RESUMEN

BACKGROUND: Current guidelines regarding oxytocin stimulation are not tailored to individuals as they are based on randomised controlled trials. The objective of the study was to develop an artificial intelligence (AI) model for individual prediction of the risk of caesarean delivery (CD) in women with a cervical dilatation of 6 cm after oxytocin stimulation for induced labour. The model included not only variables known when labour induction was initiated but also variables describing the course of the labour induction. METHODS: Secondary analysis of data from the CONDISOX randomised controlled trial of discontinued vs. continued oxytocin infusion in the active phase of induced labour. Extreme gradient boosting (XGBoost) software was used to build the prediction model. To explain the impact of the predictors, we calculated Shapley additive explanation (SHAP) values and present a summary SHAP plot. A force plot was used to explain specifics about an individual's predictors that result in a change of the individual's risk output value from the population-based risk. RESULTS: Among 1060 included women, 160 (15.1%) were delivered by CD. The XGBoost model found women who delivered vaginally were more likely to be parous, taller, to have a lower estimated birth weight, and to be stimulated with a lower amount of oxytocin. In 108 women (10% of 1060) the model favoured either continuation or discontinuation of oxytocin. For the remaining 90% of the women, the model found that continuation or discontinuation of oxytocin stimulation affected the risk difference of CD by less than 5% points. CONCLUSION: In women undergoing labour induction, this AI model based on a secondary analysis of data from the CONDISOX trial may help predict the risk of CD and assist the mother and clinician in individual tailored management of oxytocin stimulation after reaching 6 cm of cervical dilation.

Asunto(s)

RESUMEN

BACKGROUND: The increasing aging population and limited health care resources have placed new demands on the healthcare sector. Reducing the number of hospitalizations has become a political priority in many countries, and special focus has been directed at potentially preventable hospitalizations. OBJECTIVES: We aimed to develop an artificial intelligence (AI) prediction model for potentially preventable hospitalizations in the coming year, and to apply explainable AI to identify predictors of hospitalization and their interaction. METHODS: We used the Danish CROSS-TRACKS cohort and included citizens in 2016-2017. We predicted potentially preventable hospitalizations within the following year using the citizens' sociodemographic characteristics, clinical characteristics, and health care utilization as predictors. Extreme gradient boosting was used to predict potentially preventable hospitalizations with Shapley additive explanations values serving to explain the impact of each predictor. We reported the area under the receiver operating characteristic curve, the area under the precision-recall curve, and 95% confidence intervals (CI) based on five-fold cross-validation. RESULTS: The best performing prediction model showed an area under the receiver operating characteristic curve of 0.789 (CI: 0.782-0.795) and an area under the precision-recall curve of 0.232 (CI: 0.219-0.246). The predictors with the highest impact on the prediction model were age, prescription drugs for obstructive airway diseases, antibiotics, and use of municipality services. We found an interaction between age and use of municipality services, suggesting that citizens aged 75+ years receiving municipality services had a lower risk of potentially preventable hospitalization. CONCLUSION: AI is suitable for predicting potentially preventable hospitalizations. The municipality-based health services seem to have a preventive effect on potentially preventable hospitalizations.

Asunto(s)

RESUMEN

PURPOSE: In this study, we investigate the potential of a novel artificial intelligence-based system for autonomous follow-up of patients treated for neovascular age-related macular degeneration (AMD). METHODS: A temporal deep learning model was trained on a data set of 84 489 optical coherence tomography scans from AMD patients to recognize disease activity, and its performance was compared with a published non-temporal model trained on the same data (Acta Ophthalmol, 2021). An autonomous follow-up system was created by augmenting the AI model with deterministic logic to suggest treatment according to the observe-and-plan regimen. To validate the AI-based system, a data set comprising clinical decisions and imaging data from 200 follow-up consultations was collected prospectively. In each case, both the autonomous AI decision and original clinical decision were compared with an expert panel consensus. RESULTS: The temporal AI model proved superior at detecting disease activity compared with the model without temporal input (area under the curve 0.900 (95% CI 0.894-0.906) and 0.857 (95% CI 0.846-0.867) respectively). The AI-based follow-up system could make an autonomous decision in 73% of the cases, 91.8% of which were in agreement with expert consensus. This was on par with the 87.7% agreement rate between decisions made in the clinic and expert consensus (p = 0.33). CONCLUSIONS: The proposed autonomous follow-up system was shown to be safe and compliant with expert consensus on par with clinical practice. The system could in the future ease the pressure on public ophthalmology services from an increasing number of AMD patients.

