SEARCH · PubMed Health

Results for “Multiple Classification Analysis”

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 847 records · Page 47Linked to original sources

PubMed1993

Poor prognosis associated with elevated serum CA 19-9 level in advanced colorectal carcinoma, independent of DNA ploidy or SPF.

DNA ploidy, S-phase fraction (SPF) for the tumours, serum tumour markers such as carcinoembryonic antigen (CEA) and serum CA 19-9 and major clinical parameters were analysed as prognostic factors in 105 patients with advanced colorectal carcinoma. All 105 were treated with a three-drug schedule including low dose epirubicin and sequential methotrexate, 5-fluorouracil, followed by leucovorin rescue. In univariate analysis, gender, Karnofsky index, extent of metastases, presence of abdominal metastases, CEA and CA 19-9 correlated with survival. Age, presence of liver or of lung metastases, DNA ploidy or SPF were not significantly associated with survival. In stepwise multivariate analysis an elevated serum CA 19-9 level, a poor Karnofsky index and multiple sites of metastases were independent adverse prognostic factors. Based on the multivariate analysis, patients were grouped in three categories. Group 1 consisted of 32 patients with Karnofsky > or = 80, with a normal serum CA 19-9 level and a single site of metastases. Group 2 consisted of 48 patients with Karnofsky > or = 80 and with an elevated serum CA 19-9 level or multiple sites of metastases. Group 3 consisted of 14 patients with Karnofsky < or = 70. This classification gave a highly significant correlation with survival (chi 2 = 45.52, P < 0.001, log rank test). The median survival in group 1, group 2 and group 3 was 30.1 months, 13.5 months and 3.9 months, respectively. Based on these results we suggest that trials involving advanced colorectal cancer should include the measurement of serum CA 19-9 levels as one of the most important prognostic factors, but also include documentation of other independent prognostic factors.

Antigens, Tumor-Associated, Carbohydrate↗

PubMed1985

Tritiated imipramine binding distinguishes among subtypes of depression.

We studied 45 depressed patients and 20 healthy controls in order to determine if tritiated imipramine binding distinguished among subtypes of primary major depressive disorder. Mean (+/- standard deviation) values for maximal concentration of tritiated imipramine binding sites on platelet membranes were significantly lower in patients with bipolar and familial pure depressive disease (754 +/- 149 and 870 +/- 241 femtomoles [fmole]/mg of protein, respectively) than in patients with depressive spectrum and sporadic depressive disease (1,236 +/- 241 and 1,188 +/- 325 fmole/mg of protein, respectively), neither of which differed from healthy controls (1,238 +/- 201 fmole/mg of protein). Multiple linear regression analysis revealed that these differences could not be attributed to differences in age, sex, Hamilton Rating Scale score, presence of psychotic features, hospitalization status, or medication history. This association of a biological finding with distinct clinically defined subtypes of depression may lead to a classification of affective disorders useful in further research.

Adult↗

PubMed2005

Application of a human error framework to conduct train accident/incident investigations.

Accident/incident investigations are an important qualitative approach to understanding and managing transportation safety. To better understand potential safety implications of recently introduced remote control locomotive (RCL) operations in railroad yard switching, researchers investigated six railroad accidents/incidents. To conduct the investigations, researchers first modified the human factors analysis and classification system (HFACS) to optimize its applicability to the railroad industry (HFACS-RR) and then developed accident/incident data collection and analysis tools based on HFACS-RR. A total of 36 probable contributing factors were identified among the six accidents/incidents investigated. Each accident/incident was associated with multiple contributing factors, and, for each accident/incident, active failures and latent conditions were identified. The application of HFACS-RR and a theoretically driven approach to investigating accidents/incidents involving human error ensured that all levels of the system were considered during data collection and analysis phases of the investigation and that investigations were systematic and thorough. Future work is underway to develop a handheld software tool that incorporates these data collection and analysis tools.

Accidents↗

PubMed1990

The validity of diagnostic systems for common mental disorders: a comparison between the ID-CATEGO and the DSM-III systems.

