Computer-generated Predictions Research Articles

Prediction of patient disposition: comparison of computer and human approaches and a proposed synthesis.

To compare the accuracy of computer versus physician predictions of hospitalization and to explore the potential synergies of hybrid physician-computer models. A single-center prospective observational study in a tertiary pediatric hospital in Boston, Massachusetts, United States. Nine emergency department (ED) attending physicians participated in the study. Physicians predicted the likelihood of admission for patients in the ED whose hospitalization disposition had not yet been decided. In parallel, a random-forest computer model was developed to predict hospitalizations from the ED, based on data available within the first hour of the ED encounter. The model was tested on the same cohort of patients evaluated by the participating physicians. 198 pediatric patients were considered for inclusion. Six patients were excluded due to incomplete or erroneous physician forms. Of the 192 included patients, 54 (28%) were admitted and 138 (72%) were discharged. The positive predictive value for the prediction of admission was 66% for the clinicians, 73% for the computer model, and 86% for a hybrid model combining the two. To predict admission, physicians relied more heavily on the clinical appearance of the patient, while the computer model relied more heavily on technical data-driven features, such as the rate of prior admissions or distance traveled to hospital. Computer-generated predictions of patient disposition were more accurate than clinician-generated predictions. A hybrid prediction model improved accuracy over both individual predictions, highlighting the complementary and synergistic effects of both approaches. The integration of computer and clinician predictions can yield improved predictive performance.

Production of high-resolution forest-ecosite maps based on model predictions of soil moisture and nutrient regimes over a large forested area

Forest ecosite reflects the local site conditions that are meaningful to forest productivity as well as basic ecological functions. Field assessments of vegetation and soil types are often used to identify forest ecosites. However, the production of high-resolution ecosite maps for large areas from interpolating field data is difficult because of high spatial variation and associated costs and time requirements. Indices of soil moisture and nutrient regimes (i.e., SMR and SNR) introduced in this study reflect the combined effects of biogeochemical and topographic factors on forest growth. The objective of this research is to present a method for creating high-resolution forest ecosite maps based on computer-generated predictions of SMR and SNR for an area in Atlantic Canada covering about 4.3 × 106 hectares (ha) of forestland. Field data from 1,507 forest ecosystem classification plots were used to assess the accuracy of the ecosite maps produced. Using model predictions of SMR and SNR alone, ecosite maps were 61 and 59% correct in identifying 10 Acadian- and Maritime-Boreal-region ecosite types, respectively. This method provides an operational framework for the production of high-resolution maps of forest ecosites over large areas without the need for data from expensive, supplementary field surveys.

IFIM: a database of integrated fitness information for microbial genes.

Knowledge of an organism’s fitness for survival is important for a complete understanding of microbial genetics and effective drug design. Current essential gene databases provide only binary essentiality data from genome-wide experiments. We therefore developed a new database that Integrates quantitative Fitness Information for Microbial genes (IFIM). The IFIM database currently contains data from 16 experiments and 2186 theoretical predictions. The highly significant correlation between the experiment-derived fitness data and our computational simulations demonstrated that the computer-generated predictions were often as reliable as the experimental data. The data in IFIM can be accessed easily, and the interface allows users to browse through the gene fitness information that it contains. IFIM is the first resource that allows easy access to fitness data of microbial genes. We believe this database will contribute to a better understanding of microbial genetics and will be useful in designing drugs to resist microbial pathogens, especially when experimental data are unavailable.Database URL: http://cefg.uestc.edu.cn/ifim/ or http://cefg.cn/ifim/

Reliability of Hard-and Soft-tissue Computer-generated Predictions of Orthognathic Surgical Patients using VistaDent Software

Reliability of Hard-and Soft-tissue Computer-generated Predictions of Orthognathic Surgical Patients using VistaDent Software

Mutations in yhiT enable utilization of exogenous pyrimidine intermediates in Salmonella enterica serovar Typhimurium

