PubMed Health⌕ Search

SEARCH · PubMed Health

Results for “algorithms”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 325 records · Page 18Linked to original sources
PubMed1997

Validation of the ACAS TIA/stroke algorithm.

BACKGROUND AND PURPOSE: An easily administered questionnaire and algorithm classifying transient ischemic attacks (TIAs) or strokes, and also their distribution, could be invaluable for identifying endpoints in epidemiologic studies or clinical trials of prevention and therapy of cerebral ischemia. The Asymptomatic Carotid Atherosclerosis Study (ACAS) devised a symptom-based questionnaire and algorithm for detecting events in the trial. The purpose of this study was to determine sensitivity, specificity, and agreement rates of the questionnaire and algorithm against diagnoses of a panel of cerebrovascular disease authorities. METHODS: Three hundred eighty-one men and women at eight medical centers reported symptoms of stroke, TIA, or other neurologic illness. The questionnaire was administered by trained interviewers and the responses were analyzed using the algorithm. A standardized neurologic examination was performed by a neurologist. Data were submitted to two or more external reviewers. Sensitivity, specificity, and the kappa statistic (kappa) were used to evaluate the relationship between the algorithm and the external reviewers' diagnosis. RESULTS: Of the 381 reviews, 196 were diagnosed as TIA or stroke by the external panel. The algorithm's agreement with the diagnosis of TIA or stroke was 80.1%, and kappa was 0.60. Sensitivity was 87.8%, and specificity was 71.9%. CONCLUSION: While statistical agreement rates depend on the method of sample selection, the algorithm has a high agreement with an external panel of experts and is a sensitive tool for event detection. The lower specificity indicates that careful neurologic evaluation may be required to confirm or refute events identified by the screening algorithm.

Adult↗
PubMed2004

Evaluation of three pose estimation algorithms for model-based roentgen stereophotogrammetric analysis.

Model-based roentgen stereophotogrammetric analysis (RSA) uses a three-dimensional surface model of an implant in order to estimate accurately the pose of that implant from a stereo pair of roentgen images. The technique is based on minimization of the difference between the actually projected contour of an implant and the virtually projected contour of a model of that same implant. The advantage of model-based RSA over conventional marker-based RSA is that it is not necessary to attach markers to the implant. In this paper, three pose estimation algorithms for model-based RSA are evaluated. The algorithms were assessed on the basis of their sensitivities to noise in the actual contour, to the amount of drop-outs in the actual contour, to the number of points in the actual contour and to shrinkage or expansion of the actual contour. The algorithms that were studied are the iterative inverse perspective matching (IIPM) algorithm, an algorithm based on minimization of the difference (DIF) between the actual contour and the virtual contour, and an algorithm based on minimization of the non-overlapping area (NOA) between the actual and virtual contour. The results of the simulation and phantom experiments show that the NOA algorithm does not fulfil the high accuracy that is necessary for model-based RSA. The IIPM and DIF algorithms are robust to the different distortions, making model-based RSA a possible replacement for marker-based RSA.

Algorithms↗
PubMed2006

Fast three-step phase-shifting algorithm.

We propose a new three-step phase-shifting algorithm, which is much faster than the traditional three-step algorithm. We achieve the speed advantage by using a simple intensity ratio function to replace the arctangent function in the traditional algorithm. The phase error caused by this new algorithm is compensated for by use of a lookup table. Our experimental results show that both the new algorithm and the traditional algorithm generate similar results, but the new algorithm is 3.4 times faster. By implementing this new algorithm in a high-resolution, real-time three-dimensional shape measurement system, we were able to achieve a measurement speed of 40 frames per second at a resolution of 532 x 500 pixels, all with an ordinary personal computer.

Algorithms↗
PubMed1997

An algorithm for pulmonary screening of military pilots in Israel.