It is argued that latent trait analysis provides a way of examining the construct validity of diagnostic concepts which are used to categorize common mental illnesses. The present study adds two additional aspects of validity using multiple discriminant analysis applied to two widely used taxonomic systems. Scales of anxiety and depression derived from previous latent trait analyses are applied to individuals reaching criteria for 'caseness' on the ID-CATEGO system and the DSM-III system, both at initial diagnosis and six months later. The first multiple discriminant analysis is carried out on the initial scale scores, and the results are interpreted in terms of concurrent validity. The second analysis uses improvement scores on the two scales and relates to predictive validity. It is argued that the ID-CATEGO system provides a better classification for common mental illnesses than the DSM-III system, since it allows a better discrimination to be made between anxiety and depressive disorders.

Adjustment Disorders↗

PubMed2001

Assessment of absolute risk of death after myocardial infarction by use of multiple-risk-factor assessment equations: GISSI-Prevenzione mortality risk chart.

AIMS: To present and discuss a comprehensive and ready to use prediction model of risk of death after myocardial infarction based on the very recently concluded follow-up of the large GISSI-Prevenzione cohort and on the integrated evaluation of different categories of risk factors: those that are non-modifiable, and those related to lifestyles, co-morbidity, background, and other conventional clinical complications produced by the index myocardial infarction. METHODS: The 11-324 men and women recruited in the study within 3 months from their index myocardial infarction have been followed-up to 4 years. The following risk factors have been used in a Cox proportional hazards model: non-modifiable risk factors: age and sex; complications after myocardial infarction: indicators of left ventricular dysfunction (signs or symptoms of acute left ventricular failure during hospitalization, ejection fraction, NYHA class and extent of ventricular asynergy at echocardiography), indicators of electrical instability (number of premature ventricular beats per hour, sustained or repetitive arrhythmias during 24-h Holter monitoring), indicators of residual ischaemia (spontaneous angina pectoris after myocardial infarction, Canadian Angina Classification class, and exercise testing results); cardiovascular risk factors: smoking habits, history of diabetes mellitus and arterial hypertension, systolic and diastolic blood pressure, blood total and HDL cholesterol, triglycerides, fibrinogen, leukocytes count, intermittent claudication, and heart rate. Multiple regression modelling was assessed by receiver operating characteristic (ROC) analysis. Generalizability of the models was assessed through cross validation and bootstrapping techniques. POPULATION AND RESULTS: During the 4 years of follow-up, a total of 1071 patients died. Age and left ventricular dysfunction were the most relevant predictors of death. Because of pharmacological treatments, total blood cholesterol, triglycerides, and blood pressure values were not significantly associated with prognosis. Sex-specific prediction equations were formulated to predict risk of death according to age, simple indicators of left ventricular dysfunction, electrical instability, and residual ischaemia along with the following cardiovascular risk factors: smoking habits, history of diabetes mellitus and arterial hypertension, blood HDL cholesterol, fibrinogen, leukocyte count, intermittent claudication, and heart rate. The predictive models produced on the basis of information available in the routine conditions of clinical care after myocardial infarction provide ready to use and highly discriminant criteria to guide secondary prevention strategies. CONCLUSIONS AND IMPLICATIONS: Besides documenting what should be the preferred and practicable focus of clinical attention for today's patients, the experience of GISSI-Prevenzione suggests that periodically and prospectively collected databases on naturalistic' cohorts could be an important option for updating and verifying the impact of guidelines, which should incorporate the different components of the complex profile of cardiovascular risk. The GISSI Prevenzione risk function is a simple tool to predict risk of death and to improve clinical management of subjects with recent myocardial infarction. The use of predictive risk algorithms can favour the shift from medical logic, based on the treatment of single risk factors, to one centred on the patient as a whole as well as the tailoring of medical interventions according to patients' overall risk.

