Estimation of parameters and unobserved components for nonlinear systems from noisy time series

We study the problem of simultaneous estimation of parameters and unobserved states from noisy data of nonlinear time-continuous systems, including the case of additive stochastic forcing. We propose a solution by adapting the recently developed statistical method of unscented Kalman filtering to this problem. Due to its recursive and derivative-free structure, this method minimizes the cost function in a computationally efficient and robust way. It is found that parameters as well as unobserved components can be estimated with high accuracy, including confidence bands, from heavily noise-corrupted data.

Use of the intubating laryngeal mask in children: an evaluation using video-endoscopic monitoring

BACKGROUND AND OBJECTIVE

The intubating laryngeal mask is designed to act as a ventilatory device and as an aid to blind tracheal intubation in adults. The aim of this study was to evaluate the efficacy of the intubating laryngeal mask for ventilation of the lungs and tracheal intubation in children using video-endoscopic control.

METHODS

The handling and efficacy of the size 3 intubating laryngeal mask for tracheal intubation in 80 children weighing > or = 25 kg were assessed under video-endoscopic control. Ease of intubating laryngeal mask insertion, adequacy of lung ventilation through the intubating laryngeal mask and airway sealing pressures were recorded. Tracheal intubation was performed blindly by the intubator, while the supervisor observed the procedure on the video display. If blind intubation failed at the first attempt, the monitor view was used to guide the tracheal tube into the trachea. The success rate and time required for successfully placing the tracheal tube were recorded.

RESULTS

Insertion of the intubating laryngeal mask was easy in all children. Lung ventilation through the intubating laryngeal mask was uniformly excellent. Blind tracheal intubation at the first attempt was successful in 53 children (66%) within 18.8 +/- 4.1 s. Twenty-four of the 27 failed blind intubation attempts were successfully intubated with video-endoscopic guidance within 28.6 +/- 9.4 s. Two children required replacing the intubating laryngeal mask, one child had to be intubated conventionally.

CONCLUSIONS

The size 3 intubating laryngeal mask provides an airway that is easy to establish in children > or = 25 kg with excellent ventilation conditions and allows blind tracheal intubation at the first attempt with a high success rate. Endoscopic monitoring improves its safety and intubation success rate.

[Optimized hemodilution with hydroxyethyl starch. A blood saving method in malocclusion operations]

OPTIMIZED HEMODILUTION: To avoid blood transfusions, the method of optimized hemodilution (OHD) was developed. OHD consists of preoperative hypervolemic hemodilution with 500 ml hydroxyethyl starch (HES) and consecutive volume replacement with HES in the case of intraoperative intravasal fluid loss. This study verifies the efficacy of OHD.

RESULTS

Mono- and bimaxillary operations (n = 213) performed between January 1995 and May 1999 were evaluated retrospectively with respect to reduction of blood transfusions due to application of OHD. The average blood loss in the OHD group (n = 127) was 593 ml, and no blood transfusion was required. In contrast, the control group (n = 86) showed an average blood loss of 738 ml, and 15% of these patients (n = 13) required transfusion of a total of 18 blood units. Based on our results, provision of autologous blood is only necessary for patients who do not receive OHD. The costs of fluid replacement and transfusion are increased by a factor of 5.7 when OHD is not used.

DISCUSSION

OHD should be preferably used in cases of elective surgery on patients with good cardiopulmonary health. It is an easy to handle infusion regimen, which does not require additional monitoring, has no risk of infection, and results in a significant cost reduction.

Topography of foramen ovale electrodes by 3D image reconstruction

To date, the foramen ovale (FO) electrode recording technique has been used in 234 patients at our center to assist in the evaluation of epilepsy surgery. Most of the patients suffered from mesial temporal lobe epilepsy (MTLE) and were candidates for a selective amygdalohippocampectomy. Knowledge of the exact topography of the FO electrodes is mandatory for a more precise anatomo-electro-clinical correlation of seizures and for a better understanding of FO electrode recorded electroencephalogram (EEG) and evoked potentials generated in the hippocampal formation or in nearby thalamic relays or brainstem structures. Here, we describe and illustrate a 3D image reconstruction of FO electrodes in situ as an important step to better define the generators of MTLE seizures as well as of interictal spikes and physiological EEG signals recorded with FO electrodes.