Mutants capable of utilizing the pyrimidine biosynthetic intermediates carbamoylaspartate and dihydroorotate for growth were derived from pyrimidine auxotrophs of Salmonella enterica serovar Typhimurium LT2. The gain-of-function phenotypes both resulted from mutations in a single gene, yhiT, the third gene of a putative four-gene operon, yhiVUTS, for which there is no homologous region in Escherichia coli. Notably, when a mutant yhiT allele was transferred to a pyrimidine-requiring E. coli strain, the transformant was then capable of using carbamoylaspartate or dihydrorotate as a pyrimidine source. The operon arrangement of the yhiVUTS genes was supported by genetic analyses and studies employing RT-PCR, coupled to the determination of the transcriptional start site using 5'-random amplification of cDNA ends (RACE). Computer-generated predictions indicated that YhiT is an integral membrane protein with 12 putative transmembrane domains typical of bacterial transport proteins. Competition experiments showed that mutant YhiT interacts with the C4-dicarboxylates succinate and malate, as well as the amino acids aspartate and asparagine. The native function of wild-type YhiT remains undetermined, but the collective results are consistent with a role as a general transporter of C4-dicarboxylates and other compounds with a similar basic structure.

The Future Role of Humans in the Weather Forecasting Process - To Provide Input to a System that Mechanically Integrates Judgmental (Human) and Automated Predictions?

There is an increasing interest in the question of what might be the appropriate future role for the human in the forecast process. It is asserted that computer-generated forecasts are unable (by themselves) to fully replicate the decision-making processes of human forecasters. Similarly, it is also asserted that human forecasters are unable (by themselves) to optimally integrate into the forecasting process, guidance from computer-generated predictions. However, there is the accepted mathematical concept that two or more inaccurate but independent predictions of the same future events may be combined to yield predictions that are, on the average, more accurate than either of them taken individually. Automated and human forecasts might be expected to bring to the table different knowledge sets, and this suggests the development of a weather forecasting system that mechanically combines human and computer-generated predictions. This paper reports on the evaluation of a knowledge based system, modified in order to mechanically combine human and computer-generated predictions. The system's output is firstly evaluated over a real-time trial of 100 days duration. The trial reveals that forecasts generated by mechanically combining the predictions explain 7.7% additional variance of weather (rainfall amount, sensible weather, minimum temperature, and maximum temperature) over that explained by the human (official) forecasts. In the light of the results of the 100-day trial, a number of minor modifications are made to the system and the trial is then continued. After 365 Day-1 to Day-7 forecasts, that is, 2555 individual predictions, the average lift in percentage variance of weather explained is 7.9% over that explained by the current official forecasts. With computer-generated forecasts unable to fully incorporate human forecasters' valuable domain and contextual knowledge, there should be a need for the human forecaster well into the future. That future role may be as an input to a system that mechanically combines human predictions with computer-generated forecasts.

Predictability of soft tissue profile changes following bimaxillary surgery in skeletal class III Chinese patients

The aim of this study was to evaluate the accuracy of soft tissue profile predictions generated by a recently developed computer program (Computer-Assisted Simulation System for Orthognathic Surgery [CASSOS] 2001, SoftEnable Technology, Ltd, Hong Kong) in Chinese skeletal Class III patients treated with bimaxillary surgery. Presurgical and posttreatment lateral cephalograms of 35 Chinese patients who had combined 1-piece Le Fort I and bilateral sagittal split osteotomies were digitized using the CASSOS 2001 program, and a cranial base superimposition was performed. The hard tissue movements were simulated on the presurgical cephalogram until good superimposition of the hard tissues on the presurgical and posttreatment cephalograms was achieved. A customized cephalometric analysis consisting of 32 linear measurements was used to analyze the differences in the soft tissue profile between the actual posttreatment results and the computer-generated predictions. Comparison of the predicted and actual changes found that 16 of the 32 soft tissue measurements were significantly different ( P <.05). Most of the significant prediction errors were observed in the upper and lower lip region. The software tended to underestimate the vertical position of both the upper and lower lip and overestimate the horizontal position of the lower lip. However, the mean differences were relatively small, with the greatest mean difference being 2 mm in the vertical position of stomium inferius. The CASSOS 2001 program produced a clinically useful prediction of soft tissue profile changes following bimaxillary surgery in skeletal Class III Chinese patients.

Computer-aided resolution of an experimental paradox in bacterial chemotaxis.