BACKGROUND: Medical screening is used routinely to qualify and classify candidates for pilot training. The respiratory system assumes even greater importance owing to the increased stress of flying high-performance aircraft in a hostile environment characterized by high altitude, varying acceleration ("G" forces), and the possibility of rapid decompression. Any respiratory dysfunction may threaten the pilot's health, flight safety, and completion of the mission. Only those candidates with the highest psychophysical score are accepted to undergo special aeromedical screening. Physical suitability is an important factor in the selection and classification of candidates for flight training programs, and pulmonary function testing is central within this screening protocol. METHODS: We developed a respiratory algorithm for this unique screening process. The algorithm represents a practical and efficient approach for large-scale screening of healthy candidates for flight training. The algorithm deals with the major pulmonary health problems encountered in a previously screened healthy population aged 17 to 25 years. If by anamnesis, physical examination results, or baseline spirometry findings there is reason to suspect a respiratory problem that could emerge to endanger the pilot's life, a specially designed evaluation is performed according to the algorithm. We explain, step by step, the basis for each suggested test in order to reach a decision (operational specifications). The pulmonary function studies we recommend are reasonably priced and can be easily and reliably performed by regular medical staff technicians. The major justification for performing pulmonary function studies in a healthy population that has already gone through a preliminary medical screening and has been found fit is to identify occult or latent abnormalities. These abnormalities may have no or minimal clinical expression under ordinary circumstances but, under the stress of flight during the ensuing 5 to 10 years, may produce serious limitation in function. RESULTS: Two cases, seen in the Air Force Medical Center, are presented to illustrate how the algorithm is implemented. The algorithm has been in use for more than 5 years, and has been applied to the screening of several thousand candidates. Follow-up of the accepted candidates has not revealed any significant defects in the decision-making process. CONCLUSION: Use of the algorithm is highly cost-effective since it allows for nonspecialist physicians to carry out pulmonary screening and involves the pulmonary specialist only infrequently, ie, when a particularly complicated and/or borderline case is encountered. It is our contention that a similar algorithm would be useful in many other settings where large-scale screening is required.

Adolescent↗
PubMed2003

A spirometry-based algorithm to direct lung function testing in the pulmonary function laboratory.

OBJECTIVE: To design a spirometry-based algorithm to predict pulmonary restrictive impairment and reduce the number of patients undergoing unnecessary lung volume testing. DESIGN: Two prospective studies of 259 consecutive patients and 265 consecutive patients used to derive and validate the algorithm, respectively. SETTING: A pulmonary function laboratory of a tertiary care hospital. PATIENTS: Consecutive adults referred to the laboratory for lung volume measurements and spirometry. MEASUREMENTS: The sensitivity of the algorithm for predicting pulmonary restriction and the cost savings associated with its use. RESULTS: Total lung capacity correlated strongly with FVC (r = 0.66) and showed an inverse correlation with the FEV(1)/FVC ratio (r = - 0.41). According to the algorithm, only patients with an FVC < 85% of predicted and an FEV(1)/FVC ratio >or= 55% required lung volume measurements following spirometry. The algorithm had a high sensitivity for predicting restriction and a high negative predictive value (NPV) for excluding restriction (sensitivity, 96%; NPV, 98%). The diagnostic properties of the algorithm were reproducible in the validation study. Application of the algorithm would eliminate the need for lung volume testing in 48 to 49% of patients referred to the pulmonary function test (PFT) laboratory, reducing costs by 33%. CONCLUSIONS: A spirometry-based algorithm accurately excludes pulmonary restriction and reduces unnecessary lung volume testing in the PFT laboratory almost in half.

Algorithms↗
PubMed2007

Characteristics and value of directed algorithms in high content screening.

High content screening requires image processing algorithms that can accurately and robustly analyze large image numbers without requiring human intervention. Thus, a suite of algorithms that are directed by an understanding of the biology being studied was developed for the optimized automated acquisition and quantitation of cellular images. Two categories of directed algorithms were developed: Developer Tools for assay development and Specific Algorithms for turnkey screening of specific biological situations. The same basic sequence of analysis steps are used in these directed algorithms: 1. Primary object identification. 2. Measurement of primary object properties. 3. Identification and measurements of associated targets. 4. Analysis of raw measurements for specific biological problems. The detailed application of these steps is guided by the biology being studied and the expected phenotypic changes. Most cell biological problems to be analyzed using high content screening can be categorized by either the phenotype of the problem or labeling pattern, or by a standard biological response behavior of the cells. This enables application of directed algorithms optimized for these categories. Examples of the use of directed algorithms for specific categories are discussed, as well as the detailed analysis steps for a specific directed algorithm.