Provision of supplementary fluids to breast fed infants and later breast feeding success

It has been shown that altering hospital policies in a way to avoid interference of routine prescriptions with initiation of breast feeding and to provide active encouragement to mothers and personnel can result in significant benefit for later breast feeding success. It is less clear, however, which of the elements of a promotional programme such as UNICEF/WHO's "ten steps to successful breast feeding" are absolutely essential and which can be adapted to local cultural habits. We performed an open randomized multicenter study in Switzerland to evaluate, whether restriction of supplementary fluids for breast fed infants in the first week of life and strict avoidance of artificial teats and pacifiers affects later breast feeding success. Follow up to 6 months was ensured by mailed questionnaires. 602 mother infant pairs were enrolled. Of 294 infants in the intervention group 39% were excluded from the final analysis because of protocol violations, mainly maternal request for the use of pacifiers or bottles. Though the number of dextrin maltose supplements during the first two days (1.7 vs. 2.2 on day 1, 2.2 vs. 2.6 on day 2) and the percentage of infants receiving any supplement (85% vs. 96.6%) was significantly smaller in the intervention group, the difference was disappointingly small. The prevalence of breast feeding was 100% vs. 99% at day 5, 88% vs. 88% at 2 months, 75% vs. 71% at 4 months and 57% vs. 55% at 6 months, none of the differences being significant. We conclude that rigorous adherence to all of the ten steps may encounter obstinate resistance from cultural habits even in a population highly favourable to breast feeding. An improvement in adherence does not necessarily lead to better breast feeding success. The results of the few comparable studies in the literature show also that cultural practices during the first months of life may influence profoundly the long term effects of interventions during the first days of life.

Pressure-induced quenching of the Jahn-Teller distortion and insulator-to-metal transition in LaMnO(3)

LaMnO(3) was studied by synchrotron x-ray diffraction, optical spectroscopies, and transport measurements under pressures up to 40 GPa. The cooperative Jahn-Teller (JT) distortion is continuously reduced with increasing pressure. There is strong indication that the JT effect and the concomitant orbital order are completely suppressed above 18 GPa. The system, however, retains its insulating state to approximately 32 GPa, where it undergoes a bandwidth-driven insulator-metal transition. Delocalization of electron states, which suppresses the JT effect but is insufficient to make the system metallic, appears to be a key feature of LaMnO(3) at 20-30 GPa.

Kinetic evaluation of an i.v. bolus of MR contrast media

Currently, it is assumed that the pharmacokinetic properties of the first minutes of an I.V. MR contrast media bolus are similar to those of an i.v. iodinated contrast media bolus used in CT. Correct timing of an MRA examination is crucial for obtaining sufficient arterial contrast. This study sought to evaluate the temporal change of arterial signal intensity within 150 s after i.v. bolus injection of Gd-DTPA. Thirty consecutive patients (14 women/16 men; mean age: 51 +/- 11 years) were prospectively examined with a 1.0 Tesla clinical scanner. A single axial slice was acquired in 1.25 sec with manufacturer provided gradient echo sequence through the aorta at the level of the renal arteries. Investigation was started simultaneously to the application of contrast media (0.1 mmol/kg bodyweight Gd-DTPA at three different rates 2 mL/sec, 3 mL/sec and 4 mL/sec) and repeated for 2.5 min. An additional echo Doppler examination excluded patients with any cardiac disorders. Maximum signal (1300% increase compared to the basic value) in the aorta was achieved 20 +/- 6 sec after start of bolus injection. Then a plateau phase was maintained for the remaining investigation time (2.5 min). No significant difference was shown for different injection rates. After a bolus injection of Gd-DTPA the arterial contrast remains on a high level for at least 2 min. However, correct timing of the bolus arrival is still crucial to discriminate arteries and veins. An injection rate between 2 mL/sec and 4 mL/sec has no influence on early contrast media dynamics.

