Comparative Analysis between Statistical and Artificial Intelligence Models in Business Failure Prediction

A growing number of predicting corporate failure models has emerged since 60s. Economic and social consequences of business failure can be dramatic, thus it is not surprise that the issue has been of growing interest in academic research as well as in business context. The main purpose of this study is to compare the predictive ability of five developed models based on three statistical techniques (Discriminant Analysis, Logit and Probit) and two models based on Artificial Intelligence (Neural Networks and Rough Sets). The five models were employed to a dataset of 420 non-bankrupt firms and 125 bankrupt firms belonging to the textile and clothing industry, over the period 2003–09. Results show that all the models performed well, with an overall correct classification level higher than 90%, and a type II error always less than 2%. The type I error increases as we move away from the year prior to failure. Our models contribute to the discussion of corporate financial distress causes. Moreover it can be used to assist decisions of creditors, investors and auditors. Additionally, this research can be of great contribution to devisers of national economic policies that aim to reduce industrial unemployment.

Semi-automatic 3D Segmentation Of Costal Cartilage In CT Data From Pectus Excavatum Patients

One of the current frontiers in the clinical management of Pectus Excavatum (PE) patients is the prediction of the surgical outcome prior to the intervention. This can be done through computerized simulation of the Nuss procedure, which requires an anatomically correct representation of the costal cartilage. To this end, we take advantage of the costal cartilage tubular structure to detect it through multi-scale vesselness filtering. This information is then used in an interactive 2D initialization procedure which uses anatomical maximum intensity projections of 3D vesselness feature images to efficiently initialize the 3D segmentation process. We identify the cartilage tissue centerlines in these projected 2D images using a livewire approach. We finally refine the 3D cartilage surface through region-based sparse field level-sets. We have tested the proposed algorithm in 6 noncontrast CT datasets from PE patients. A good segmentation performance was found against reference manual contouring, with an average Dice coefficient of 0.75±0.04 and an average mean surface distance of 1.69±0.30mm. The proposed method requires roughly 1 minute for the interactive initialization step, which can positively contribute to an extended use of this tool in clinical practice, since current manual delineation of the costal cartilage can take up to an hour.

A hybrid PSO-ANFIS approach for short-term wind power prediction in Portugal

The increased integration of wind power into the electric grid, as nowadays occurs in Portugal, poses new challenges due to its intermittency and volatility. Wind power prediction plays a key role in tackling these challenges. The contribution of this paper is to propose a new hybrid approach, combining particle swarm optimization and adaptive-network-based fuzzy inference system, for short-term wind power prediction in Portugal. Significant improvements regarding forecasting accuracy are attainable using the proposed approach, in comparison with the results obtained with five other approaches.

Valor preditivo do "Clinical Risk Index for Babies" para o risco de mortalidade neonatal

OBJETIVO: Comparar o valor preditivo do CRIB (Clinical Risk Index for Babies) para o risco de mortalidade neonatal ao peso de nascimento (PN) e idade gestacional (IG). MÉTODO: Numa coorte prospectiva foram estudados, durante o ano de 1996, 71 recém-nascidos admitidos na unidade de terapia intensiva, com PN < 1.500 g e/ou IG < 31 semanas. RESULTADOS: A taxa de mortalidade foi de 29,6%; para PN < 1.000 g ou IG < 29 semanas foi de 60%, enquanto que para o escore de CRIB > 10 foi de 100%. CONCLUSÕES: O escore de CRIB > 10 correspondeu a maiores especificidade e valor preditivo positivo em relação aos demais parâmetros. A área determinada pela "receiver operating characteristic" relativa ao CRIB também foi superior. O CRIB mostrou-se um marcador mais acurado na previsão de risco de mortalidade quando comparado ao PN ou IG isoladamente.

Monitorização da insuficiência cardíaca por parâmetros fisiológicos em portadores de cardioversor-desfibrilhador implantável e/ou sistema de ressincronização cardíaca: HF-PREDICT

Mestrado em Tecnologia de Diagnóstico e Intervenção Cardiovascular. Área de especialização: Intervenção Cardiovascular.