Algorithms↗
PubMed2004

Morphological detection algorithms for the automatic implantable cardioverter/defibrillator (AICD).

To prevent sudden cardiac death of patients who are at risk from long standing tachyarrhythmia the implantable cardioverter defibrillator (ICD) is the first choice therapy. ICDs use a range of electrostimuli up to defibrillation, which is a non synchronous high energy shock, whereas cardioversion is synchronous with the ECG. In order to know when and how to react, a detection algorithm, which analyses an intracardial electrocardiogram (ECG) and classifies the heart rhythm, is implemented in every ICD. All detection algorithms use the heart rate to classify the different heart rhythms roughly. If a tachycardia is detected, it is important to discriminate between a ventricular tachycardia, which is life threatening and a supraventricular tachycardia, which is much less threatening. To be able to make this distinction the detection algorithms analyse the behaviour of the heart cycle intervals, the ECG-morphology or in addition to the ventricular ECG, an atrial ECG. In this paper morphological algorithms will be evaluated and newly developed algorithms will be presented. Recent algorithms use the mathematical wavelet theory. The evaluation shows that these get better results than all but one of the simpler classical morphological algorithms. A new wavelet based algorithm, developed by the authors, exhibits the best detection results.

Algorithms↗
PubMed2001

Human performance models and rear-end collision avoidance algorithms.

Collision warning systems offer a promising approach to mitigate rear-end collisions, but substantial uncertainty exists regarding the joint performance of the driver and the collision warning algorithms. A simple deterministic model of driver performance was used to examine kinematics-based and perceptual-based rear-end collision avoidance algorithms over a range of collision situations, algorithm parameters, and assumptions regarding driver performance. The results show that the assumptions concerning driver reaction times have important consequences for algorithm performance, with underestimates dramatically undermining the safety benefit of the warning. Additionally, under some circumstances, when drivers rely on the warning algorithms, larger headways can result in more severe collisions. This reflects the nonlinear interaction among the collision situation, the algorithm, and driver response that should not be attributed to the complexities of driver behavior but to the kinematics of the situation. Comparisons made with experimental data demonstrate that a simple human performance model can capture important elements of system performance and complement expensive human-in-the-loop experiments. Actual or potential applications of this research include selection of an appropriate algorithm, more accurate specification of algorithm parameters, and guidance for future experiments.

Accidents, Traffic↗
PubMed2006

Does assessment of signs and symptoms add to the predictive value of an algorithm to rule out pregnancy?

BACKGROUND: A World Health Organization-endorsed algorithm, widely published in international guidance documents and distributed in the form of a 'pregnancy checklist', has become a popular tool for ruling out pregnancy among family planning clients in developing countries. The algorithm consists of six criteria excluding pregnancy, all conditional upon a seventh 'master criterion' relating to signs or symptoms of pregnancy. Few data exist on the specificity to pregnancy among family planning clients of long-accepted signs and symptoms of pregnancy. The aim of the present study was to assess whether reported signs and symptoms of pregnancy add to the predictive value of an algorithm to rule out pregnancy. METHODS: Data from a previous observational study were used to assess the performance of the algorithm with and without the 'signs and symptoms' criterion. The study group comprised 1852 new, non-menstruating family planning clients from seven clinics in Kenya. RESULTS: Signs and symptoms of pregnancy were rare (1.5%) as was pregnancy (1%). Signs and symptoms were more common (18.2%) among the 22 clients who tested positive for pregnancy than among the 1830 clients (1.3%) who tested negative, but did not add significantly to the predictive value of the algorithm. Most women with signs or symptoms were not pregnant and would have been unnecessarily denied a contraceptive method using the current criteria. CONCLUSIONS: The 'signs and symptoms' criterion did not substantially improve the ability of the algorithm to exclude pregnant clients, but several reasons (including use of the algorithm for intrauterine device clients) render it unlikely that the algorithm will be changed.