Adult↗

PubMed1992

The excretion of 17-ketosteroids and 17-hydroxycorticosteroids in night urine of elite rowers during altitude training.

During an altitude training camp (23 days, 1850 m above sea level) we collected night urine from 36 German elite rowers in order to measure the excretion of urea, creatinine, 17-ketosteroids (17-KS), and 17-hydroxycorticosteroids (17-OHCS). 17-KS and 17-OHCS represent the major metabolites from endogenous anabolic and catabolic steroid hormone systems. There was no significant change in the excretion of 17-OHCS during the training camp (mean value: females 0.0025 +/- 0.0011, males 0.0033 +/- 0.0012 mg/kgbw.h). Significant increases in the excretion of 17-KS (mean value: females 0.007 +/- 0.003, males 0.0092 +/- 0.0039 mg/kgbw.h) and in the ratio 17-KS/17-OHCS (mean value: females 3.13 +/- 1.65, males 3.11 +/- 1.71) were found during the first week and towards the end of the training camp. To investigate training effects on the excretion of these hormone metabolites, we used a recently described classification for rowing training (categories I to IV of rowing-specific training according to lactate levels and non-rowing specific categories) and a corresponding rowing data base. More than 80% of training was performed at a lactate level lower than 2 mmol/l. Using multiple regression analysis, the general finding was that in males the ratio of 17-KS/17-OHCS increased with rowing specific and non specific training regimes where a lactate level below 2 mmol/l was observed. Training at higher lactate levels caused a decrease in the ratio that may be interpreted as a shift in the production from endogenous androgenic steroid hormones to cortisol. No significant effects of single training variables were found in female rowers, which indicates major training influences on testosterone metabolism. The influence of further factors such as a relationship between urine urea and 17-KS in males are described and need further explanation.

17-Hydroxycorticosteroids↗

PubMed1999

Widespread pain and neck pain in schoolchildren. A prospective one-year follow-up study.

The aims of this study were to prospectively follow up population-based cohorts of children with widespread pain, children with neck pain and pain-free children, in order to evaluate 1-y changes in pain symptoms and to evaluate predictors for persistent widespread pain and for the change of neck pain to widespread pain. A structured pain questionnaire, the Children's Depression Inventory, and a sleep questionnaire were completed by the pre-adolescent cohorts, and clinical evaluation with tender point palpation and pain threshold measurements was carried out in both years. The Child Behavior Checklist, the Teacher's Report Form and a sociodemographic questionnaire were completed at baseline. More children in the 2 pain cohorts reported pain at follow-up than did controls (p < 0.0001). Children with persistent widespread pain had lower pain thresholds compared with those whose pain classification changed. In the neck pain group, 19 (20.4%) reported widespread pain at follow-up. Although depressive symptoms and sleep problems were associated with the change of neck pain to widespread pain, neither they nor other measured factors showed independent explanatory power in multiple logistic regression analysis. In conclusion, fluctuation of pain symptoms occurred mainly among pain cohorts. Depressive and sleep problems may have an effect on the spreading of regional neck pain to widespread pain. Pain threshold measurement and evaluation of depressive and sleep problems may be useful tools for secondary prevention of musculoskeletal pain in pre-adolescents.

Adolescent↗

PubMedDate not supplied

Objective assessment of dyskinesia in children with cerebral palsy.

The clinical classification of children with cerebral palsy is limited by multiple factors. Distinguishing between spasticity and dyskinesia is critical, because the outcome after standard orthopaedic and neurosurgical interventions is less predictable in children with cerebral palsy who have a significant dyskinetic component. This study applied computer-based analysis of gait to assess objectively the presence of significant dyskinesia in children with cerebral palsy. Three-dimensional gait analysis was performed on 18 normal children, 17 children with principally spastic cerebral palsy, and 23 children with significantly dyskinetic cerebral palsy. Children were assigned to the spastic or dyskinetic groups prospectively, based on clinical analysis by an experienced physician and physical therapist. The children with dyskinesia were found to have a significantly wider, and more variable normalized dynamic base of support, a smaller step profile (step length divided by step width), and a greater and more variable maximal lateral acceleration than the spastic and normal groups (mixed model analysis of variance, p = 0.0001). A predictive model of dyskinesia, (developed by logistic regression analysis), using these gait parameters, exhibited excellent sensitivity, correctly classifying 20 (87%) of 23 children as dyskinetic. This study shows that children with dyskinetic cerebral palsy have distinct gait parameters and that objective assessment of dyskinesia in children with cerebral palsy is possible with computer-based analysis of gait.

