An Incremental EM-based Learning Approach for On-Line Prediction of Hospital Resource Utilization

Objective: Inpatient length of stay (LOS) is an important measure of hospital activity, health care resource consumption, and patient acuity. This research work aims at developing an incremental expectation maximization (EM) based learning approach on mixture of experts (ME) system for on-line prediction of LOS. The use of a batchmode learning process in most existing artificial neural networks to predict LOS is unrealistic, as the data become available over time and their pattern change dynamically. In contrast, an on-line process is capable of providing an output whenever a new datum becomes available. This on-the-spot information is therefore more useful and practical for making decisions, especially when one deals with a tremendous amount of data. Methods and material: The proposed approach is illustrated using a real example of gastroenteritis LOS data. The data set was extracted from a retrospective cohort study on all infants born in 1995-1997 and their subsequent admissions for gastroenteritis. The total number of admissions in this data set was n = 692. Linked hospitalization records of the cohort were retrieved retrospectively to derive the outcome measure, patient demographics, and associated co-morbidities information. A comparative study of the incremental learning and the batch-mode learning algorithms is considered. The performances of the learning algorithms are compared based on the mean absolute difference (MAD) between the predictions and the actual LOS, and the proportion of predictions with MAD < 1 day (Prop(MAD < 1)). The significance of the comparison is assessed through a regression analysis. Results: The incremental learning algorithm provides better on-line prediction of LOS when the system has gained sufficient training from more examples (MAD = 1.77 days and Prop(MAD < 1) = 54.3%), compared to that using the batch-mode learning. The regression analysis indicates a significant decrease of MAD (p-value = 0.063) and a significant (p-value = 0.044) increase of Prop(MAD

Retention, adherence and compliance: Important considerations for home- and group-based resistance training programs for older adults

Reports on the efficacy of physical activity intervention trials usually only include discussion of the primary outcomes. However, assessing factors such as participant retention, adherence and compliance can assist in the accurate interpretation of the overall impact of a program in terms of reach and appeal. A quasi-randomised trial was carried out to assess and compare retention and adherence rates, and compliance with, a twice weekly resistance training program provided either individually at home or in a group format. Retirement villages (n=6) were assigned to either 'Have A Try' (HAT, home-based) or 'Come Have A Try' (CHAT, group-based); both programs included nine strength and two balance exercises. The program involved a 20-week Intervention Phase a 24-week Maintenance Phase and a 20-week On-going Maintenance Phase. One hundred and nineteen participants (mean age 80 +/- 6 years) were recruited (HAT = 38, CHAT = 81). There was no difference in retention rates at the end of the Intervention Phase, but significantly more HAT than CHAT participants had dropped out of the study (p < 0.01) after the Maintenance Phase and the On-going Maintenance Phase. During the Intervention Phase, over half the HAT and CHAT participants completed >= 75% of the prescribed activity sessions, but adherence was significantly greater in CHAT than HAT during the Maintenance Phase (p < 0.01). Participants in CHAT were significantly more compliant than HAT participants (p < 0.05). Both home- and group-based formats were successful over the short-term, but, in retirement villages, the group program had better adherence and compliance in the longer-term. (c) 2006 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved.

Community center-based resistance training for the maintenance of glycemic control in adults with type 2 diabetes

