Erik D. Gommer

Dynamic Cerebral Autoregulation in Subjects with Alzheimer's Disease, Mild Cognitive Impairment, and Controls: Evidence for Increased Peripheral Vascular Resistance with Possible Predictive Value

Cerebrovascular dysfunction plays a role not only in vascular causes of cognitive impairment but also in Alzheimers disease (AD). We hypothesized that cerebral autoregulation is impaired in patients with AD compared to subjects with mild cognitive impairment (MCI) and controls. Dynamic cerebral autoregulation (dCA) was investigated in 17 AD patients, 19 MCI subjects, and 20 controls (C). Groups were matched for age, gender, and level of education. Electrocardiogram and non-invasive finger arterial blood pressure were measured and transcranial doppler ultrasonography was used to measure cerebral blood flow velocity in right and left middle cerebral artery (MCA). Cerebrovascular resistance index (CVRi) was also computed. dCA in supine position was quantified based on spontaneous blood pressure variations by computation of the linear transfer function between arterial blood pressure and MCA cerebral blood flow velocity. dCA gain and phase were evaluated for different frequency bands. Results were also evaluated using a 3-parameter windkessel model (WKM). CVRi was significantly higher in AD (2.9 ± 0.2) compared to both MCI (2.3 ± 0.1, p = 0.02) and C (2.1 ± 0.1 mmHgs/cm, p = 0.002). Five MCI patients who converted to AD during the course of the study also had higher CVRi compared to non-converters (2.8 ± 0.6 versus 2.1 ± 0.5 mmHgs/cm, p < 0.05). No significant differences in dCA gain and phase were found. In terms of the WKM approach, in the order C→MCI→AD groups showed about equal arterial resistance and peripheral compliance, but increased peripheral vasculature resistance (26 ± 2 versus 36 ± 3 mmHgs/ml in C resp. AD, p = 0.004). In conclusion, AD patients compared to MCI patients and controls have increased CVRi, whereas dCA parameters do not seem to differentiate AD patients. For MCI patients, CVRi might have predictive value in developing AD.

Baroreflex function is reduced in Alzheimer's disease: a candidate biomarker?

The baroreflex (BR) reflects autonomic blood pressure control. Alzheimers disease (AD) affects the autonomic system. Detailed properties of BR in AD are unknown. We hypothesized that BR is reduced in AD, and is influenced by autonomic effects of cholinesterase inhibitors (ChEI). BR was determined in 18 AD patients, 11 patients with mild cognitive impairment (MCI) and 19 healthy control subjects. In AD, BR was measured again after ChEI treatment. Receiver operating characteristic analysis was used to define a BR cutoff value, which was then tested in an independent validation sample of 16 AD, 18 MCI, and 18 control subjects. BR was lower in AD compared with MCI (p < 0.05) and in MCI compared with healthy control subjects (p < 0.01). Receiver operating characteristic analysis between AD and healthy control subjects yielded a sensitivity of 89% and a specificity of 94%. ChEI treatment increased BR with 66% (p < 0.01). BR was reduced in AD and increased after treatment with ChEI. BR might be a good biomarker to further explore the link between cardiovascular disease and AD.

Validity of a Non-invasive Determination of the Isovolumetric Bladder Pressure During Voiding in Men With LUTS

Invasive pressure flow analysis is the gold standard for discriminating between hypocontractile bladder muscle function and infravesical obstruction in male patients with lower urinary tract symptoms. Here a non‐invasive method to determine the isovolumetric bladder pressure to judge contractility is presented. This is based on interruption of urine flow by sudden occlusion of a specially fixed condom catheter. The pressure inside the condom is recorded and used to estimate the isovolumetric bladder pressure. Combined with, for example, home uroflowmetry, this non‐invasive method may overcome some of the disadvantages (e.g., invasiveness, cost) of the conventional pressure flow test. To determine the isovolumetric bladder pressure reliably with this non‐invasive method, two constraints have to be met. First, the bladder neck and urethra have to remain open after occlusion of the condom catheter. This was tested combining the non‐invasive test with radiography in five patients. Second, a steady state has to be reached, i.e., the flow in the urethra, due to the elastic properties of the biological and the condom systems, should come to a stop when the bladder pressure and the condom pressure equilibrate. This was investigated by comparing the non‐invasively recorded condom pressure with the simultaneously invasively recorded intravesical pressure in 52 patients. In these patients, three different methods of condom fixation were evaluated. The results show that the bladder neck and urethra remain open during the test. However, a steady state is often not reached. In more than 80% of the cases with the best condom fixation, the bladder pressure has not stabilized, although the condom pressure reached a plateau. Therefore, this method of sudden occlusion is not yet clinically applicable for determining the isovolumetric bladder pressure. Neurourol. Urodynam. 18:477–486, 1999.