Adolescent↗

PubMed1998

Valve Resistance in Mitral Stenosis: Its Determinants and its Role in the Evaluation of the Disease.

To evaluate the value and the determinants of valve resistance in mitral stenosis, 95 patients with pure mitral stenosis were examined by Doppler echocardiography during their clinical follow-up, measuring cavity dimensions, left ventricular function, mitral area (by planimetry and pressure half time), mean transmitral pressure gradient, aortic flow, and pulmonary artery systolic pressure. The mitral resistance was calculated as mean transmitral pressure gradient/aortic flow ratio. To graduate the severity of the morphological abnormalities in valvular structure, we used a point score system with evaluation of leaflet and subvalvular thickness, calcification, and valvular mobility. The functional class was determined according to NYHA classification. In this study, both mitral area (r = -0.79, P < 0.001 and r(p) = -0.60, P < 0.001) and mitral score (r = 0.68, P < 0.001 and r(p) = 0.25, P = 0.013) were independent determinants of mitral resistance. In multivariate analysis, mitral resistance and female gender were selected by multiple linear regression analysis as determinants of pulmonary artery systolic pressure, and mitral area and pulmonary artery systolic pressure were selected by logistic linear regression analysis as determinants of NYHA functional class. In patients with moderate or severe mitral stenosis, the estimated probability for III and IV NYHA functional class considering mitral area 1 cm(2) or below went from 51.1-86.4% when mitral resistance below or above 130 dynes.sec.cm(-5), respectively, was considered together. Thus, mitral valve resistance should be used as a complement to the mitral area method in assessment of mitral stenosis, adding the effects of the reduction in mitral area and the damage in mitral valve apparatus.

Journal Article↗

PubMed1985

Average, binge and maximum alcohol intake in healthy young men: discriminant function analysis.

Blood samples and data on alcohol consumption, diet, exercise, smoking and drug use were collected from 89 undergraduate men affiliated with Greek houses. This population has previously been identified as one that consumed large amounts of alcohol among university undergraduates. Energy and nutrient intake from foods, exercise and drug use (except caffeine) were not different between those reporting consuming more than or less than 1 oz of alcohol per day. Multiple discriminant analysis using a panel of 35 blood tests correctly classified all subjects according to their self-reported intake of more than or less than 2 oz of alcohol per day; 96% were classified correctly at a cut point of 1.5 oz per day. Correct classification was 96-99% using a measure of binge consumption (quantity-frequency of heavy drinking). When maximum number of drinks per occasion was used in the discriminant function, 90-92% were correctly placed. Discriminant analysis based on a panel of common blood tests appears to be a promising technique to identify young drinkers who consume high amounts of alcohol.

Adolescent↗

PubMed2004

Pediatric stroke among Hong Kong Chinese subjects.