Algorithms↗
PubMed2004

A blood-conservation algorithm to reduce blood transfusions after total hip and knee arthroplasty.

BACKGROUND: Donation of autologous blood before total joint arthroplasty is inconvenient and costly, causes a phlebotomy-induced anemia, and may be wasteful and unnecessary for the nonanemic patient. We developed a blood-conservation algorithm that does not require predonation of autologous blood, employs selective use of epoetin alfa, and uses evidence-based transfusion criteria. Our hypothesis was that use of this algorithm would reduce the rate of transfusion after unilateral total hip and knee arthroplasty as compared with the rates described in previous reports. METHODS: We retrospectively reviewed the records of 500 consecutive patients in whom unilateral primary total hip or knee arthroplasty had been performed by a single surgeon. The same blood-conservation algorithm was recommended to all patients. Two groups of patients were identified: the first group consisted of 433 patients in whom the algorithm was followed, and the second group consisted of sixty-seven patients in whom the algorithm was not followed. RESULTS: In the group in which the algorithm was followed, the rates of allogeneic transfusion after total knee and total hip arthroplasty were 1.4% (three of 220) and 2.8% (six of 213), respectively. The overall rate of transfusion in this group was only 2.1% (nine of 433). The prevalence of transfusion in the group in which the algorithm was not followed was 16.4% (eleven of sixty-seven). This difference was significant (p = 0.0001). CONCLUSIONS: The use of this blood-conservation algorithm resulted in a significant reduction in the need for allogeneic blood transfusions after unilateral total hip and knee arthroplasty, and the results compare favorably with the rates of transfusion described in previous reports.

Aged↗
PubMed2006

Can decisional algorithms replace global introspection in the individual causality assessment of spontaneously reported ADRs?

AIM: The usefulness of algorithms for assessing the causality of suspected adverse drug reactions (ADRs) has yet to be established and, since the validation of causality algorithms depends upon their sensitivity and specificity, our study was carried out to evaluate these measures. METHOD: In this study, an expert panel assessed causality of adverse reports by using the WHO global introspection (GI) method. The same reports were independently assessed using 15 published algorithms. The causality assessment level 'possible' was considered the lower limit for a report to be considered to be drug related. For a given algorithm, sensitivity was determined by the proportion of reports simultaneously classified as drug related by the algorithm and the GI method. Specificity was measured as the proportion of reports simultaneously considered non-drug related. The analysis was performed for the total sample and within serious or unexpected events. RESULTS: Five hundred adverse reports were studied. Algorithms presented high rates of sensitivity (average of 93%, positive predictive value of 89%) and low rates of specificity (average of 7%, negative predictive value of 31%). CONCLUSION: Decisional algorithms are sensitive methods for the detection of ADRs, but they present poor specificity. A reference method was not identified. Algorithms do not replace GI and are not definite alternatives in the individual causality assessment of suspected ADRs.

Adverse Drug Reaction Reporting Systems↗
PubMed2005

A point-selection algorithm based on spatial-stiffness analysis of rigid registration.

OBJECTIVE: We propose a model of shape-based registration that leads to a task-specific algorithm for preoperatively selecting a set of model registration points. MATERIALS AND METHODS: We performed five sets of computer simulations using registration points generated by our algorithm and two noise amplification index (NAI) algorithms on the basis of the research of Simon 20. We used several different bone surface models (distal radius, proximal femur and tibia) computed from CT images of patient volunteers. The number of registration points used varied between 6 and 30. RESULTS: Our algorithm was faster than the NAI-based algorithms by factors of approximately 4 and 200. It had equal or better performance in terms of target registration error (TRE) when compared with the other algorithms. Our simulations also showed that point selection can have a large effect on TRE behavior; in particular, poor point selection does not necessarily decrease TRE as more registration points are added. CONCLUSIONS: Our point-selection algorithm produces model registration points with similar or better TRE behavior than the NAI-based algorithms we tested, and it does so with significantly less computation time.

Algorithms↗
PubMedDate not supplied

Three multizone photorefractive keratectomy algorithms for myopia. The Melbourne Excimer Laser Group.