Between-centre variability in transfer function analysis, a widely used method for linear quantification of the dynamic pressure–flow relation: The CARNet study☆

Transfer function analysis (TFA) is a frequently used method to assess dynamic cerebral autoregulation (CA) using spontaneous oscillations in blood pressure (BP) and cerebral blood flow velocity (CBFV). However, controversies and variations exist in how research groups utilise TFA, causing high variability in interpretation. The objective of this study was to evaluate between-centre variability in TFA outcome metrics. 15 centres analysed the same 70 BP and CBFV datasets from healthy subjects (n=50 rest; n=20 during hypercapnia); 10 additional datasets were computer-generated. Each centre used their in-house TFA methods; however, certain parameters were specified to reduce a priori between-centre variability. Hypercapnia was used to assess discriminatory performance and synthetic data to evaluate effects of parameter settings. Results were analysed using the Mann-Whitney test and logistic regression. A large non-homogeneous variation was found in TFA outcome metrics between the centres. Logistic regression demonstrated that 11 centres were able to distinguish between normal and impaired CA with an AUC>0.85. Further analysis identified TFA settings that are associated with large variation in outcome measures. These results indicate the need for standardisation of TFA settings in order to reduce between-centre variability and to allow accurate comparison between studies. Suggestions on optimal signal processing methods are proposed.

A lumped parameter model of cerebral blood flow control combining cerebral autoregulation and neurovascular coupling

Cerebral blood flow regulation is based on a variety of different mechanisms, of which the relative regulatory role remains largely unknown. The cerebral regulatory system expresses two regulatory properties: cerebral autoregulation and neurovascular coupling. Since partly the same mechanisms play a role in cerebral autoregulation and neurovascular coupling, this study aimed to develop a physiologically based mathematical model of cerebral blood flow regulation combining these properties. A lumped parameter model of the P2 segment of the posterior cerebral artery and its distal vessels was constructed. Blood flow regulation is exerted at the arteriolar level by vascular smooth muscle and implements myogenic, shear stress based, neurogenic, and metabolic mechanisms. In eight healthy subjects, cerebral autoregulation and neurovascular coupling were challenged by squat-stand maneuvers and visual stimulation using a checkerboard pattern, respectively. Cerebral blood flow velocity was measured using transcranial Doppler, whereas blood pressure was measured by finger volume clamping. In seven subjects, the model proposed fits autoregulation and neurovascular coupling measurement data well. Myogenic regulation is found to dominate the autoregulatory response. Neurogenic regulation, although only implemented as a first-order mechanism, describes neurovascular coupling responses to a great extent. It is concluded that our single, integrated model of cerebral blood flow control may be used to identify the main mechanisms affecting cerebral blood flow regulation in individual subjects.

EEG Feature Pre-processing for Neonatal Epileptic Seizure Detection

Aim of our project is to further optimize neonatal seizure detection using support vector machine (SVM). First, a Kalman filter (KF) was used to filter both feature and classifier output time series in order to increase temporal precision. Second, EEG baseline feature correction (FBC) was introduced to reduce inter patient variability in feature distributions. The performance of the detection methods is evaluated on 54 multi channel routine EEG recordings from 39 both term and pre-term newborns. The area under the receiver operating characteristics curve (AUC) as well as sensitivity and specificity are used to evaluate the performance of the classification method. SVM without KF and FBC achieves an AUC of 0.767 (sensitivity 0.679, specificity 0.707). The highest AUC of 0.902 (sensitivity 0.801, specificity 0.831) is achieved on baseline corrected features with a Kalman smoother used for training data pre-processing and a KF used to filter the classifier output. Both FBC and KF significantly improve neonatal epileptic seizure detection. This paper introduces significant improvements for the state of the art SVM based neonatal epileptic seizure detection.

Visually evoked blood flow responses and interaction with dynamic cerebral autoregulation: Correction for blood pressure variation

Visually evoked flow responses recorded using transcranial Doppler ultrasonography are often quantified using a dynamic model of neurovascular coupling. The evoked flow response is seen as the models response to a visual step input stimulus. However, the continuously active process of dynamic cerebral autoregulation (dCA) compensating cerebral blood flow for blood pressure fluctuations may induce changes of cerebral blood flow velocity (CBFV) as well. The effect of blood pressure variability on the flow response is evaluated by separately modeling the dCA-induced effects of beat-to-beat measured blood pressure related CBFV changes. Parameters of 71 subjects are estimated using an existing, well-known second order dynamic neurovascular coupling model proposed by Rosengarten et al., and a new model extending the existing model with a CBFV contributing component as the output of a dCA model driven by blood pressure as input. Both models were evaluated for mean and systolic CBFV responses. The model-to-data fit errors of mean and systolic blood pressure for the new model were significantly lower compared to the existing model: mean: 0.8%±0.6 vs. 2.4%±2.8, p<0.001; systolic: 1.5%±1.2 vs. 2.2%±2.6, p<0.001. The confidence bounds of all estimated neurovascular coupling model parameters were significantly (p<0.005) narrowed for the new model. In conclusion, blood pressure correction of visual evoked flow responses by including cerebral autoregulation in model fitting of averaged responses results in significantly lower fit errors and by that in more reliable model parameter estimation. Blood pressure correction is more effective when mean instead of systolic CBFV responses are used. Measurement and quantification of neurovascular coupling should include beat-to-beat blood pressure measurement.

