Early Identification of High-Risk TIA or Minor Stroke Using Artificial Neural Network.

Background and Purpose: The risk of recurrent stroke following a transient ischemic attack (TIA) or minor stroke is high, despite of a significant reduction in the past decade. In this study, we investigated the feasibility of using artificial neural network (ANN) for risk stratification of TIA or minor stroke patients. Methods: Consecutive patients with acute TIA or minor ischemic stroke presenting at a tertiary hospital during a 2-year period were recruited. We collected demographics, clinical and imaging data at baseline. The primary outcome was recurrent ischemic stroke within 1 year. We developed ANN models to predict the primary outcome. We randomly down-sampled patients without a primary outcome to 1:1 match with those with a primary outcome to mitigate data imbalance. We used a 5-fold cross-validation approach to train and test the ANN models to avoid overfitting. We employed 19 independent variables at baseline as the input neurons in the ANN models, using a learning algorithm based on backpropagation to minimize the loss function. We obtained the sensitivity, specificity, accuracy and the c statistic of each ANN model from the 5 rounds of cross-validation and compared that of support vector machine (SVM) and Naïve Bayes classifier in risk stratification of the patients. Results: A total of 451 acute TIA or minor stroke patients were enrolled. Forty (8.9%) patients had a recurrent ischemic stroke within 1 year. Another 40 patients were randomly selected from those with no recurrent stroke, so that data from 80 patients in total were used for 5 rounds of training and testing of ANN models. The median sensitivity, specificity, accuracy and c statistic of the ANN models to predict recurrent stroke at 1 year was 75%, 75%, 75%, and 0.77, respectively. ANN model outperformed SVM and Naïve Bayes classifier in our dataset for predicting relapse after TIA or minor stroke. Conclusion: This pilot study indicated that ANN may yield a novel and effective method in risk stratification of TIA and minor stroke. Further studies are warranted for verification and improvement of the current ANN model.

Reconstruction of continuous brachial artery pressure wave from continuous finger arterial pressure in humans.

Generalized transfer functions (GTFs) are available to compute the more relevant proximal blood pressure (BP) waveform from a more easily measured peripheral BP waveform. However, GTFs are based on the black box model. This paper presents a practical approach to reconstruct brachial artery pressure (BAP) distally from finger artery pressure (FAP). We assume that continuous BAP can be simply approximated by summing two halves of the continuous FAP shifted by the time delay. We firstly showed that the pressure wave in the finger artery can be considered twice as much as the forward/backward wave in the finger. A simplified individualized transfer function was then derived so as to estimate BAP from FAP. The effectiveness of the method was examined by experiment involving 26 healthy volunteers (26.7 ± 3.8 years old) in a resting state. By comparing with a reference BAP, we found that the proposed method can correct the FAP. The errors of the proposed method in estimating systolic and diastolic pressures are - 0.6 ± 6.0 and - 0.6 ± 3.7 mmHg, respectively. These results agree with the standard of Association for the Advancement of Medical Instrumentation (AAMI). We also found that the reconstructed BAP from FAP by terminal arterial occlusion technology (TAOT) is comparable to that of the artery occlusion technology (AOT). Our method or TAOT is promising in estimating continuous proximal blood pressure from peripheral blood pressure in practice.

Compromised dynamic cerebral autoregulation in patients with generalized anxiety disorder: a study using transfer function analysis.

BACKGROUND: Patients with generalized anxiety disorder (GAD) usually present with various neurological symptoms, but the mechanisms remain unclear. We aimed to analyze the characteristics of dynamic cerebral autoregulation (dCA) in patients with GAD. METHODS: Patients (aged ≥18 years) who were diagnosed with GAD were enrolled in this study. Medically and psychiatrically healthy volunteers were recruited as controls. Subjects received the Hamilton Rating Scale for Anxiety (HAMA) and 17-item Hamilton Depression Rating Scale (HAMD) evaluation. Noninvasive continuous arterial blood pressure and bilateral middle cerebral artery blood flow velocity were recorded simultaneously from each subject. Transfer function analysis was used to derive the autoregulatory parameters, including phase difference, gain, and coherence function. RESULTS: A total of 57 patients with GAD and 40 healthy volunteers were enrolled. We found that the phase difference values were significantly compromised in patients with GAD. In the Spearman correlation analysis, the phase difference values were negatively correlated with the HAMA scores and the HAMD scores. In the multiple linear regression analysis, GAD is negatively correlated with the phase difference values, whereas age is positively correlated with the phase difference values. CONCLUSIONS: Our results suggested that the dCA was compromised in patients with GAD and negatively correlated with the score of anxiety. Improving the dCA may be a potential therapeutic method for treating the neurological symptoms of GAD patients.

Compromised Dynamic Cerebral Autoregulation in Patients with Epilepsy.

OBJECTIVE: The aim of this study is to analyze dynamic cerebral autoregulation (dCA) in patients with epilepsy. METHODS: One hundred patients with epilepsy and 100 age- and sex-matched healthy controls were recruited. Noninvasive continuous cerebral blood flow velocity of the bilateral middle artery and arterial blood pressure were recorded. Transfer function analyses were used to analyze the autoregulatory parameters (phase difference and gain). RESULTS: The overall phase difference of patients with epilepsy was significantly lower than that of the healthy control group (p = 0.046). Furthermore, patients with interictal slow wave had significant lower phase difference than the slow-wave-free patients (p = 0.012). There was no difference in overall phase between focal discharges and multifocal discharges in patients with epilepsy. Simultaneously, there was no difference in mean phase between the affected and unaffected hemispheres in patients with unilateral discharges. In particular, interictal slow wave was an independent factor that influenced phase difference in patients with epilepsy (p = 0.016). CONCLUSIONS: Our study documented that dCA is impaired in patients with epilepsy, especially in those with interictal slow wave. The impairment of dCA occurs irrespective of the discharge location and type. Interictal slow wave is an independent factor to predict impaired dCA in patients with epilepsy. CLINICAL TRIAL IDENTIFIER: This trial is registered with NCT02775682.