OBJECTIVE: To compare the efficacy and complications of three different excimer laser algorithms for multizone photorefractive and photoastigmatic keratectomy. METHODS: Three different software algorithms were applied to treat myopia and myopic astigmatism with the VISX 20/20 excimer laser. Each algorithm had a maximum ablation zone of 6 mm but differed in the number of zones employed, the proportion of the total treatment allocated to each ablation zone, and the treatment of astigmatism. The Melbourne multizone technique equally divided myopia correction into a maximum of three ablation zones. The Pop multizone technique biased myopia treatment into the smaller diameter zones to a maximum of six ablation zones, with one central island pretreatment. The Alpins multizone technique equally divided myopia treatment through all zones up to a maximum of six, with one central island pretreatment. RESULTS: A total of 585 patients (780 eyes) were treated and 625 eyes (80%) were followed for more than 6 months. The mean baseline spherical equivalent refractive error was -5.63 D (-1.00 to -19.50 D). Between 71 and 79% of eyes were treated for astigmatism. There was no statistically significant differences in baseline refractive error or other characteristics among the three groups. At 6 months, the Alpins multizone algorithm had more eyes with a refractive error within +/- 1.00 D of emmetropia (p = 0.01) and more within +/- 2.00 D of emmetropia (p < 0.01). This new algorithm produced more eyes with an uncorrected visual acuity of 20/20 or better at 6 months (p < 0.01). When multiple logistic regression was used to correct for any differences in baseline myopia among the three groups, this algorithm also had a higher odds ratio for achieving 20/20 or better uncorrected visual acuity (OR = 1.58). CONCLUSION: At 6 months, all three algorithms were effective in the reduction of myopia. Significantly better visual acuity and refractive results were achieved with the Alpins multizone algorithm that spread the total treatment over a larger number of ablation zones, with an equal number of diopters of treatment in each zone.

Adolescent↗
PubMed2006

Outcome of a 4-step treatment algorithm for depressed inpatients.

OBJECTIVE: The aim of this study was to examine the efficacy and the feasibility of a 4-step treatment algorithm for inpatients with major depressive disorder. METHOD: Depressed inpatients, meeting DSM-IV criteria for major depressive disorder, were enrolled in the algorithm that consisted of sequential treatment steps (washout period, anti-depressant monotherapy, lithium addition, treatment with a nonselective monoamine oxidase inhibitor, electroconvulsive therapy). Definition of nonresponse and progression through the steps of the algorithm was dependent on the score on the 17-item Hamilton Rating Scale for Depression (HAM-D) at predefined evaluation times. Patients were admitted from April 1997 through July 2001. RESULTS: Of the 203 patients studied, 149 were treated according to the full algorithm, and 54 patients were immediately entered into step 3. Of the 203 patients, 165 (81%) achieved response (> or = 50% reduction in HAM-D score) and 101 (50%) remitted (final HAM-D score < or = 7). Of the 149 patients treated according to the full algorithm, 129 (87%) responded and 89 (60%) remitted. Twenty-four patients (16%) dropped out from the algorithm. CONCLUSION: Although response with antidepressant monotherapy was less than 50%, successive treatment according to the 4-step algorithm was very effective in a sample of depressed inpatients. The adherence to the algorithm was good as shown by a low dropout rate. This study emphasizes the importance of persisting with standardized antidepressant treatment in patients who are initially nonresponders to the first antidepressant. By the end of the study, more than 80% of the patients responded and 50% achieved full remission.

Adult↗
PubMed2002

[The use of the expectation-maximization (EM) algorithm for maximum likelihood estimation of gametic frequencies of multilocus polymorphic codominant systems based on sampled population data].