Cancellation of self-induced retinal image motion during smooth pursuit eye movements

When our eyes are tracking a target that is moving in front of a structured background, global motion of equal speed is induced in the opposite direction. This effect has been termed reafference, which, astonishingly, does not significantly affect the execution of such pursuit eye movements. Employing brief and unexpected injections of full-field motion during ongoing human smooth pursuit, we demonstrate that the sensitivity for full-field motion is reduced strongly in the direction opposite to the eye movement, i.e. the direction of reafferent background motion. Our experiments further characterize this asymmetry in visual motion processing and provide a preliminary explanation for the accuracy of the pursuit system despite self-induced motion.

Teaching and supervising tracheal intubation in paediatric patients using videolaryngoscopy

In 85 patients (mean age 4.6 +/- 3.1 years, range 0-10 years), a prototype of a new videolaryngoscope was used to provide a display of the intubation procedure in a paediatric anaesthesia teaching unit. The device provided important information to the teacher about the trainee's problems related to direct laryngoscopy, difficulties with tube insertion into the larynx as well as the need and extent for anterior laryngeal pressure. Because the videolaryngoscope did not impair the normal intubation procedure, it seems to be a safe and a benefical tool for teaching and supervising tracheal intubation in children. In addition, it provides a potentially useful aid during difficult laryngoscopy, since the monitor view of the vocal cords was significantly better compared with a direct laryngoscopic view (P < 0.001).

The effects of sevoflurane and halothane anesthesia on cerebral blood flow velocity in children

We compared cerebral blood flow velocity during anesthesia with sevoflurane and halothane in 23 children admitted for elective surgery (age, 0.4-9.7 yr; median age, 1.9 yr; ASA physical status I-II). Inhaled induction was performed in a randomized sequence with sevoflurane or halothane. Under steady-state conditions, cerebral blood flow velocity (systolic [V(s)], mean [V(mn)], and diastolic [VD]) were measured by a blinded investigator using transcranial pulsed Doppler ultrasonography. The anesthetic was then changed. CBFV measurements were repeated after washout of the first anesthetic and after steady-state of the second (equivalent minimal alveolar concentration to first anesthetic). The resistance index was calculated. VD and V(mn) were significantly lower during sevoflurane (V(mn) 1.35 m/s) than during halothane (V(mn) 1.50 m/s; P = 0.001), whereas V(s) was unchanged. The resistance index was lower during halothane (P < 0.001). Our results indicate lower vessel resistance and higher mean velocity during halothane than during sevoflurane.

IMPLICATIONS

The mean cerebral blood flow velocity is significantly decreased in children during inhaled anesthesia with sevoflurane than during halothane. This might be relevant for the choice of anesthetic in children with risk of increased intracranial pressure, neurosurgery, or craniofacial osteotomies.

[Endotracheal intubation of patients with Pierre-Robin sequence. Successful use of video intubation laryngoscope]

Patients with Pierre Robin sequence are a classic model for patients with a difficult airway. In these patients tracheal intubation may be facilitated using the lateral approach. In a 3-day-old newborn with Pierre Robin sequence, scheduled for anaesthesia to take a mould of the cleft palate, we used a video-intubation laryngoscope to give a video-display of the lateral approach intubation technique to the junior colleagues. The video-laryngoscopic view allowed detailed demonstration of the intubation procedure to all persons present. In addition, video-transmission of the laryngoscope picture enabled the attendant anaesthetist to quickly recognise the need for suctioning the hypopharynx and to coordinate the direction and extent of laryngeal pressure according the video-laryngoscopic findings. Of special value for the intuboscopist was the improved view of the cords provided by the video-display compared with direct laryngoscopy. This facilitated a non-traumatic insertion of the tube into the trachea and allowed reliable confirmation of the intubation depth by the tube marking between the vocal cords.

