Escala de Resultado de Glasgow , Procedimentos Neurocirúrgicos/métodos , Aprendizado de Máquina , Estudos RetrospectivosRESUMO
AIMS: Patient admission is a decision relying on sparsely available data. This study aims to provide prediction models for discharge versus admission for ward observation or intensive care, and 30 day-mortality for patients triaged with the Manchester Triage System. METHODS: This is a single-centre, observational, retrospective cohort study from data within ten minutes of patient presentation at the interdisciplinary emergency department of the Kepler University Hospital, Linz, Austria. We trained machine learning models including Random Forests and Neural Networks individually to predict discharge versus ward observation or intensive care admission, and 30 day-mortality. For analysis of the features' relevance, we used permutation feature importance. RESULTS: A total of 58323 adult patients between 1 December 2015 and 31 August 2020 were included. Neural Networks and Random Forests predicted admission to ward observation with an AUC-ROC of 0.842â ±â 0.00 with the most important features being age and chief complaint. For admission to intensive care, the models had an AUC-ROC of 0.819â ±â 0.002 with the most important features being the Manchester Triage category and heart rate, and for the outcome 30 day-mortality an AUC-ROC of 0.925â ±â 0.001. The most important features for the prediction of 30 day-mortality were age and general ward admission. CONCLUSION: Machine learning can provide prediction on discharge versus admission to general wards and intensive care and inform about risk on 30 day-mortality for patients in the emergency department.
Assuntos
Hospitalização , Triagem , Adulto , Humanos , Estudos Retrospectivos , Serviço Hospitalar de Emergência , Aprendizado de MáquinaRESUMO
Electronic health records (EHRs) have been successfully used in data science and machine learning projects. However, most of these data are collected for clinical use rather than for retrospective analysis. This means that researchers typically face many different issues when attempting to access and prepare the data for secondary use. We aimed to investigate how raw EHRs can be accessed and prepared in retrospective data science projects in a disciplined, effective, and efficient way. We report our experience and findings from a large-scale data science project analyzing routinely acquired retrospective data from the Kepler University Hospital in Linz, Austria. The project involved data collection from more than 150,000 patients over a period of 10 years. It included diverse data modalities, such as static demographic data, irregularly acquired laboratory test results, regularly sampled vital signs, and high-frequency physiological waveform signals. Raw medical data can be corrupted in many unexpected ways that demand thorough manual inspection and highly individualized data cleaning solutions. We present a general data preparation workflow, which was shaped in the course of our project and consists of the following 7 steps: obtain a rough overview of the available EHR data, define clinically meaningful labels for supervised learning, extract relevant data from the hospital's data warehouses, match data extracted from different sources, deidentify them, detect errors and inconsistencies therein through a careful exploratory analysis, and implement a suitable data processing pipeline in actual code. Only few of the data preparation issues encountered in our project were addressed by generic medical data preprocessing tools that have been proposed recently. Instead, highly individualized solutions for the specific data used in one's own research seem inevitable. We believe that the proposed workflow can serve as a guidance for practitioners, helping them to identify and address potential problems early and avoid some common pitfalls.
RESUMO
Our world is filled with texture. For the human visual system, this is an important source of information for assessing environmental and material properties. Indeed-and presumably for this reason-the human visual system has regions dedicated to processing textures. Despite their abundance and apparent relevance, only recently the relationships between texture features and high-level judgments have captured the interest of mainstream science, despite long-standing indications for such relationships. In this study, we explore such relationships, as these might be used to predict perceived texture qualities. This is relevant, not only from a psychological/neuroscience perspective, but also for more applied fields such as design, architecture, and the visual arts. In two separate experiments, observers judged various qualities of visual textures such as beauty, roughness, naturalness, elegance, and complexity. Based on factor analysis, we find that in both experiments, ~75% of the variability in the judgments could be explained by a two-dimensional space, with axes that are closely aligned to the beauty and roughness judgments. That a two-dimensional judgment space suffices to capture most of the variability in the perceived texture qualities suggests that observers use a relatively limited set of internal scales on which to base various judgments, including aesthetic ones. Finally, for both of these judgments, we determined the relationship with a large number of texture features computed for each of the texture stimuli. We find that the presence of lower spatial frequencies, oblique orientations, higher intensity variation, higher saturation, and redness correlates with higher beauty ratings. Features that captured image intensity and uniformity correlated with roughness ratings. Therefore, a number of computational texture features are predictive of these judgments. This suggests that perceived texture qualities-including the aesthetic appreciation-are sufficiently universal to be predicted-with reasonable accuracy-based on the computed feature content of the textures.
RESUMO
Little is known about the influence of visual characteristics other than colour on flavor perception, and the complex interactions between more than two sensory modalities. This study focused on the effects of recognizability of visual (texture) information on flavor perception of odorized sweet beverages. Participants rated the perceived sweetness of odorized sucrose solutions in the presence or absence of either a congruent or incongruent visual context. Odors were qualitatively reminiscent of sweet foods (strawberry and caramel) or not (savoury). Visual context was either an image of the same sweet foods (figurative context) or a visual texture derived from this product (non-figurative context). Textures were created using a texture synthesis method that preserved perceived food qualities while removing object information. Odor-taste combinations were rated sweeter within a figurative than a non-figurative context. This behaviour was exhibited for all odor-taste combinations, even in trials without images, indicating sustained priming by figurative visual context. A non-figurative context showed a transient sweetening effect. Sweetness was generally enhanced most by the strawberry odor. We conclude that the degree of recognizability of visual information (figurative versus non-figurative), influences flavor perception differently. Our results suggest that this visual context priming is mediated by separate sustained and transient processes that are differently evoked by figurative and non-figurative visual contexts. These components operate independent of the congruency of the image-odor-taste combinations.