BACKGROUND: The incidence of pediatric stroke was estimated to be 2.5 to 2.7 cases per 100,000 children per year in North America and 13 cases per 100,000 children per year in France. Stroke is among the top 10 causes of death among children in the United States, with the highest incidence in the first 1 year of life. The annual mortality rate was 0.34 deaths per 100,000 person-years, with an average of 244 deaths per year. Interethnic differences have been demonstrated to be important in pediatric stroke. However, most population-based studies on pediatric stroke were from Europe or North America, and there was a lack of data on the incidence of stroke among Chinese or Asian children. Whether the etiologic patterns and risk factors for death and morbidity among Chinese children with stroke were similar to those described for other ethnic groups was unknown. OBJECTIVES: To calculate the incidence of stroke among Chinese children in Hong Kong and to examine the clinical spectrum, causes, patterns, risk factors, and outcomes of pediatric stroke among Chinese subjects. METHODS: The population of Hong Kong was 6.7 million in 2001, and >98% of our population is Chinese in origin. In Hong Kong, public hospitals under the Hospital Authority provide >95% of the hospital service for the region. We identified children (>1 month to <15 years of age) who were admitted and given a discharge diagnosis of stroke from the Clinical Data Analysis and Reporting System, which is a centralized computerized database for all public hospitals. The discharge coding of stroke used codes from the International Classification of Diseases, 9th Revision, Clinical Modification. Only first admissions during the study period were included. We excluded any subsequent admissions by using multiple demographic characteristics of the patients. The incidence of pediatric stroke was estimated as the number of first hospitalizations divided by the person-years at risk. Since 1991, we had been collecting a database on pediatric stroke (ages of 1 month to 16 years) from a single center (the university-affiliated pediatric unit). The clinical presentation, causes, risk factors, and outcomes for those in the Hong Kong Children's Stroke Registry with follow-up data for > or =2 years were analyzed. Data on outcomes, in terms of survival and neurologic deficits, were studied. For survivors, neurologic deficits were defined as short-term if they resolved within 3 months and long-term if they persisted for >3 months. The severity of deficits was defined as mild when function was minimally affected and the patient remained independent in activities of daily living, moderate when the patient required supervision or partial assistance in activities of daily living or when the deficit caused delay in developmental milestones, and severe when the patient required total or near-total care in activities of daily living. Potential risk factors for death and poor neurologic outcomes, including gender, age at the time of stroke, clinical presentation, causes, and neuroimaging findings, were analyzed. RESULTS: Using projections from census data in 2001, the number of children <15 years of age in Hong Kong from 1998 to 2001 was estimated to be 1,104100 to 1,158800, resulting in 4,545300 person-years. During the same period, 94 children with discharge coding of stroke were identified. Therefore, the estimated incidence of pediatric stroke between 1998 and 2001 was 2.1 cases per 100,000 children-years. The average number of new cases treated annually was 4.5 (0-15 cases/year). Fifty children (28 boys and 22 girls; male/female ratio: 1.27:1) were identified in the 11-year period. The mean age at presentation was 5.6 +/- 4.9 years. Thirty-six strokes (72%) were ischemic and 14 (28%) were hemorrhagic. Despite evaluation for possible underlying causes, 12% (6 cases) remained idiopathic. Eighteen patients with ischemic strokes had cerebral thrombosis, whereas 15 had cerebral embolism. We did not observe any case of sinovenous thrombosis. The 36 cases of ischemic stroke were subtyped according to vascular territories. Eleven cases had infarction involving the middle cerebral artery territory; 2 were limited to the cortical region, 3 were limited to subcortical structures such as the basal ganglia or internal capsule or both, and 6 had complete middle cerebral artery involvement, with cortical and subcortical stroke. Involvement of the anterior cerebral artery occurred in 2 cases, with involvement of cerebellar/basilar artery territories in another 2 cases. The remaining 15 cases had multiple sites of infarction. Three patients experienced secondary hemorrhagic transformation after the initial thrombotic event. Of the 14 patients with hemorrhagic strokes, only 1 had subarachnoid hemorrhage. All others had intracerebral bleeding, at single (N = 9) or multiple (N = 4) loci. Important causes included complications related to congenital heart diseases (N = 15, 30%), vascular diseases (N = 13, 26%), and hematologic diseases (N = 14, 28%). Six cases had no determined causes. One case involved mitochondrial encephalopathy with lactic acidosis and stroke-like episodes and constituted the only case with a metabolic cause. For the 7 patients for whom prothrombotic screening was performed, findings were negative. Seizures (52%) and hemiplegia (34%) were the most common presenting features. Other presenting clinical features included headaches (22%), decreased consciousness (30%), visual field defects (12%), dysphasia (10%), and lethargy (8%). Only 1 patient, with moyamoya disease, had a family history of stroke. The median follow-up time was 8.7 years (range: 2-12.4 years). Nine patients (18%) died, 5 with ischemic stroke and 4 with hemorrhagic stroke. Among the 5 cases of death with ischemic stroke, 3 involved hemorrhagic transformation before death. Seven patients (77%) died within 31 days (range: 2-31 days), whereas the other 2 died 6 months and 2.5 years after the episode. Recurrence occurred in 5 cases (10%). Long-term neurologic deficits occurred among 41% of survivors, including mental retardation (N = 11), epilepsy (N = 7), and hemiplegia (N = 10). The functional deficits were classified as severe in 7 cases, moderate in 3 cases, and mild in 7 cases, for patients with long-term neurologic deficits. Decreased levels of consciousness, hematologic causes, and hemorrhagic transformation (applicable only in ischemic stroke) were significant risk factors associated with high mortality rates. For the 41 patients who survived, the only significant risk factor for long-term neurologic deficits was seizures at the initial presentation. Other factors, such as gender, age, other clinical features, stroke type, vascular territory, other causes, and recurrence of stroke, were all insignificant for both death and long-term deficits. The 3 risk factors identified for death were analyzed in multivariate logistic regression analyses, with adjustment for the confounding variables, and only decreased levels of consciousness remained significant (odds ratio = 15.6). CONCLUSIONS: The incidence of stroke among Chinese children was slightly lower than that in Europe or North America. The etiologic pattern was different in our cohort, and there was no sickle cell anemia, thrombophilia, or sinovenous thrombosis. Despite these differences, however, mortality and long-term neurologic deficit rates were similar.