Estimation of gametic frequencies in multilocus polymorphic systems based on the numerical distribution of multilocus genotypes in a population sample ("analysis without pedigrees") is difficult because some gametes are not recognized in the data obtained. Even in the case of codominant systems, where all alleles can be recognized by genotypes, so that direct estimation of the frequencies of genes (alleles) is possible ("complete data"), estimation of the frequencies of multilocus gametes based on the data on multilocus genotypes is sometimes impossible, whether population data or even family data are used for studying genotypic segregation or analysis of linkage ("incomplete data"). Such "incomplete data" are analyzed based on the corresponding genetic models using the expectation-maximization (EM) algorithm. In this study, the EM algorithm based on the random-marriage model for a nonsubdivided population was used to estimate gametic frequencies. The EM algorithm used in the study does not set any limitations on the number of loci and the number of alleles of each locus. Locus and alleles are identified by numeration making possible to arrange loops. In each combination of alleles for a given combination of m out of L loci (L is the total number of loci studied), all alleles are assigned value 1, and the remaining alleles are assigned value 0. The sum of zeros and unities for each gamete is its gametic value (h), and the sum of the gametic values of the gametes that form a given genotype is the genotypic value (g) of this genotype. Then, gametes with the same h are united into a single class, which reduces the number of the estimated parameters. In a general case of m loci, this procedure yields m + 1 classes of gametes and 2m + 1 classes of genotypes with genotypic values g = 0, 1, 2, ..., 2m. The unknown frequencies of the m + 1 classes of gametes can be represented as functions of the gametic frequencies whose maximum likelihood estimations (MLEs) have been obtained in all previous EM procedures and the only unknown frequency (Pm(m)) that is to be estimated in the given EM procedure. At the expectation step, the expected frequencies (Fm(g) of the genotypes with genotypic values g are expressed in terms of the products of the frequencies of m + 1 classes of gametes. The data on genotypes are the numbers (ng) of individuals with genotypic values g = 0, 1, 2, 3, ..., 2m. The maximization step is the maximization of the logarithm of the likelihood function (LLF) for ng values. Thus, the EM algorithm is reduced, in each case, to solution of only one equation with one unknown parameter with the use of the ng values, i.e., the numbers of individuals after the corresponding regrouping of the data on the individuals' genotypes. Treatment of the data obtained by Kurbatova on the MNSs and Rhesus systems with alleles C, Cw, c, D, d, E, e with the use of Weir's EM algorithm and the EM algorithm suggested in this study yielded similar results. However, the MLEs of the parameters obtained with the use of either algorithm often converged to a wrong solution: the sum of the frequencies of all gametes (4 and 12 gametes for MNSs and Rhesus, respectively) was not equal to 1.0 even if the global maximum of LLF was reached for each of them (as it was for MNSs with the use of Weir's EM algorithm), with each parameter falling within admissible limits (e.g., [0, min(PN,Ps)] for PNs). The chi 2 function is suggested to be used as a goodness-of-fit function for the distribution of genotypes in a sample in order to select acceptable solutions. However, the minimum of this function only guarantee the acceptability of solutions if all limitations on the parameters are met: the sum of estimations of gametic frequencies is 1.0, each frequency falls within the admissible limits, and the "gametic algebra" is complied with (none of the frequencies is negative).

Algorithms↗
PubMed2004

Bio-ALIRT biosurveillance detection algorithm evaluation.

INTRODUCTION: Early detection of disease outbreaks by a medical biosurveillance system relies on two major components: 1) the contribution of early and reliable data sources and 2) the sensitivity, specificity, and timeliness of biosurveillance detection algorithms. This paper describes an effort to assess leading detection algorithms by arranging a common challenge problem and providing a common data set. OBJECTIVES: The objectives of this study were to determine whether automated detection algorithms can reliably and quickly identify the onset of natural disease outbreaks that are surrogates for possible terrorist pathogen releases, and do so at acceptable false-alert rates (e.g., once every 2-6 weeks). METHODS: Historic de-identified data were obtained from five metropolitan areas over 23 months; these data included International Classification of Diseases, Ninth Revision (ICD-9) codes related to respiratory and gastrointestinal illness syndromes. An outbreak detection group identified and labeled two natural disease outbreaks in these data and provided them to analysts for training of detection algorithms. All outbreaks in the remaining test data were identified but not revealed to the detection groups until after their analyses. The algorithms established a probability of outbreak for each day's counts. The probability of outbreak was assessed as an "actual" alert for different false-alert rates. RESULTS: The best algorithms were able to detect all of the outbreaks at false-alert rates of one every 2-6 weeks. They were often able to detect for the same day human investigators had identified as the true start of the outbreak. CONCLUSIONS: Because minimal data exists for an actual biologic attack, determining how quickly an algorithm might detect such an attack is difficult. However, application of these algorithms in combination with other data-analysis methods to historic outbreak data indicates that biosurveillance techniques for analyzing syndrome counts can rapidly detect seasonal respiratory and gastrointestinal illness outbreaks. Further research is needed to assess the value of electronic data sources for predictive detection. In addition, simulations need to be developed and implemented to better characterize the size and type of biologic attack that can be detected by current methods by challenging them under different projected operational conditions.