Optimal training dataset composition for SVM-based, age-independent, automated epileptic seizure detection

Automated seizure detection is a valuable asset to health professionals, which makes adequate treatment possible in order to minimize brain damage. Most research focuses on two separate aspects of automated seizure detection: EEG feature computation and classification methods. Little research has been published regarding optimal training dataset composition for patient-independent seizure detection. This paper evaluates the performance of classifiers trained on different datasets in order to determine the optimal dataset for use in classifier training for automated, age-independent, seizure detection. Three datasets are used to train a support vector machine (SVM) classifier: (1) EEG from neonatal patients, (2) EEG from adult patients and (3) EEG from both neonates and adults. To correct for baseline EEG feature differences among patients feature, normalization is essential. Usually dedicated detection systems are developed for either neonatal or adult patients. Normalization might allow for the development of a single seizure detection system for patients irrespective of their age. Two classifier versions are trained on all three datasets: one with feature normalization and one without. This gives us six different classifiers to evaluate using both the neonatal and adults test sets. As a performance measure, the area under the receiver operating characteristics curve (AUC) is used. With application of FBC, it resulted in performance values of 0.90 and 0.93 for neonatal and adult seizure detection, respectively. For neonatal seizure detection, the classifier trained on EEG from adult patients performed significantly worse compared to both the classifier trained on EEG data from neonatal patients and the classier trained on both neonatal and adult EEG data. For adult seizure detection, optimal performance was achieved by either the classifier trained on adult EEG data or the classifier trained on both neonatal and adult EEG data. Our results show that age-independent seizure detection is possible by training one classifier on EEG data from both neonatal and adult patients. Furthermore, our results indicate that for accurate age-independent seizure detection, it is important that EEG data from each age category are used for classifier training. This is particularly important for neonatal seizure detection. Our results underline the under-appreciated importance of training dataset composition with respect to accurate age-independent seizure detection.

Hemodilution Combined With Hypercapnia Impairs Cerebral Autoregulation During Normothermic Cardiopulmonary Bypass

OBJECTIVE To investigate the influence of hemodilution and arterial pCO2 on cerebral autoregulation and cerebral vascular CO2 reactivity. DESIGN Prospective interventional study. SETTING University hospital-based single-center study. PARTICIPANTS Forty adult patients undergoing elective cardiac surgery using normothermic cardiopulmonary bypass. INTERVENTIONS Blood pressure variations induced by 6/minute metronome-triggered breathing (baseline) and cyclic 6/min changes of indexed pump flow at 3 levels of arterial pCO2. MEASUREMENTS AND MAIN RESULTS Based on median hematocrit on bypass, patients were assigned to either a group of a hematocrit ≥28% or<28%. The autoregulation index was calculated from cerebral blood flow velocity and mean arterial blood pressure using transfer function analysis. Cerebral vascular CO2 reactivity was calculated using cerebral tissue oximetry data. Cerebral autoregulation as reflected by autoregulation index (baseline 7.5) was significantly affected by arterial pCO2 (median autoregulation index amounted to 5.7, 4.8, and 2.8 for arterial pCO2 of 4.0, 5.3, and 6.6 kPa, p≤0.002) respectively. Hemodilution resulted in a decreased autoregulation index; however, during hypocapnia and normocapnia, there were no significant differences between the two hematocrit groups. Moreover, the autoregulation index was lowest during hypercapnia when hematocrit was<28% (autoregulation index 3.3 versus 2.6 for hematocrit ≥28% and<28%, respectively, p = 0.014). Cerebral vascular CO2 reactivity during hypocapnia was significantly lower when perioperative hematocrit was<28% (p = 0.018). CONCLUSIONS Hemodilution down to a hematocrit of<28% combined with hypercapnia negatively affects dynamic cerebral autoregulation, which underlines the importance of tight control of both hematocrit and paCO2 during CPB.

Unravelling gene interactions to find the cause of artherosclerosis, a multigenic disease, using an artificial neural network

To understand the etiology of multigenic diseases like atherosclerosis, a polymerase chain reaction (PCR) based gene array containing 65 single nucleotide polymorphisms (SNPs) was analyzed. To asses the possibilities of pattern recognition techniques in detecting unfavorable genetic combinations, two approaches were analysed. A selection of these 65 SNPs formed the input both to binary logistic regression models and to self-learning artificial neural networks (ANNs). Repeated analyses showed that both methods performed equally well. Further research to improve the differentiating power of both methods should focus first on decreasing the number of otherwise indeterminable polymorphisms.