Asunto(s)

RESUMEN

PURPOSE: To meet the demands imposed by the continuing growth of the Age-related macular degeneration (AMD) patient population, automation of follow-ups by detecting retinal oedema using deep learning might be a viable approach. However, preparing and labelling data for training is time consuming. In this study, we investigate the feasibility of training a convolutional neural network (CNN) to accurately detect retinal oedema on optical coherence tomography (OCT) images of AMD patients with labels derived directly from clinical treatment decisions, without extensive preprocessing or relabelling. METHODS: A total of 50 439 OCT images with associated treatment information were retrieved from databases at the Department of Ophthalmology, Rigshospitalet, Copenhagen, Denmark between 01.06.2007 and 01.06.2018. A CNN was trained on the retrieved data with the recorded treatment decisions as labels and validated on a subset of the data relabelled by three ophthalmologists to denote presence of oedema. RESULTS: Moderate inter-grader agreement on presence of oedema in the relabelled data was found (76.4%). Despite different training and validation labels, the CNN performed on par with inter-grader agreement in detecting oedema on OCT images (AUC 0.97, accuracy 90.9%) and previously published models based on relabelled datasets. CONCLUSION: The level of performance shown by the current model might make it valuable in detecting disease activity in automated AMD patient follow-up systems. Our approach demonstrates that high accuracy is not necessarily constrained by incongruent training and validation labels. These results might encourage the use of existing clinical databases for development of deep learning based algorithms without labour-intensive preprocessing in the future.

Asunto(s)

RESUMEN

Problem framing is critical to developing risk prediction models because all subsequent development work and evaluation takes place within the context of how a problem has been framed and explicit documentation of framing choices makes it easier to compare evaluation metrics between published studies. In this work, we introduce the basic concepts of framing, including prediction windows, observation windows, window shifts and event-triggers for a prediction that strongly affects the risk of clinician fatigue caused by false positives. Building on this, we apply four different framing structures to the same generic dataset, using a sepsis risk prediction model as an example, and evaluate how framing affects model performance and learning. Our results show that an apparently good model with strong evaluation results in both discrimination and calibration is not necessarily clinically usable. Therefore, it is important to assess the results of objective evaluations within the context of more subjective evaluations of how a model is framed.

RESUMEN

Acute critical illness is often preceded by deterioration of routinely measured clinical parameters, e.g., blood pressure and heart rate. Early clinical prediction is typically based on manually calculated screening metrics that simply weigh these parameters, such as early warning scores (EWS). The predictive performance of EWSs yields a tradeoff between sensitivity and specificity that can lead to negative outcomes for the patient. Previous work on electronic health records (EHR) trained artificial intelligence (AI) systems offers promising results with high levels of predictive performance in relation to the early, real-time prediction of acute critical illness. However, without insight into the complex decisions by such system, clinical translation is hindered. Here, we present an explainable AI early warning score (xAI-EWS) system for early detection of acute critical illness. xAI-EWS potentiates clinical translation by accompanying a prediction with information on the EHR data explaining it.

Asunto(s)

RESUMEN

BACKGROUND: The timeliness of detection of a sepsis incidence in progress is a crucial factor in the outcome for the patient. Machine learning models built from data in electronic health records can be used as an effective tool for improving this timeliness, but so far, the potential for clinical implementations has been largely limited to studies in intensive care units. This study will employ a richer data set that will expand the applicability of these models beyond intensive care units. Furthermore, we will circumvent several important limitations that have been found in the literature: (1) Model evaluations neglect the clinical consequences of a decision to start, or not start, an intervention for sepsis. (2) Models are evaluated shortly before sepsis onset without considering interventions already initiated. (3) Machine learning models are built on a restricted set of clinical parameters, which are not necessarily measured in all departments. (4) Model performance is limited by current knowledge of sepsis, as feature interactions and time dependencies are hard-coded into the model. METHODS: In this study, we present a model to overcome these shortcomings using a deep learning approach on a diverse multicenter data set. We used retrospective data from multiple Danish hospitals over a seven-year period. Our sepsis detection system is constructed as a combination of a convolutional neural network and a long short-term memory network. We assess model quality by standard concepts of accuracy as well as clinical usefulness, and we suggest a retrospective assessment of interventions by looking at intravenous antibiotics and blood cultures preceding the prediction time. RESULTS: Results show performance ranging from AUROC 0.856 (3 h before sepsis onset) to AUROC 0.756 (24 h before sepsis onset). Evaluating the clinical utility of the model, we find that a large proportion of septic patients did not receive antibiotic treatment or blood culture at the time of the sepsis prediction, and the model could, therefore, facilitate such interventions at an earlier point in time. CONCLUSION: We present a deep learning system for early detection of sepsis that can learn characteristics of the key factors and interactions from the raw event sequence data itself, without relying on a labor-intensive feature extraction work. Our system outperforms baseline models, such as gradient boosting, which rely on specific data elements and therefore suffer from many missing values in our dataset.

Asunto(s)