Escherichia coli responds to its environment by means of a network of intracellular reactions which process signals from membrane-bound receptors and relay them to the flagellar motors. Although characterization of the reactions in the chemotaxis signaling pathway is sufficiently complete to construct computer simulations that predict the phenotypes of mutant strains with a high degree of accuracy, two previous experimental investigations of the activity remaining upon genetic deletion of multiple signaling components yielded several contradictory results (M. P. Conley, A. J. Wolfe, D. F. Blair, and H. C. Berg, J. Bacteriol. 171:5190-5193, 1989; J. D. Liu and J. S. Parkinson, Proc. Natl. Acad. Sci. USA 86:8703-8707, 1989). For example, "building up" the pathway by adding back CheA and CheY to a gutted strain lacking chemotaxis genes resulted in counterclockwise flagellar rotation whereas "breaking down" the pathway by deleting chemotaxis genes except cheA and cheY resulted in alternating episodes of clockwise and counterclockwise flagellar rotation. Our computer simulation predicts that trace amounts of CheZ expressed in the gutted strain could account for this difference. We tested this explanation experimentally by constructing a mutant containing a new deletion of the che genes that cannot express CheZ and verified that the behavior of strains built up from the new deletion does in fact conform to both the phenotypes observed for breakdown strains and computer-generated predictions. Our findings consolidate the present view of the chemotaxis signaling pathway and highlight the utility of molecularly based computer models in the analysis of complex biochemical networks.

CLINICAL USE OF prédiction REGIONS FOR MOTION ANALYSIS

This study assesses the diagnostic capability of statistically defined prediction regions, developed by a 'bootstrap' method, for assessing the curves of angular rotation of joints in children as they walk. The prediction regions had been previously developed in the authors' laboratory from a study of 309 normal children. The goal of the present study was to determine whether these computer-generated prediction regions could be used as a screen in clinical gait analysis, to determine whether a movement falls outside the normal range of variability. Kinematic analysis of 38 consecutive children referred to the motion analysis laboratory for clinical gait assessment provided 912 curves of lower-extremity joint angle dynamics. An experienced observer first inspected the patients' curves with mean normal curves superimposed and designated the curves as normal or abnormal. The performance of the computer-generated prediction regions was judged by comparison with the experienced observer's designations. The prediction regions were found to have a high sensitivity (81%), indicating that they can be used as an initial screen to identify deficits in lower limb function.

Theoretical prediction of the Infrared and Raman spectra ofO-ethylS-2-diisopropylaminoethylmethylphosphonothiolate

We have extended our computations of the structure and of the infrared and Raman spectra of methylphosphonates and related compounds to the O-ethyl S-2-diisopropylaminoethylmethylphosphonothiolate molecule (we abbreviate the name to ESD). We have computed the optimized geometry and the vibrational infrared and Raman frequencies of ESD by means of the Guassian 92 Program Package using 6–31G* basis sets. We assign the vibrational frequencies and we correct each frequency by multiplying it with a previously derived 6–31G* correction factor. The result is a computer-generated prediction of the IR and Raman spectra of ESD. The agreement between our theoretical predictions and the experimental IR spectrum of ESD is surprisingly good. © 1994 John Wiley & Sons, Inc.

Insect Phenology and Integrated Pest Management

Success of any pest management program requires knowing when to apply a treatment. Due to year to year weather variation, the calendar method is the least precise. Plant phenology is a more precise way to time treatments but it is subject to years when insect and plant development are not well synchronized. Pheromone trapping can be very precise but few tree and shrub insects have had their pheromones analyzed or synthesized. Insect phenology, the direct relationship of insect development to weather, can be precise and has been adapted for computer application. Lower developmental threshold, degree day, cumulative degree days, model, normal temperatures, and observed temperatures, terminology used in computer-generated predictions of insect development, are defined. One computer program created by the University of Minnesota Extension Service for predicting insect development and developing custom models is briefly described.

A computer model for unconscious spread of anxiety-linked inhibition in cognitive networks.