Video-intuboscopic monitoring of tracheal intubation in pediatric patients

PURPOSE

To evaluate the efficacy of video-intuboscopic monitoring during orotracheal intubation in a pediatric anesthesia teaching unit.

METHODS

In 100 pediatric patients direct laryngoscopy performed by residents or nurse anesthetists was monitored on a video display using a flexible, ultralight video-endoscopic system (O.D. 2.8 mm/length 1.8 m) within the endotracheal tube (ETT). Best direct laryngoscopic view was assessed by the intubator. The instructor observed the intubation procedure on the video display and noted number of intubation attempts, best laryngoscopic monitor view, tube passage through the larynx and final tube position. If required he gave instructions or corrections. After removal of the laryngoscope, tracheal tube position was adjusted using the monitor view. Difficulties attributed to the fibreoptic endoscope during intubation and removal from the ETT were recorded.

RESULTS

The trachea was successfully intubated at the first attempt in all patients. The supervisor detected one esophageal and 12 endobronchial intubations that were immediately corrected before starting ventilation. Final visualized endotracheal tube adjustment above the carina was possible in all patients except in one with copious tracheal secretions. Compared with direct laryngoscopy the video display provided an improved view of the vocal cords during intubation. In six patients direct laryngoscopy was difficult but the tracheas were intubated using the monitor view. No difficulties with the equipment were encountered except that the black coated endoscope compromised recognition of the black ETT depth-markings in small tubes.

CONCLUSIONS

Video-intuboscopic monitoring is a useful monitor of tracheal intubation. "The improved view of the cords may provide assistance during difficult laryngoscopy.

Quantitative human in vivo evaluation of high resolution MRI for vessel wall morphometry after percutaneous transluminal angioplasty

Visualization of the vessel wall after transluminal angioplasty is important to monitor the restenosis progress. Intravascular ultrasound proved its capabilities as an invasive procedure in many studies. The aim of this study was to evaluate the feasibility of high-resolution MRI as a non-invasive tool for follow-up after PTA. High-resolution magnetic resonance images (pixel size: 0.49 * 0.49 mm(2)) were acquired on a 1.0 T clinical scanner. Morphometry was conducted after conversion of DICOM images into TIFF format using ScionImage on a PC. In-vitro studies using a polyvinylchloride tube were evaluated by two independent investigators. Goldstandard was a caliper rule and direct radiography. Five patients were monitored before and 24 h, six weeks, three months and six months after PTA. In vivo measurements promised a good concordance for both investigators for area as well as for diameter measurements. Area measurements showed correlations up to r = 0.86 (p < 0.001) whereas the correlations of diameters were slightly inferior (r between 0.58 and 0.84; p < 0.005). Relocation of the same slice position in the follow up studies could be guaranteed using anatomic landmarks in the images. As a non-invasive tool to assess restenosis after PTA high-resolution MRI promises to be a reproducible technique. It is easy to identify the same vascular region in different studies due to neighboring anatomic landmarks. Progression of disease as well as success of pharmacologic treatment to prevent restenosis may be monitored.

Difficult airway management: comparison of the Bullard laryngoscope with the video-optical intubation stylet

PURPOSE

To evaluate, whether the video-optical intubation stylet (VOIS) was more successful for difficult tracheal intubation than the Bullard laryngoscope (BL).

METHODS

An intubation mannequin head was modified so that, using a Macintosh blade size 3, only the epiglottis was visible at direct laryngoscopy, representing a grade III laryngoscopic view. Forty anesthesiologists attempted tracheal intubation using each technique. Tracheal intubation with the Bullard laryngoscope was performed using the attached non-malleable intubating stylet preloaded with an endotracheal tube. The video-optical intubation stylet inserted into an endotracheal tube was used with direct laryngoscopy. During conventional laryngoscopy, the video-view from the stylet tip allowed the tracheal tube to be guided behind the epiglottis into the trachea. Ten attempts with each technique were performed by each anesthesiologist in randomized order. Intubation time, and failed intubation (> 60 sec/esophageal intubation) were recorded. The operators assessed the degree of difficulty of each method using a Likert-scale.