Adolescent↗

PubMed1987

Multimodal evoked potentials in multiple sclerosis: a contribution to diagnosis and classification.

The adoption of multimodal evoked potential procedures for the diagnosis and classification of multiple sclerosis is recommended as capable of revealing subclinical lesions of the central nervous system. The CSF test is also helpful, though in a different way. We present the results obtained by VEP, BAEP, SEP-median and SEP-tibial examinations and by CSF analysis (OB+) in a group of 189 MS patients, of whom 103 were diagnosed as definite. 58 as probable and 28 as suspected MS cases according to McDonald and Halliday, criteria. EPs frequently displayed clinically silent lesions, enabling us to reclassify 14 of the 37 potentially reclassifiable subjects (37.8%). The CSF test (OB+) allowed the reclassification of 31 out of the 58 reclassifiable cases (53.4%). In the present report we discuss the usefulness of the two methods.

Adolescent↗

PubMed2006

CAFTAN: a tool for fast mapping, and quality assessment of cDNAs.

BACKGROUND: The German cDNA Consortium has been cloning full length cDNAs and continued with their exploitation in protein localization experiments and cellular assays. However, the efficient use of large cDNA resources requires the development of strategies that are capable of a speedy selection of truly useful cDNAs from biological and experimental noise. To this end we have developed a new high-throughput analysis tool, CAFTAN, which simplifies these efforts and thus fills the gap between large-scale cDNA collections and their systematic annotation and application in functional genomics. RESULTS: CAFTAN is built around the mapping of cDNAs to the genome assembly, and the subsequent analysis of their genomic context. It uses sequence features like the presence and type of PolyA signals, inner and flanking repeats, the GC-content, splice site types, etc. All these features are evaluated in individual tests and classify cDNAs according to their sequence quality and likelihood to have been generated from fully processed mRNAs. Additionally, CAFTAN compares the coordinates of mapped cDNAs with the genomic coordinates of reference sets from public available resources (e.g., VEGA, ENSEMBL). This provides detailed information about overlapping exons and the structural classification of cDNAs with respect to the reference set of splice variants. The evaluation of CAFTAN showed that is able to correctly classify more than 85% of 5950 selected "known protein-coding" VEGA cDNAs as high quality multi- or single-exon. It identified as good 80.6 % of the single exon cDNAs and 85 % of the multiple exon cDNAs. The program is written in Perl and in a modular way, allowing the adoption of this strategy to other tasks like EST-annotation, or to extend it by adding new classification rules and new organism databases as they become available. We think that it is a very useful program for the annotation and research of unfinished genomes. CONCLUSION: CAFTAN is a high-throughput sequence analysis tool, which performs a fast and reliable quality prediction of cDNAs. Several thousands of cDNAs can be analyzed in a short time, giving the curator/scientist a first quick overview about the quality and the already existing annotation of a set of cDNAs. It supports the rejection of low quality cDNAs and helps in the selection of likely novel splice variants, and/or completely novel transcripts for new experiments.

