The key informant technique

The clinical outcome of 152 patients aged 65 years or over who were referred to the author's institute between August 1990 and August 1991 with certain specified gastrointestinal malignancies and acute, life-threatening abdominal conditions, were audited concurrently. Two groups were considered: patients aged 65-79 years and those over 80 years. The mortality rate within 30 days of surgery was 14 per cent in both age groups, although significantly fewer patients aged over 80 years (35 of 54) were considered suitable for surgery than in the 65-79 years age group (84 of 98) (0.01 > P > 0.001). Morbidity after operation and cost of treatment were not significantly different between the two groups. Two years after surgery 40 per cent of the patients aged over 80 years and 58 per cent of those aged 65-79 years were alive. Quality of life in these survivors was good with 85 per cent of those aged over 80 years living at home and 72 per cent fit enough to undertake light work.     

Effect of crystallinity on PVC physical properties

Crystallinity in PVC contributes significantly to the strength and resiliency of the polymer. Two types of crystallites have been described: a primary crystallite in the virgin powder state pictured as a platelet or flat needle-like lamellar crystallite, and a secondary, fringed micellar crystallite from melts or solutions. Both crystallites create loose, crosslinking networks. The secondary crystallite forms when plastisol melts are cooled or solutions are gelled. Crystallites exert a major effect on rheological properties. The thermal destruction of the primary crystallite networks in the melt phase results in a decrease in elastic modulus. Cooling from the fused state creates secondary crystallites that affect tensile and elongation.     

Artificial intelligence and symbols

The introduction of massive parallelism and the renewed interest in neural networks gives a new need to evaluate the relationship of symbolic processing and artificial intelligence. The physical symbol hypothesis has encountered many difficulties coping with human concepts and common sense. Expert systems are showing more promise for the early stages of learning than for real expertise. There is a need to evaluate more fully the inherent limitations of symbol systems and the potential for programming compared with training. This can give more realistic goals for symbolic systems, particularly those based on logical foundations.     

Distributed Sensing and Control Technology for Intelligent Civil Infrastructure Systems

Abstract: Sensing and control functions are the fundamental technologies on which many civil infrastructure control and monitoring systems are constructed. Early civil infrastructure control and monitoring systems employed a centralized computing model for sensor processing and actuator control. These systems required extensive wiring to route power, sensor, and control signals among the distributed input and output devices. Furthermore, the operation of the entire system depended on one or two computers. Abstractly, these civil infrastructure applications are a realtime data acquisition and control system. Due to the dramatic decrease in computing costs, it is now feasible to construct these real-time systems using dozens or even hundreds of distributed processors linked with a data communication network. This allows the controllers to be located adjacent to sensors and actuators, thereby reducing the need for long lengths of multiconductor wire that is subject to noise and breakage. Also, failure of a single processor does not necessarily cause the entire system to fail. This paper introduces the reader to a technology called LonWorks for constructing distributed sensing and control systems. The LonWorks technology is essentially a data communication network that links together extremely small computers called Neurons. Particular emphasis is placed on describing the communication architecture and emphasizing how the distributed Neuron communication processors are designed to ensure a truly modular and open architecture.     

Improved wetland remote sensing in Yellowstone National Park using classification trees to combine TM imagery and ancillary environmental data

The U.S. Fish and Wildlife Service uses the term palustrine wetland to describe vegetated wetlands traditionally identified as marsh, bog, fen, swamp, or wet meadow. Landsat TM imagery was combined with image texture and ancillary environmental data to model probabilities of palustrine wetland occurrence in Yellowstone National Park using classification trees. Model training and test locations were identified from National Wetlands Inventory maps, and classification trees were built for seven years spanning a range of annual precipitation. At a coarse level, palustrine wetland was separated from upland. At a finer level, five palustrine wetland types were discriminated: aquatic bed (PAB), emergent (PEM), forested (PFO), scrub-shrub (PSS), and unconsolidated shore (PUS). TM-derived variables alone were relatively accurate at separating wetland from upland, but model error rates dropped incrementally as image texture, DEM-derived terrain variables, and other ancillary GIS layers were added. For classification trees making use of all available predictors, average overall test error rates were 7.8% for palustrine wetland/upland models and 17.0% for palustrine wetland type models, with consistent accuracies across years. However, models were prone to wetland over-prediction. While the predominant PEM class was classified with omission and commission error rates less than 14%, we had difficulty identifying the PAB and PSS classes. Ancillary vegetation information greatly improved PSS classification and moderately improved PFO discrimination. Association with geothermal areas distinguished PUS wetlands. Wetland over-prediction was exacerbated by class imbalance in likely combination with spatial and spectral limitations of the TM sensor. Wetland probability surfaces may be more informative than hard classification, and appear to respond to climate-driven wetland variability. The developed method is portable, relatively easy to implement, and should be applicable in other settings and over larger extents.     

Intra-seasonal NDVI change projections in semi-arid Africa

Early warning systems (EWS) tend to focus on the identification of slow onset disasters such famine and epidemic disease. Since hazardous environmental conditions often precede disastrous outcomes by many months, effective monitoring via satellite and in situ observations can successfully guide mitigation activities. Accurate short term forecasts of NDVI could increase lead times, making early warning earlier. This paper presents a simple empirical model for making 1 to 4 month NDVI projections. These statistical projections are based on parameterized satellite rainfall estimates (RFE) and relative humidity demand (RHD). A quasi-global, 1 month ahead, 1° study demonstrates reasonable accuracies in many semi-arid regions. In Africa, a 0.1° cross-validated skill assessment quantifies the technique's applicability at 1 to 4 month forecast intervals. These results suggest that useful projections can be made over many semi-arid, food insecure regions of Africa, with plausible extensions to drought prone areas of Asia, Australia and South America.     

The role of dietary fiber in the etiology of non-insulin-dependent diabetes mellitus. The San Luis Valley Diabetes Study

Accuracy of patient positioning and dose delivery in mantle field irradiations was investigated on 29 patients treated between August 1990 and December 1991. Patients were treated in two different centers, University Hospital St. Rafa?l in Leuven and Institut Gustave Roussy in Villejuif, where different techniques and procedures were used. Measurements were performed on 341 portal films and entrance doses were measured in 518 treatment set-ups. The impact of systematic errors occurring during treatment preparation and day-to-day variations on the accuracy of treatment execution were separately analysed. Daily reproducibility, defined as the deviation from the respective mean measured value for a treatment was demonstrated to be good for both the treated volume and the delivered dose and no difference between the two techniques was shown. Comparing the successive portal films of individual patients (reproducibility of a treatment, once it has started), only small day-to-day variations are found: the SD is 3.4 mm for craniocaudal movements and 2.6 mm for lateral movements. For dose delivery very narrow distributions are obtained with SDs of, respectively, 1.5% and 1.85% for the Leuven and the Villejuif group. This suggests that the position of the patient, which is often thought as the critical point in this complex set-up can be done in a very accurate way, regardless of the position used. To assess the global accuracy of the treatment, the actually treated volume and delivered dose were compared with the planned values. Apart from reproducibility this also takes into account the whole preparatory procedure between planning and the start of the therapy (first session).(ABSTRACT TRUNCATED AT 250 WORDS)