We have compared sleep staging by an automated neural network (ANN) system, BioSleep (Oxford BioSignals) and a human scorer using the Rechtschaffen and Kales scoring system. Sleep study recordings from 114 patients with suspected obstructed sleep apnoea syndrome (OSA) were analysed by ANN and by a blinded human scorer. We also examined human scorer reliability by calculating the agreement between the index scorer and a second independent blinded scorer for 28 of the 114 studies. For each study, we built contingency tables on an epoch-by-epoch (30 s epochs) comparison basis. From these, we derived kappa (kappa) coefficients for different combinations of sleep stages. The overall agreement of automatic and manual scoring for the 114 studies for the classification {wake / light-sleep / deep-sleep / REM} was poor (median kappa = 0.305) and only a little better (kappa = 0.449) for the crude {wake / sleep} distinction. For the subgroup of 28 randomly selected studies, the overall agreement of automatic and manual scoring was again relatively low (kappa = 0.331 for {wake light-sleep / deep-sleep REM} and kappa = 0.505 for {wake / sleep}), whereas inter-scorer reliability was higher (kappa = -0.641 for {wake / light-sleep / deep-sleep / REM} and kappa = 0.737 for {wake / sleep}). We conclude that such an ANN-based analysis system is not sufficiently accurate for sleep study analyses using the R&K classification system.

Although simple in design, the home flowmeter actually shows greater accuracy than might be expected when used repeatedly to study the flow rates of men. Simple flow devices such as this could be used in conjunction with voiding diaries to give a more representative picture of patients' day-to-day voiding function.

The analysis of measurements of Electrical Bioimpedance (EBI) is on the increase for performing non-invasive assessment of health status and monitoring of pathophysiological mechanisms. EBI measurements might contain measurements artefacts that must be carefully removed prior to any further analysis. Cole model-based analysis is often selected when analysing EBI data and might lead to miss-conclusion if it is applied on data contaminated with measurement artefacts. The recently proposed Correction Function to eliminate the influence of the Hook Effect from EBI data and the fitting to the real part of the Cole model to extract the Cole parameters have been validated on experimental measurements. The obtained results confirm the feasible experimental use of these promising pre-processing tools that might improve the outcome of EBI applications using Cole model-based analysis.

The diagnostic accuracy of maximum flow rate (Q max ) for bladder outlet obstruction (BOO) is limited. 1,2 In a much-cited study, Reynard et al. 3 concluded that diagnostic accuracy can be improved by taking the maximum Q max of multiple flow rate measurements, observing that it may be more representative of the man's usual Q max . This conclusion was based on improved specificity from multiple voids. Yet this might also be achieved from a single void simply by lowering the threshold for classifying obstruction. We revisited the data published by Reynard et al. to compare diagnostic accuracy from the highest Q max of multiple measurements, with that from a single measurement of Q max . ASSESSMENT OF DIAGNOSTIC VALUE BY RECEIVER-OPERATING CHARACTERISTICTo assess diagnostic value we used receiver-operating characteristic (ROC) curves, which demonstrate the trade-off between sensitivity and specificity. The area under the ROC curve (AUC) is a measure of the diagnostic accuracy of the test. 4 The ROC is normally constructed by measuring, from experimental data on individual patients, sensitivities and specificities for the diagnosis of BOO using a range of cut-off values of Q max . This can also be calculated from the population means and standard deviations of Q max .Reynard et al. published comprehensive data on 157 men with LUTS. In order to assess the diagnostic accuracy of the various protocols (single void; highest of two voids; highest of three; and highest of four voids) we have re-plotted their data as ROC curves for each protocol in turn (Fig. 1) using the sensitivity and specificity pairs given in the article for various Q max cut-off values.The key observation is that the AUCs for the four protocols are almost identical (range 0.75-0.77), with overlapping 95% confidence intervals (Table I). This demonstrates, in contrast to the interpretation by Reynard et al., that taking the highest from multiple voids does not improve the diagnostic accuracy of Q max . DIAGNOSTIC ACCURACY OF A SINGLE MEASURE OF Q MAX FROM SUMMARY POPULATION STATISTICSAlternatively, the ROC curve for maximum flow rate can be predicted from the means and standard deviations of Q max in normal and obstructed men. The result can be seen as a measure of overlap between the flow rates of normal and obstructed men, and represents the fundamental limitation of Q max as a diagnostic test.We derived ROC curves in this way from means and standard deviations reported in two well-cited studies (Table II). We also employed the same method using the summary population statistics from Reynard et al.Notably, the diagnostic values were similar for the two well-cited studies and for Reynard et al. (Table II). Despite differing inclusion criteria and different definitions of obstruction, the AUCs never reach 0.8. We believe this represents a ceiling that fundamentally limits the diagnostic accuracy of Q max .Finally, using the data from Reynard et al., we compared the diagnostic value determined using the two alternative methods. Importantly, we observe ...