Chromosome Mapping↗

PubMed1988

An analysis of commitments made by student clinicians in speech-language pathology and audiology.

This study identified and classified commitments made by student clinicians in speech-language pathology and audiology supervisory conferences and provides a methodology for documenting change in supervisees' behavior as a result of supervisory interaction. Five different types of commitments were isolated and served as the basis of the Commitment Classification System. They included clinical procedures, clinical process administration, supervisory procedures, supervisory process administration, and academic information/teaching function. The independent variables were condition (written agreement or no written agreement between supervisee and supervisor), order of conditions, experience level of supervisee, week number within each condition, and type of commitments. Frequency data were collected for 6 consecutive weeks. The resulting 1,389 commitments made by 64 student clinicians in 384 individual conferences at 12 universities were analyzed within a multiple analysis of variance. Significant findings (p less than .05) indicated that the greatest number of commitments made involved planning, analysis, and evaluation of the clinical process with particular attention on client behavior; and the number of commitments made was a function of written accountability and order of conditions.

Audiology↗

PubMed2000

Lymphoma discrimination by computerized triple matrix analysis of list mode data from three-color flow cytometric immunophenotypes of bone marrow aspirates.

BACKGROUND: The goal of this study was to evaluate a self-learning algorithm for the computer classification of information extracted from flow cytometric immunophenotype list mode files from high-grade non-Hodgkin's lymphoma (NHL), Hodgkin's disease (HD), and multiple myeloma (MM). Materials and Methods Bone marrow aspirates (BMA) were obtained from untreated NHL (n = 51), HD (n = 9), or MM (n = 13) patients. Bone marrow aspirates were not infiltrated in NHL and HD patients as confirmed by thorough histologic and cytologic investigation; however, MM patients showed an infiltration rate >50% by malignant myeloma cells. Peripheral blood leukocyte (PBL) samples were taken from age-matched healthy volunteers (n = 44) as easily available control material. A second control group of 15 healthy volunteers, from whom BMA and PBL samples were available, allowed us to differentiate whether the observed classification results on malignant samples were due to the malignant process or simply to the inherent differences between BMA and PBL. Bone marrow aspirates and PBL were analyzed by the same immunophenotyping antibody panel (CD45/14/20, CD4/8/3, kappa/CD19/5, lambda/CD19/5). The acquired list mode data files were analyzed and classified by the self-learning triple matrix classification algorithms CLASSIF1 following a priori separation of the data into a learning set and unknown test set. After completion of the learning phase, known patient samples were reclassified and unknown samples prospectively classified by the algorithm. RESULTS: Highly discriminatory information was extracted for the various lymphoma entities. The most discriminating information was encountered in antibody binding, antibody binding ratios, and relative antibody surface density parameters of leukocytes rather than in percentage frequencies of discrete leukocyte subpopulations. Samples from healthy controls were classified as normal in 97.2% of the cases, whereas those of NHL, HD, and MM patients were on average correctly classified in 80. 8% of the cases. CONCLUSIONS: Although no detectable lymphoma cells were present in BMA of NHL and HD patients, the CLASSIF1 classification of the immunophenotypes of morphologically normal cells provided a surprisingly good disease discrimination equal or better than that obtained by examining pathological lymph nodes according to the respective literature. The results are suggestive for a lymphoma-related and disease-specific antigen expression shift on normal hematopoietic bone marrow cells that can be used to discriminate the underlying disease (specificity of unspecific changes), i.e., in this case NHL from HD. Multiple myeloma patients were discriminated by changes on malignant as well as on normal bone marrow cells.