Clinical and laboratory parameters in dapsone acute intoxication

OBJECTIVE: To determine the severity of dapsone (DDS) acute intoxication – an uncommon medical event – using clinical and laboratory parameters. METHODS: Two hundred and seventy four patients with acute DDS intoxication, aged 1 month to 50 years old, were studied and classified into four age groups. Clinical evaluation was assessed through a protocol and correlated with laboratory parameters. Spectrophotometric methods were used to analyze methemoglobinemia (MHbp) and dapsonemia (DDSp). RESULTS: The most prevalent clinical sign of intoxication was cyanosis, seen in 65.7% of the patients and in 100% of children less than 5 years of age. According to laboratory criteria, MHbp-related severe clinical intoxication was seen in 56.2% and DDSp-related occurred in 58% of the patients. Regarding DDSp, intoxication was considered severe when 20 tablets (100 mg each) were ingested, a median of 29 mug/ml. Regarding MHbp, intoxication was severe when 7.5 tablets were ingested, a median of 38% of the total Hb. The correlation between MHbp and DDSp was statistically significant (n=144, r=0.32, p<0.05). Negative correlation was observed between MHbp and the time elapsed since DDS intake (n=124, r=-0.34, p<0.001). There was also a negative correlation between DDSp and the time elapsed since DDS intake (n=63, r=-0.35, p<0.0001). CONCLUSIONS: Longitudinal analysis showed a significant association between methemoglobinemia and the time elapsed after the intake (t), according to the equation: Dapsonemia = 12.9256 - 0.0682t + 0.234 methemoglobinemia

Trypanosoma cruzi prevalence and clinical forms in blood donor candidates in Brazil

The prevalence and clinical forms of Trypanosoma cruzi were evaluated among blood donor candidates attended at a general hospital in Rio de Janeiro, Brazil, from January 1997 to April 1999. The investigation was done by means of the indirect hemagglutination test and was confirmed via ELISA. Data were collected from clinical examinations, conventional electrocardiogram, chest radiography and echocar-diography. The results showed that despite Trypanosoma cruzi prevalence of 1.17% (128 patients), mainly in males aged 40 years or over, 70.8% of these patients, mainly males aged 19 to 39 years, demonstrated abnormalities that allowed the diagnosis of cardiopathy and/or esophagopathy. This once again corroborates the importance of Trypanosoma cruzi infection in urban centers.

Neonatal anthropometry: a tool to evaluate the nutritional status and predict early and late risks

Neonatal anthropometry is an inexpensive, noninvasive and convenient tool for bedside evaluation, especially in sick and fragile neonates. Anthropometry can be used in neonates as a tool for several purposes: diagnosis of foetal malnutrition and prediction of early postnatal complications; postnatal assessment of growth, body composition and nutritional status; prediction of long-term complications including metabolic syndrome; assessment of dysmorphology; and estimation of body surface. However, in this age group anthropometry has been notorious for its inaccuracy and the main concern is to make validated indices available. Direct measurements, such as body weight, length and body circumferences are the most commonly used measurements for nutritional assessment in clinical practice and in field studies. Body weight is the most reliable anthropometric measurement and therefore is often used alone in the assessment of the nutritional status, despite not reflecting body composition. Derived indices from direct measurements have been proposed to improve the accuracy of anthropometry. Equations based on body weight and length, mid-arm circumference/head circumference ratio, and upper-arm cross-sectional areas are among the most used derived indices to assess nutritional status and body proportionality, even though these indices require further validation for the estimation of body composition in neonates.

Clinical Education - a fundamental curricular unit for all health practitioners

Clinical education is recognized as being crucial for the training of health professionals. This subject is debated amongst teachers, students and professionals. Besides the clinical and research skills, we look for other competencies such as oratory, creative thinking or leadership. We present the results of a study with 4th graders. It’s a exploratory study; the main purpose was to evaluate the outcomes of a unit of clinical education prepared according a new set of competencies and methodologies. The competencies were seen as valuable. Organization, leading or supporting a colleague, rethinking a program to serve client and family are equally important.

Novel method using a laser scanning cytometer for detection of Mycobacteria in clinical samples

In order to evaluate the capacity of laser scanning cytometry (LSC) to detect acid-fast bacilli directly on clinical samples, a comparison between Kinyoun-stained smears analyzed under light microscopy and propidium iodide-auramine-stained smears analyzed by LSC was performed. The results were compared with those for culture on BACTEC MGIT 960. LSC is a new, reliable methodology to detect Mycobacteria.

