Olafur P. Palsson

Retinal vessel oxygen saturation in healthy individuals.

PURPOSE We measured oxygen saturation in retinal vessels of healthy eyes to determine the effects of age, sex, and cardiovascular parameters, as well as the reliability of the measurements and topographic differences. METHODS The Oxymap T1 retinal oximeter is based on a fundus camera. It simultaneously captures retinal images at two different wavelengths and estimates retinal vessel oxygen saturation. Mean saturation of main retinal arterioles and venules was measured in 120 healthy individuals aged 18-80 years (median 47 years). Of the 120 participants 44 (37%) were male (49 years) and 76 (63%) female (44 years). RESULTS Oxygen saturation was 92.2 ± 3.7% (mean ± SD) in retinal arterioles and 55.6 ± 6.3% in venules. No significant difference in oxygen saturation was found between left and right eyes. The inferotemporal quadrant had lower oxygen saturation in arterioles and venules (P < 0.0001). Arteriolar oxygen saturation was stable with age. Venular oxygen saturation in males decreased by 1.9 ± 0.6% (mean ± SEM) per 10 years of age (P = 0.003) and by 0.7 ± 0.4% in females (P = 0.068). Arteriovenous (AV) difference increased by 1.5 ± 0.5% per 10 years in males (P = 0.004) and 1.0 ± 0.4% (P = 0.007) in females. For every 10 mm Hg increase in ocular perfusion pressure, oxygen saturation in arterioles increased by 0.9 ± 0.4% (P = 0.024) and in venules by 1.2 ± 0.7% (P = 0.075). CONCLUSIONS Retinal arteriolar oxygen saturation is stable in healthy individuals, while there is a significant decrease in venular oxygen saturation with age in males and a similar trend in females. AV difference increases significantly with age for both sexes. Our study provided normative data for spectrophotometric retinal oximetry in the Caucasian population.

Generalized predictive control for non-stationary systems

This paper shows how the generalized predictive control (GPC) can be extended to non-stationary (time-varying) systems. If the time-variation is slow, then the classical GPC can be used in context with an adaptive estimation procedure of a time-invariant ARIMAX model. However, in this paper prior knowledge concerning the nature of the parameter variations is assumed available. The GPC is based on the assumption that the prediction of the system output can be expressed as a linear combination of present and future controls. Since the Diophantine equation cannot be used due to the time-variation of the parameters, the optimal prediction is found as the general conditional expectation of the system output. The underlying model is of an ARMAX-type instead of an ARIMAX-type as in the original version of the GPC (Clarke, D. W., C. Mohtadi and P. S. Tuffs (1987). Automatica, 23, 137–148) and almost all later references. This implies some further modifications of the classical GPC.

Stochastic heat storage problem—solved by the progressive hedging algorithm

Abstract In the present study, a problem concerning operation of a heat storage tank in connection with a combined heat and power (CHP) plant is considered. The heat storage is used to supply the district heating system when the CHP plant is producing electric power alone and also to redistribute optimally over time the required heat production. Stochasticity is assumed attached to the future power production, and it is assumed that accurate predictions of the future heat consumption are available. A receding horizon idea is used. The problem is solved by the Progressive Hedging Algorithm (PHA), a recently developed method to deal with multi-period optimization problems under uncertainty. The application of the method is explained in detail.

On flow and supply temperature control in district heating systems

This paper discusses how the control of the flow and the supply temperature in district heating systems can be optimized, utilizing stochastic modelling, prediction and control methods. The main objective is to reduce heat production costs and heat losses in the transmission and distribution net by minimizing the supply temperature at the district heating plant. This control strategy is reasonable, in particular, if the heat production takes place at a combined heat and power (CHP) plant. The control strategy is subject to some restrictions, e.g. that the total heat requirement for all consumers is supplied at any time, and each individual consumer is guaranteed some minimum supply temperature at any time. Another important restriction is that the variation in time of the supply temperature is kept as small as possible. This concept has been incorporated in the program package, PRESS, developed at the Technical University of Denmark. PRESS has been applied and tested, e.g. at Vestkraft in Esbjerg, Denmark, and significant saving potentials have been documented. PRESS is now distributed by the Danish District Heating Association.

Individualised risk assessment for diabetic retinopathy and optimisation of screening intervals: a scientific approach to reducing healthcare costs