Algorithms↗

PubMed2005

Borrowing information from relevant microarray studies for sample classification using weighted partial least squares.

With an increasing number of publicly available microarray datasets, it becomes attractive to borrow information from other relevant studies to have more reliable and powerful analysis of a given dataset. We do not assume that subjects in the current study and other relevant studies are drawn from the same population as assumed by meta-analysis. In particular, the set of parameters in the current study may be different from that of the other studies. We consider sample classification based on gene expression profiles in this context. We propose two new methods, a weighted partial least squares (WPLS) method and a weighted penalized partial least squares (WPPLS) method, to build a classifier by a combined use of multiple datasets. The methods can weight the individual datasets depending on their relevance to the current study. A more standard approach is first to build a classifier using each of the individual datasets, then to combine the outputs of the multiple classifiers using a weighted voting. Using two quite different datasets on human heart failure, we show first that WPLS/WPPLS, by borrowing information from the other dataset, can improve the performance of PLS/PPLS built on only a single dataset. Second, WPLS/WPPLS performs better than the standard approach of combining multiple classifiers. Third, WPPLS can improve over WPLS, just as PPLS does over PLS for a single dataset.

Gene Expression Profiling↗

PubMedDate not supplied

Classification of inhibitors of protein tyrosine phosphatase 1B using molecular structure based descriptors.

Loss of Protein Tyrosine Phosphatase 1B (PTP 1B) activity is known to enhance insulin sensitivity and resistance to weight gain. So potent and orally active PTP1B inhibitors could be potential pharmacological agents for the treatment of Type 2 diabetes and obesity. Classification models of PTP1B inhibitors are developed using a data set containing 128 compounds. Their inhibitory concentrations ranged from -1.59 to 1.68 log units. Initially a two-class (active, inactive) problem is tackled using a number of different methods. The data set was divided into active and inactive classes on the basis of inhibitory activity of the compounds. Molecular structure-based descriptors were calculated and used in the model development. Descriptors encoding the flexibility of the molecules were investigated. Classification models were generated using k-nearest neighbors (k-NN), linear discriminant analysis (LDA), and radial basis function neural network (RBFNN). All models are tested using an external prediction set, compounds not used anywhere during the model development procedure. A five-descriptor model is developed that produces a classification rate of 85.7% for an external prediction set. Then a three-class (active, moderately active, inactive) problem was explored. This time the data set was divided into highly active, moderate, and inactive classes on the basis of inhibitory activity of the compounds. The best classification rate achieved for an external prediction set was 85%. The classification rates achieved indicate that these models could serve as a screening mechanism, to identify potentially useful PTP 1B inhibitors. In addition multiple linear regression and computational neural network models are also developed for prediction of log IC(50) values. All QSAR models are tested using the same external prediction set.

Benzofurans↗

PubMed1987

Attention and learning processes in the identification and categorization of integral stimuli.

The relationship between subjects' identification and categorization learning of integral-dimension stimuli was studied within the framework of an exemplar-based generalization model. The model was used to predict subjects' learning in six different categorization conditions on the basis of data obtained in a single identification learning condition. A crucial assumption in the model is that because of selective attention to component dimensions, similarity relations may change in systematic ways across different experimental contexts. The theoretical analysis provided evidence that, at least under unspeeded conditions, selective attention may play a critical role in determining the identification-categorization relationship for integral stimuli. Evidence was also provided that similarity among exemplars decreased as a function of identification learning. Various alternative classification models, including prototype, multiple-prototype, average distance, and "value-on-dimensions" models, were unable to account for the results.

Attention↗