Unconscious inhibitory processes, triggered by a potential anxiety reaction, are reviewed in the context of an emerging rapprochement between psychodynamic and cognitive approaches in experimental psychology. Conditions underlying spread of inhibitory action to other cognitive networks are first explored in three tachistoscopic experiments utilizing words posthypnotically tied to a potential anxiety, pleasure, or neutral reaction. Response times of subjects, instructed to ignore those words while naming pictures or solving anagrams as quickly as possible, reveal a highly differentiated pattern of circumstances governing likelihood of inhibitory spread from anxiety-linked words to target stimuli. Next a computer model is constructed to simulate cognitive processes from onset of display to eventual response, and the model is then tested for its fit to the empirical data. Finally, an illustrative study shows that a subset of computer-generated predictions for spread of inhibitory action is verifiable experimentally.

Procainamide in the dog: antiarrhythmic plasma concentrations after intravenous administration.

Procainamide hydrochloride was administered to ouabain-intoxicated dogs to determine an antiarrhythmic plasma concentration of procainamide. Ventricular arrhythmias were produced in dogs following intravenous injections of ouabain. After a sustained ventricular tachycardia was achieved, procainamide was administered and plasma samples collected for assay. Plasma procainamide was assayed by fluorescence polarization immunoassay. Procainamide was administered at increasingly higher constant rate infusions in order to achieve intermittent, steady-state plasma concentrations. Infusion rates were calculated on the basis of previous pharmacokinetic information. All six dogs that received procainamide converted to a normal sinus cardiac rhythm after attaining a mean plasma concentration of 33.8 micrograms/ml with a range of 48.5 micrograms/ml-25.0 micrograms/ml. It was observed that the computer-generated prediction of plasma concentrations based upon previous pharmacokinetic data produced an underestimate of the actual plasma concentrations. These data may suggest that plasma concentrations of procainamide for controlling some cardiac arrhythmias in dogs may be higher than plasma concentrations cited for human patients.

In vitro measurement and computer modelling of the diffusion of antibiotic in bone cement

The diffusion coefficient for a particular antibiotic (benzyl penicillin) in various types of bone cement has been calculated from direct measurement of the passage of radio-labelled penicillin through thin sections of cement at intervals over a period of 90 days, assuming the transport processes to be Fickian. Since the concentration gradient across the test sample is not uniform in the earlier stages of the experiment, a finite-difference approximation to the diffusion equation was used to take account of the transient behaviour in determining the coefficient. This method of analysis may also be used to predict the release, as a function of time, from the surface of antibiotic-loaded bone cement and from implanted pre-loaded PMMA beads used to provide antibiotic cover following bone or joint surgery. Our computer-generated predictions appear to agree very well with previously published figures arising from in vivo and in vitro studies and we conclude therefore, that Fick's Law accurately describes the process, even when the cement mix is inhomogeneous. We have by this means a reliable method by which the diffusion coefficients for various combinations of cement, antibiotic and other additives may be retrospectively determined and compared.

Strategy for computer-generated theoretical and quantum chemical prediction of toxicity and toxicology (and pharmacology in general)

Theoretical and quantum chemical prediction of toxicity and toxicology is even more challenging than prediction of pharmacology, which is usually a one-stage event at a target site. For toxicity and toxicology, the concept of the “toxic triggering event” was developed which then leads to the cascade of subsequent physiological events. The strategy for computer-generated predictions in this area includes as the major components chemical automated substructure and “toxicophore” identification by powerful chemical substructure searching techniques developed in Europe, geometry optimizations (desirably by ab-initio intramolecular atom class-atom class pair-pair and three-body potentials), quantum chemical calculations (desirably ab-initio, incorporating optimal strategies for such computations on large molecules) on both the toxicant and its metabolites (the structures of which were generated by computer-assisted tracing of metabolic pathways), generation of the three-dimensional electrostatic molecular potential contour maps around the toxicants and their metabolites and matching of these by reverse image holography for new compounds whose toxicity has not yet been tested experimentally against those with a known toxic mechanism, matching of intermolecular interaction maps of untested compounds with known toxicants combined with matching observed physiologic signs and symptoms with “toxic triggering events” and specific pathologies, and using the concepts of systems analysis and control theory and catastrophe theory to track both the dynamic balance of endogenous biomolecules and interactions with these biomolecules. The necessary program modules are described, and the necessary data bases (both theoretical and physiological) are identified along with the form in which they may be used most expeditiously.