Objective To validate a mathematical algorithm that calculates risk of diabetic retinopathy progression in a diabetic population with UK staging (R0–3; M1) of diabetic retinopathy. To establish the utility of the algorithm to reduce screening frequency in this cohort, while maintaining safety standards. Research design and methods The cohort of 9690 diabetic individuals in England, followed for 2 years. The algorithms calculated individual risk for development of preproliferative retinopathy (R2), active proliferative retinopathy (R3A) and diabetic maculopathy (M1) based on clinical data. Screening intervals were determined such that the increase in risk of developing certain stages of retinopathy between screenings was the same for all patients and identical to mean risk in fixed annual screening. Receiver operating characteristic curves were drawn and area under the curve calculated to estimate the prediction capability. Results The algorithm predicts the occurrence of the given diabetic retinopathy stages with area under the curve =80% for patients with type II diabetes (CI 0.78 to 0.81). Of the cohort 64% is at less than 5% risk of progression to R2, R3A or M1 within 2 years. By applying a 2 year ceiling to the screening interval, patients with type II diabetes are screened on average every 20 months, which is a 40% reduction in frequency compared with annual screening. Conclusions The algorithm reliably identifies patients at high risk of developing advanced stages of diabetic retinopathy, including preproliferative R2, active proliferative R3A and maculopathy M1. Majority of patients have less than 5% risk of progression between stages within a year and a small high-risk group is identified. Screening visit frequency and presumably costs in a diabetic retinopathy screening system can be reduced by 40% by using a 2 year ceiling. Individualised risk assessment with 2 year ceiling on screening intervals may be a pragmatic next step in diabetic retinopathy screening in UK, in that safety is maximised and cost reduced by about 40%.

Estimation of the Length Distribution of Marine Populations in the Gaussian-multinomial Setting using the Method of Moments

Abstract In this paper, the problem of estimation of the length distribution of marine populations in the Gaussian-multinomial model is considered. For the purpose of the mean and covariance parameter estimation, the method of moments estimators are developed. That is, minimum variance linear unbiased estimator for the mean frequency vector is derived and a consistent estimator for the covariance matrix of the length observations is presented. The usefulness of the proposed estimators is illustrated with an analysis of real cod length measurement data.

Application of Predictive Control in District Heating Systems

In district heating systems, and in particular if the heat production takes place at a combined heat and power (CHP) plant, a reasonable control strategy is to keep the supply temperature from the district heating plant as low as possible. However, the control is subject to some restrictions, for example, that the total heat requirement for all consumers is supplied at any time and each individual consumer is guaranteed some minimum supply temperature at any time. A lower supply temperature implies lower heat loss from the transport and the distribution network, and lower production costs. A district heating system is an example of a non-stationary system, and the model parameters have to be time varying. Hence, the classical predictive control theory has to be modified. Simulation experiments are performed in order to study the performance of modified predictive controllers. The systems are, however, described by transfer function models identified from real data.

Predictor-Based Optimal Control of Supply Temperature in District Heating Systems

Abstract A modified version of the linear quadratic Gaussian controller is presented, which is build upon the prediction form of the model. This implies that the controller is more capable of handling non-stationarities, like time-varying model parameters, than the classical type of linear quadratic Gaussian controller, which usually is based on the solution of the Diophantine or the Riccati Equation (in the state-space case). The linear quadratic Gaussian controller is used in a simulation study together with a transfer function model, with time-varying parameters, that describes the relations between supply temperature of the water from a district heating plant and the supply temperature at specific locations in the distribution network. The simulation results show that the variance of the differenced control signal can be reduced drastically without affecting the performance of the controller significantly.

Estimation of extreme mass flowrate in district heating systems

Abstract This paper focuses on how extreme mass flowrate conditions, i.e. T year events, in district heating systems can be estimated by extreme value analysis, namely the method based on annual maximum values and the peaks over threshold (POT) method. Only a few years of mass flowrate data are available, which is far too short a time to be of any practical use for extreme analysis. However, 42 years of climate data are available, which are used to generate daily mass flow data using an autoregressive external (ARX) model. The simulated mass flow contains uncertainty which must be considered in the analysis. This is not possible using a traditional analysis without some modifications. Furthermore, it is also of interest to consider the uncertainty in the climate. A bootstrap technique is applied in order to answer these questions and a comparison is made of the traditional analysis and the bootstrap approach. The results show that when using the bootstrap method the estimated T year events and their uncertainties are always higher compared to traditional methods, whether based on the annual maximum or the POT method. The difference is, however, quite moderate in most cases. The POT method yields somewhat higher estimates than the annual maximum method. It turns out that the predicted load in severe cold spells is not drastically higher compared to, for example, what has been experienced in Reykjavik over the last ten years when the climate has been quite normal. For instance, the 100 year event is estimated to be approximately 20 per cent higher than the average maximum mass flowrate during the last ten years.

Online Fouling Detection of Domestic Hot Water Heat Exchangers

Due to hardness of cold water supply in many countries, there is a risk of fouling in domestic hot water (DHW) counterflow plate heat exchangers. The scaling will result in increased resistance to heat transfer, which has negative effects on the economics of the district heating network. A common approach is to clean or change the heat exchanger periodically, which can be expensive if only limited fouling has occurred (unnecessary) or if a higher than expected scaling layer has formed (inefficiency). A better approach is to monitor the state of the heat exchangers and clean them when actually required. This would result in more energy-efficient operation and provide an optimum schedule for heat exchanger cleaning. This can be simple if the heat exchangers are operating under steady-state conditions; however, if large variations in the inlets are experienced, as is the case with the mass flows in DHW heat exchangers, it quickly becomes impossible with standard methods. In this paper it is proposed to monitor the state of the heat exchanger online by using measurements that are easily obtainable under normal operation and applying fast mathematical models to estimate the overall heat transfer coefficient of the heat exchanger. The results show that the methods proposed can be used to detect fouling in DHW heat exchangers.