Validade da versão em português da Clinical Dementia Rating

OBJETIVO: Analisar a validade da versão em português da Clinical Dementia Rating para classificar a função cognitiva de idosos. MÉTODOS: Utilizou-se o instrumento Mini-Mental State Examination para rastreamento de déficit cognitivo em coorte composta por 424 idosos. Foram selecionados todos que obtiveram escores <26 (108 idosos) e 48 idosos com escores >26. Os 156 idosos selecionados foram submetidos a uma avaliação clínica e testes neuropsicológicos para diagnóstico de casos de demência. Tanto os casos como os não-casos foram classificados segundo a versão em português da Clinical Dementia Rating em: normais, casos questionáveis e casos de demência leve, moderada ou grave. RESULTADOS: Entre os 156 avaliados, 122 eram não-casos, destes 62 (51%) foram classificados como normais (CDR=0) e questionáveis 60 (49%) (CDR=0,5). Entre os 34 casos de demência, 17 (50%) foram classificados como demência leve (CDR=1), 8 (23%) moderada (CDR=2) e 6 (18%) grave (CDR=3). Apenas três (9%) dos casos foram considerados questionáveis pelo Clinical Dementia Rating. Sua sensibilidade foi de 91,2% e a especificidade de 100%, com valor preditivo positivo de 100% e negativo de 97,6%. As pontuações no Mini-Mental State Examination declinaram significativamente conforme o grau de demência. CONCLUSÕES: O Clinical Dementia Rating mostrou ser instrumento válido para classificar o grau de demência entre idosos. Quase metade dos não-casos foram casos questionáveis pelo Clinical Dementia Rating e podem corresponder a casos de transtorno cognitivo leve, com maior risco de conversão em demência.

Avaliação ecocardiográfica de doentes propostos para terapêutica de ressincronização: impacto dos índices de dissincronia

Mestrado em Tecnologia de Diagnóstico e Intervenção Cardiovascular - Área de especialização: Ultrassonografia Cardiovascular.

Reliability of nasal auscultation and their relation to middle ear condition – a clinical interest to pediatrics respiratory physiotherapy

Background: Acute respiratory infections are usual in children under three years old occurring in upper respiratory tract, having an impact on child and caregiver’s quality of life predisposing to otitis media or bronchiolitis. There are few valid and reliable measures to determine the child’s respiratory condition and to guide the physiotherapy intervention. Aim: To assess the intra and inter rater reliability of nasal auscultation, to analyze the relation between sounds’ classification and middle ear’s pressure and compliance as well as with the Clinical Severity Score. Methods: A cross-sectional observational study was composed by 125 nursery children aged up to three years old. Tympanometry, pulmonary and nasal auscultation and application of Clinical Severity Score were performed to each child. Nasal auscultation sounds’ were recorded and sent to 3 blinded experts, that classified, as “obstructed” and “unobstructed”, with a 48 hours interval, in order to analyze inter and intra rater reliability. Results: Nasal auscultation revealed a substantial inter and intra rater reliability (=0,749 and evaluator A - K= 0,691; evaluator B - K= 0,605 and evaluator C - K= 0,724, respectively). Both ears’ pressure was significantly lower in children with an "unobstructed" nasal sound when compared with an “obstructed” nasal sound (t=-3,599, p<0,001 in left ear; t=-2,258, p=0,026 in right ear). Compliance in both ears was significantly lower in children with an "obstructed" nasal sound when compared with “unobstructed” nasal sound (t=-2,728, p=0,007 in left ear; t=-3,830, p<0,001 in right ear). There was a statistically significant association between sounds’ classification and tympanograms types in both ear’s (=11,437, p=0,003 in left ear; =13,535, p=0,001 in right ear). There was a trend to children with an "unobstructed" nasal sound that had a lower clinical severity score when compared with “obstructed” children. Conclusion: It was observed a good intra and substantial inter reliability for nasal auscultation. Nasal auscultation sounds’ classification was related to middle ears’ pressure and compliance.

Chronic liver disease staging classification based on ultrasound, clinical and laboratorial data

In this work the identification and diagnosis of various stages of chronic liver disease is addressed. The classification results of a support vector machine, a decision tree and a k-nearest neighbor classifier are compared. Ultrasound image intensity and textural features are jointly used with clinical and laboratorial data in the staging process. The classifiers training is performed by using a population of 97 patients at six different stages of chronic liver disease and a leave-one-out cross-validation strategy. The best results are obtained using the support vector machine with a radial-basis kernel, with 73.20% of overall accuracy. The good performance of the method is a promising indicator that it can be used, in a non invasive way, to provide reliable information about the chronic liver disease staging.