Dataset Information

Development of machine learning models to prognosticate chronic shunt-dependent hydrocephalus after aneurysmal subarachnoid hemorrhage.

ABSTRACT: BACKGROUND:Shunt-dependent hydrocephalus significantly complicates subarachnoid hemorrhage (SAH), and reliable prognosis methods have been sought in recent years to reduce morbidity and costs associated with delayed treatment or neglected onset. Machine learning (ML) defines modern data analysis techniques allowing accurate subject-based risk stratifications. We aimed at developing and testing different ML models to predict shunt-dependent hydrocephalus after aneurysmal SAH. METHODS:We consulted electronic records of patients with aneurysmal SAH treated at our institution between January 2013 and March 2019. We selected variables for the models according to the results of the previous works on this topic. We trained and tested four ML algorithms on three datasets: one containing binary variables, one considering variables associated with shunt-dependency after an explorative analysis, and one including all variables. For each model, we calculated AUROC, specificity, sensitivity, accuracy, PPV, and also, on the validation set, the NPV and the Matthews correlation coefficient (?). RESULTS:Three hundred eighty-six patients were included. Fifty patients (12.9%) developed shunt-dependency after a mean follow-up of 19.7 (±?12.6) months. Complete information was retrieved for 32 variables, used to train the models. The best models were selected based on the performances on the validation set and were achieved with a distributed random forest model considering 21 variables, with a ? =?0.59, AUC?=?0.88; sensitivity and specificity of 0.73 (C.I.: 0.39-0.94) and 0.92 (C.I.: 0.84-0.97), respectively; PPV?=?0.59 (0.38-0.77); and NPV?=?0.96 (0.90-0.98). Accuracy was 0.90 (0.82-0.95). CONCLUSIONS:Machine learning prognostic models allow accurate predictions with a large number of variables and a more subject-oriented prognosis. We identified a single best distributed random forest model, with an excellent prognostic capacity (??=?0.58), which could be especially helpful in identifying low-risk patients for shunt-dependency.

SUBMITTER: Muscas G

PROVIDER: S-EPMC7593274 | biostudies-literature | 2020 Dec

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Development of machine learning models to prognosticate chronic shunt-dependent hydrocephalus after aneurysmal subarachnoid hemorrhage.

Muscas Giovanni G Matteuzzi Tommaso T Becattini Eleonora E Orlandini Simone S Battista Francesca F Laiso Antonio A Nappini Sergio S Limbucci Nicola N Renieri Leonardo L Carangelo Biagio R BR Mangiafico Salvatore S Della Puppa Alessandro A

Acta neurochirurgica 20200708 12

<h4>Background</h4>Shunt-dependent hydrocephalus significantly complicates subarachnoid hemorrhage (SAH), and reliable prognosis methods have been sought in recent years to reduce morbidity and costs associated with delayed treatment or neglected onset. Machine learning (ML) defines modern data analysis techniques allowing accurate subject-based risk stratifications. We aimed at developing and testing different ML models to predict shunt-dependent hydrocephalus after aneurysmal SAH.<h4>Methods</ ...[more]

PMID: 32642833

Similar Datasets

Project description:Background and purposeCurrent guidelines do not support the routine use of corticosteroids in patients with aneurysmal subarachnoid hemorrhage (aSAH). However, corticosteroids use in aSAH has been practiced at some centers by convention. The aim of the study was to determine the incidence of hydrocephalus requiring ventriculoperitoneal shunt (VPS) placement as well as functional outcome on discharge and adverse events attributed to corticosteroids in patients with aSAH treated with different dexamethasone (DXM) treatment schemes.MethodsWe retrospectively analyzed 206 patients with aSAH stratified to three groups based on the DXM treatment scheme: no corticosteroids, short course of DXM (S-DXM; 4 mg every 6 h for 1 day followed by a daily total dose reduction by 25% and then by 50% on last day), and long course of DXM (L-DXM; 4 mg every 6 h for 5-7 days followed by reduction by 50% every other day). The primary outcome measure was the placement of a VPS, and the secondary outcome was a good functional outcome [modified Rankin Scale (mRS) 0-3] at hospital discharge. Safety measures were the incidence of infection (pneumonia, urinary tract infection, ventriculitis, meningitis), presence of delirium, and hyperglycemia.ResultsThere was no difference in the rate of external ventricular drain (EVD) (p = 0.164) and VPS placement (p = 0.792), nor in the rate of good outcome (p = 0.928) among three defined treatment regimens. Moreover, the median duration of treatment with EVD did not differ between subjects treated with no corticosteroids, S-DXM, and L-DXM (p = 0.905), and the probability of EVD removal was similar when stratified according to treatment regimens (log-rank; p = 0.256). Patients who received L-DXM had significantly more complications as compared to patients, who received no corticosteroids or S-DXM (78.4% vs. 58.6%; p = 0.005). After adjustment, L-DXM remained independently associated with increased risk of combined adverse events (OR = 2.72; 95%CI, 1.30-5.72; p = 0.008), infection (OR = 3.45; 95%CI, 1.63-7.30; p = 0.001) and hyperglycemia (OR = 2.05; 95%CI, 1.04-4.04; p = 0.039).ConclusionsDXM use among patients with aSAH did not relate to the rate of EVD and VPS placement, duration of EVD treatment, and functional disability at discharge but increased the risk of medical complications.

Project description:Reliable prediction of outcomes of aneurysmal subarachnoid hemorrhage (aSAH) based on factors available at patient admission may support responsible allocation of resources as well as treatment decisions. Radiographic and clinical scoring systems may help clinicians estimate disease severity, but their predictive value is limited, especially in devising treatment strategies. In this study, we aimed to examine whether a machine learning (ML) approach using variables available on admission may improve outcome prediction in aSAH compared to established scoring systems. Combined clinical and radiographic features as well as standard scores (Hunt & Hess, WFNS, BNI, Fisher, and VASOGRADE) available on patient admission were analyzed using a consecutive single-center database of patients that presented with aSAH (n = 388). Different ML models (seven algorithms including three types of traditional generalized linear models, as well as a tree bosting algorithm, a support vector machine classifier (SVMC), a Naive Bayes (NB) classifier, and a multilayer perceptron (MLP) artificial neural net) were trained for single features, scores, and combined features with a random split into training and test sets (4:1 ratio), ten-fold cross-validation, and 50 shuffles. For combined features, feature importance was calculated. There was no difference in performance between traditional and other ML applications using traditional clinico-radiographic features. Also, no relevant difference was identified between a combined set of clinico-radiological features available on admission (highest AUC 0.78, tree boosting) and the best performing clinical score GCS (highest AUC 0.76, tree boosting). GCS and age were the most important variables for the feature combination. In this cohort of patients with aSAH, the performance of functional outcome prediction by machine learning techniques was comparable to traditional methods and established clinical scores. Future work is necessary to examine input variables other than traditional clinico-radiographic features and to evaluate whether a higher performance for outcome prediction in aSAH can be achieved.

Dataset Information

Development of machine learning models to prognosticate chronic shunt-dependent hydrocephalus after aneurysmal subarachnoid hemorrhage.

Publications

Development of machine learning models to prognosticate chronic shunt-dependent hydrocephalus after aneurysmal subarachnoid hemorrhage.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets