Design, construction and evaluation of systems to predict risk in obstetrics.

We present a systematic, practical approach to developing risk prediction systems, suitable for use with large databases of medical information. An important part of this approach is a novel feature selection algorithm which uses the area under the receiver operating characteristic (ROC) curve to measure the expected discriminative power of different sets of predictor variables. We describe this algorithm and use it to select variables to predict risk of a specific adverse pregnancy outcome: failure to progress in labour. Neural network, logistic regression and hierarchical Bayesian risk prediction models are constructed, all of which achieve close to the limit of performance attainable on this prediction task. We show that better prediction performance requires more discriminative clinical information rather than improved modelling techniques. It is also shown that better diagnostic criteria in clinical records would greatly assist the development of systems to predict risk in pregnancy.

Process envelopes for stabilisation/solidification of contaminated soil using lime-slag blend.

PURPOSE: Stabilisation/solidification (S/S) has emerged as an efficient and cost-effective technology for the treatment of contaminated soils. However, the performance of S/S-treated soils is governed by several intercorrelated variables, which complicates the optimisation of the treatment process design. Therefore, it is desirable to develop process envelopes, which define the range of operating variables that result in acceptable performance. METHODS: In this work, process envelopes were developed for S/S treatment of contaminated soil with a blend of hydrated lime (hlime) and ground granulated blast furnace slag (GGBS) as the binder (hlime/GGBS = 1:4). A sand contaminated with a mixture of heavy metals and petroleum hydrocarbons was treated with 5%, 10% and 20% binder dosages, at different water contents. The effectiveness of the treatment was assessed using unconfined compressive strength (UCS), permeability, acid neutralisation capacity and contaminant leachability with pH, at set periods. RESULTS: The UCS values obtained after 28 days of treatment were up to ∼800 kPa, which is quite low, and permeability was ∼10(-8) m/s, which is higher than might be required. However, these values might be acceptable in some scenarios. The binder significantly reduced the leachability of cadmium and nickel. With the 20% dosage, both metals met the waste acceptance criteria for inert waste landfill and relevant environmental quality standards. CONCLUSIONS: The results show that greater than 20% dosage would be required to achieve a balance of acceptable mechanical and leaching properties. Overall, the process envelopes for different performance criteria depend on the end-use of the treated material.