Algorithms↗
PubMed1995

A comparison of two photon planning algorithms for 8 MV and 25 MV X-ray beams in lung.

We report results of a comparison of two photon planning algorithms, the Clarkson Scatter Integration algorithm and the Equivalent Tissue-air Ratio algorithm, using a simple lung phantom for 8 MV and 25 MV X-ray beams of field sizes 5 cm x 5cm and 10 cm x 10 cm. Central axis depth-dose distributions were measured with a thimble chamber or a Markus parallel-plate chamber. Dose profile distributions were measured with TLD rods and films. Measured dose distributions were then compared to predicted dose distributions. Both agorithms overestimate the dose at mid-lung as they do not account for the effect of electronic disequilibrium. The Clarkson algorithm consistently shows less accurate results in comparison with the ETAR algorithm. There is additional error in the case of the Clarkson algorithm because of the assumption of a unit density medium in calculating scatter, which gives an overestimate in the effective scatter-air ratios in lung. For a 5 cm x 5 cm field, the error of dose prediction (Dpredicted-Dmeasured) for 25 MV x-ray beam at mid-lung is 15.8% and 12.8% for Clarkson and ETAR algorithm respectively. At 8 MV the error is 9.3% and 5.1% respectively. In addition, both algorithms underestimate the penumbral width at mid-lung as they do not account for the penumbral flaring effect in low density medium. It is very important for medical physicists, radiation therapists and clinicians to be aware of the limitation of their radiotherapy treatment planning systems.

Algorithms↗
PubMed1997

Evaluation of an algorithm for the integrated management of childhood illness in an area with seasonal malaria in the Gambia.

Most of the 12.4 million deaths occurring every year among under-5-year-olds in developing countries could be prevented by the application of simple treatment strategies. So that health professionals who have had limited training can identify and classify the common childhood diseases, WHO developed a treatment algorithm (the Integrated Management of Childhood Illness (IMCI) or Sick Child algorithm), a prototype of which was tested in 440 Gambian children aged between 2 months and 5 years. The children were first assessed by a trained field worker using the algorithm, and then by a paediatrician whose clinical diagnosis was supported by laboratory investigations and, when indicated, a chest X-ray. Compared with the paediatrician's diagnosis, the sensitivity and specificity of the draft IMCI algorithm were, respectively, 81% and 89% for the detection of pneumonia, 67% and 96% for dehydration, 87% and 8% for malaria parasitaemia (any level), 100% and 9% for malaria parasitaemia (above 5000 parasites/microliter), 100% and 99% for measles, 31% and 97% for otitis media, and 89% and 90% for malnutrition. Among the children admitted by the physician, 45% had been recommended for admission by the algorithm. Intermittent fever, chills and sweats did not help in discriminating between malaria and non-malarious fevers; shivering or shaking of the body had a sensitivity of only 35%. While the algorithm dealt with the majority of presenting complaints, the most common problems not addressed by the chart were skin rashes (21%), mouth problems (8%), and eye problems (6%). The draft IMCI algorithm proved to be effective in the diagnosis of pneumonia, gastroenteritis, measles and malnutrition, but not malaria where its use without microscopy would result in considerable over-treatment, especially in a low transmission area or during a low transmission season in countries with seasonal malaria. The current algorithm would benefit from expansion to cover management of localized infections as well as skin, mouth and eye problems.

Algorithms↗