RESULTS

Mean intubation time (19.2+/-4.5 sec for the BL and 18.8+/-4.6 sec for the VOIS) was almost the identical. The video-optical intubation stylet was associated with fewer failed intubations (8 vs. 41; P<0.005) and had a lower degree of difficulty (1.7+/-0.65 for the VOIS and 2.6+/-0.74 for the BL; P<0.0001). No correlation was found between the anesthesiologist's experience and mean intubation time, estimated degree of difficulty or number of unsuccessful intubation.

CONCLUSION

The video-optical intubation stylet was a more effective and simpler intubation device to facilitate difficult tracheal intubation than the Bullard laryngoscope.

Hierarchical visual processing is dependent on the oculomotor system

Using functional MRI and eye movement recordings we studied the processing of hierarchical stimuli. In agreement with others, we found a minor left hemispheric dominance during local and right dominance during global processing. When attention was directed locally, well-known oculomotor cortical areas were activated, and saccades were elicited in 41% of the trials. Their latencies were similar to pro-saccades. During global processing virtually no saccades occurred. These results suggest two different operational modes of attention. Attending to local features induces a shift of attention, which simultaneously computes a saccade on any level above the brainstem with a computational burden equal to reflexive saccades. Conversely, attending to global features induces an expansion of the focus of attention, which reinforces fixation.

Linear and nonlinear time series analysis of the black hole candidate cygnus X-1

We analyze the variability in the x-ray lightcurves of the black hole candidate Cygnus X-1 by linear and nonlinear time series analysis methods. While a linear model describes the overall second order properties of the observed data well, surrogate data analysis reveals a significant deviation from linearity. We discuss the relation between shot noise models usually applied to analyze these data and linear stochastic autoregressive models. We debate statistical and interpretational issues of surrogate data testing for the present context. Finally, we suggest a combination of tools from linear and nonlinear time series analysis methods as a procedure to test the predictions of astrophysical models on observed data.

Morphology of coronary atherosclerotic lesions in patients with end-stage renal failure

BACKGROUND

An excessive rate of cardiac death is a well-known feature of renal failure. Coronary heart disease is frequent and the possibility has been raised that the natural history of the coronary plaque is different in uraemic patients. We assessed the morphology of coronary arteries in patients with end-stage renal failure and compared them with coronary arteries of matched non-uraemic control patients.

METHODS

Fifty-four cases were identified at autopsy who met the inclusion criteria: cases, end-stage renal disease (n=27); controls, non-renal patients with coronary artery disease (n=27). At autopsy all three coronary arteries were prepared at corresponding sites for investigations: (i) qualitative analysis (after Stary), (ii) quantitative measurements of intima and media thickness (by planimetry), (iii) immunohistochemical analysis of the coronary plaques and (iv) X-ray diffraction of selected calcified plaques.

RESULTS

Qualitative analysis of the coronary arteries showed significantly more calcified plaques of coronary arteries in patients with end-stage renal failure. Plaques of non-uraemic patients were mostly fibroatheromatous. Media thickness of coronary arteries was significantly higher in uraemic patients (187+/-53 microm vs 135+/-29 microm in controls) and intima thickness tended to be higher (158+/-38 microm vs 142+/-31 microm) but this difference was not statistically significant. Plaque area (4.09+/-1. 50 mm(2) vs 4.39+/-0.88 mm(2)) was comparable in both groups. Lumen area, however, was significantly lower in end-stage renal patients. Immunohistochemical analysis of the cellular infiltrate in coronary arteries showed no major differences in these advanced plaques of uraemic and non-uraemic subjects.