OBJECTIVE - The purpose of this study was to determine whether beneficial effects on glycemic control of an initial laboratory-supervised resistance training program could be sustained through a community center-based maintenance program. RESEARCH DESIGN AND METHODS - We studied 57 overweight (BMI >= 27 kg/m(2)) sedentary men and women aged 40-80 years with established (> 6 months) type 2 diabetes. initially, all participants attended a twice-weekly 2-month supervised resistance training program conducted in the exercise laboratory. Thereafter, participants undertook a resistance training maintenance program (2 times/week) for 12 months and were randomly assigned to carry this out either in a community fitness and recreation center (center) or in their domestic environment (home). Glycemic control (HbA(1c) [A1C]) was assessed at 0, 2, and 14 months. RESULTS - Pooling data from the two groups for the 2-month supervised resistance training program showed that compared with baseline, mean A1C fell by -0.4% [95% CI -0.6 to -0.2]. Within-group comparisons showed that A I C remained lower than baseline values at 14 months in the center group (- 0.4% [-0.7 to -0.03]) but not in the home group (-0.1% [-0.4 to 0.3]). However, no between-group differences were observed at each time point. Changes in AIC during the maintenance period were positively associated with exercise adherence in the center group only. CONCLUSIONS - Center-based but not home-based resistance training was associated with the maintenance of modestly improved glycemic control from baseline, which was proportional to program adherence. Our findings emphasize the need to develop and test behavioral methods to promote healthy lifestyles including increased physical activity in adults with type 2 diabetes.

Numerical method for determination of the NMR frequency of the single-qubit operation in a silicon-based solid-state quantum computer

A numerical method is introduced to determine the nuclear magnetic resonance frequency of a donor (P-31) doped inside a silicon substrate under the influence of an applied electric field. This phosphorus donor has been suggested for operation as a qubit for the realization of a solid-state scalable quantum computer. The operation of the qubit is achieved by a combination of the rotation of the phosphorus nuclear spin through a globally applied magnetic field and the selection of the phosphorus nucleus through a locally applied electric field. To realize the selection function, it is required to know the relationship between the applied electric field and the change of the nuclear magnetic resonance frequency of phosphorus. In this study, based on the wave functions obtained by the effective-mass theory, we introduce an empirical correction factor to the wave functions at the donor nucleus. Using the corrected wave functions, we formulate a first-order perturbation theory for the perturbed system under the influence of an electric field. In order to calculate the potential distributions inside the silicon and the silicon dioxide layers due to the applied electric field, we use the multilayered Green's functions and solve an integral equation by the moment method. This enables us to consider more realistic, arbitrary shape, and three-dimensional qubit structures. With the calculation of the potential distributions, we have investigated the effects of the thicknesses of silicon and silicon dioxide layers, the relative position of the donor, and the applied electric field on the nuclear magnetic resonance frequency of the donor.

Computer animation of hot spot development in bulk coal as an aid for training coal miners

The processes that take place during the development of a heating are difficult to visualise. Bulk coal self-heating tests at The University of Queensland (UQ) using a two-metre column are providing graphic evidence of the stages that occur during a heating. Data obtained from these tests, both temperature and corresponding off-gas evolution can be transformed into what is effectively a video-replay of the heating event. This is achieved by loading both sets of data into a newly developed animation package called Hotspot. The resulting animation is ideal for spontaneous combustion training purposes as the viewer can readily identify the different hot spot stages and corresponding off-gas signatures. Colour coding of the coal temperature, as the hot spot forms, highlights its location in the coal pile and shows its ability to migrate upwind. An added benefit of the package is that once a mine has been tested in the UQ two-metre column, there is a permanent record of that particular coals performance for mine personnel to view.

Using the Amazon metric to construct an image database based on what people do, not what they say

Current image database metadata schemas require users to adopt a specific text-based vocabulary. Text-based metadata is good for searching but not for browsing. Existing image-based search facilities, on the other hand, are highly specialised and so suffer similar problems. Wexelblat's semantic dimensional spatial visualisation schemas go some way towards addressing this problem by making both searching and browsing more accessible to the user in a single interface. But the question of how and what initial metadata to enter a database remains. Different people see different things in an image and will organise a collection in equally diverse ways. However, we can find some similarity across groups of users regardless of their reasoning. For example, a search on Amazon.com returns other products also, based on an averaging of how users navigate the database. In this paper, we report on applying this concept to a set of images for which we have visualised them using traditional methods and the Amazon.com method. We report on the findings of this comparative investigation in a case study setting involving a group of randomly selected participants. We conclude with the recommendation that in combination, the traditional and averaging methods would provide an enhancement to current database visualisation, searching, and browsing facilities.