CONCLUSION

Coronary plaques in patients with end-stage renal failure are characterized by increased media thickness and marked calcification. In contrast to the previous opinion the most marked difference compared to non-uraemic controls does not concern the size, but the composition of the plaque. Deposition of calcium within the plaques may contribute to the high complication rate in uraemic patients.

[Effects of priming technique on onset profile of cisatracurium]

Compared to atracurium, cisatracurium releases less laudanosine and histamine, but it has a longer onset time. The primary objective of this study was a blinded, randomized comparison of intubation scores and onset times of a threefold ED 95 of cisatracurium using the priming technique with two priming substances cisatracurium itself and pancuronium. To test the effect of priming with cisatracurium or pancuronium on the onset of cisatracurium, 45 patients were anaesthetised with 0.15-0.25 mg/kg alfentanil, 0.25-0.3 mg/kg edomidate i.v. and O2/N2O, and were randomisely divided into one of three groups. After induction, 15 patients were primed with sodium chloride and thereafter received 0.15 mg/kg cisatracurium, 15 patients were primed with 0.01 mg/kg cisatracurium, another 15 patients were primed with 0.015 mg/kg pancuronium and the last two groups received 0.14 mg/kg cisatracurium three minutes later. Neuromuscular response was monitored by adductor pollicis electromyogram (EMG) by stimulating in a TOF pattern. Times for T1 reduction to 75%, 50%, 25% and 0% and T1 recovery to 25% were taken. Intubation was performed 120 seconds after the main relaxant dose and scored in four grades. The two priming groups showed a significantly faster onset of neuromuscular blockade than the control group (cisatracurium priming group: T1 = 0: 178.4 +/- 16.3 sec., pancuronium priming group 171.2 +/- 15.3 sec. vs. control group: T1 = 0: 205.5 +/- 18.9 sec.). Both primed groups showed no significantly better intubation scores, compared with the control group. Using the priming principle, cisatracurium will give good intubation scores 120 seconds after injection with a clinical duration profile comparable to an equipotent dose of atracurium.

[Postdural puncture headache: diagnosis, prevention and therapy]

Lumbar puncture (LP) is a routine technique performed for a variety of procedures, e.g. diagnosis, administration of drugs, myelography and spinal anaesthesia. Postdural puncture headache is a common complication (30-40% in diagnostic LP). Prevention can be accomplished by using small-gauge needles (< or = 25 G) or pencil-point needles (22 G). Therapy should be carried out in a stepwise approach. The first step is bedrest, use of analgetics, i.v. fluids and an adequate guidance of the patient. The second step comprises special drug therapy. Several methods of pharmacologic management have been presented in the literature, but most of these are case reports. There is a lack of large double-blind placebo-controlled studies. Theophylline, caffeine, ACTH and sumatriptan are potentially promising agents for the treatment of postdural puncture headache. The efficacy of theophylline has been proven in a double-blind and placebo-controlled study. There are a few studies and case reports reporting that caffeine p.o. and i.v. is effective in the treatment of postdural puncture headache, but recurrence of headache after caffeine therapy is frequent. ACTH acts on a complex hormonal system. The treatment with sumatriptan has been reported in only a few case reports. The third step, and one of the most effective treatments of postdural puncture headache, is the epidural blood patch. The success rate ranges between 80 and 97%.

Asymmetry in visual motion processing

The smooth pursuit system is traditionally employed using a single small target moving on a homogeneous background. It still is not fully understood, however, how accurate tracking is sustained in the presence of a structured background, which will activate global motion processing in the opposite direction as a consequence of the ongoing eye movement. To further study this interaction, we used brief shifts of a textured background injected at various times during the initiation of smooth pursuit. While shifts opposite to the target direction did not alter smooth pursuit performance, those in the same direction resulted in a marked transient perturbation of the pursuit. These results suggest a simple yet limited mechanism that adjusts the sensitivity of global motion processing.