Classification and staging of chronic liver disease from multimodal data

Chronic liver disease (CLD) is most of the time an asymptomatic, progressive, and ultimately potentially fatal disease. In this study, an automatic hierarchical procedure to stage CLD using ultrasound images, laboratory tests, and clinical records are described. The first stage of the proposed method, called clinical based classifier (CBC), discriminates healthy from pathologic conditions. When nonhealthy conditions are detected, the method refines the results in three exclusive pathologies in a hierarchical basis: 1) chronic hepatitis; 2) compensated cirrhosis; and 3) decompensated cirrhosis. The features used as well as the classifiers (Bayes, Parzen, support vector machine, and k-nearest neighbor) are optimally selected for each stage. A large multimodal feature database was specifically built for this study containing 30 chronic hepatitis cases, 34 compensated cirrhosis cases, and 36 decompensated cirrhosis cases, all validated after histopathologic analysis by liver biopsy. The CBC classification scheme outperformed the nonhierachical one against all scheme, achieving an overall accuracy of 98.67% for the normal detector, 87.45% for the chronic hepatitis detector, and 95.71% for the cirrhosis detector.

The usefulness of ultrasound in the classification of chronic liver disease

Chronic Liver Disease is a progressive, most of the time asymptomatic, and potentially fatal disease. In this paper, a semi-automatic procedure to stage this disease is proposed based on ultrasound liver images, clinical and laboratorial data. In the core of the algorithm two classifiers are used: a k nearest neighbor and a Support Vector Machine, with different kernels. The classifiers were trained with the proposed multi-modal feature set and the results obtained were compared with the laboratorial and clinical feature set. The results showed that using ultrasound based features, in association with laboratorial and clinical features, improve the classification accuracy. The support vector machine, polynomial kernel, outperformed the others classifiers in every class studied. For the Normal class we achieved 100% accuracy, for the chronic hepatitis with cirrhosis 73.08%, for compensated cirrhosis 59.26% and for decompensated cirrhosis 91.67%.

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.

Imunohistoquímica no diagnóstico da glomerulonefrite membranosa: estudo comparativo de binómios cromogênio + coloração de contraste

A glomerulonefrite membranosa faz parte das doenças glomerulares que provocam glomerulonefrite crônica, apresentando-se como uma das causas da doença renal terminal. As técnicas de imunofluorescência são o gold standard no estudo imunológico desta patologia em biópsia renal, através da deteção de imunocomplexos (e.g. IgG e C3) e do seu padrão de distribuição granular característico. No entanto, a imunofluorescência não permite uma contextualização histológica e os fluorocromos utilizados possuem um reduzido tempo de atividade, ao contrário das técnicas imunoenzimáticas que utilizam cromogénios coloridos precipitados que permitem a obtenção de uma marcação permanente e a sua contextualização histológica por via da utilização de eficientes colorações de contraste. Com a finalidade de contribuir para a qualidade do diagnóstico da glomerulonefrite membranosa, em biópsias renais, procurou-se, com esta pesquisa, identificar uma técnica imunoenzimática, através da conjugação entre diferentes cromogênios e colorações de contraste, que permita a deteção de depósitos de IgG e C3, com padrão granular. Foram constituídos diferentes binômios cromogênio + coloração, com os cromogênios 3,3›- Diaminobenzidine Tetrahydrochloride e 3-Amino-9-ethylcarbazole e as colorações Periodic Acid Schiff, Periodic Acid Methenamine Silver e Hematoxilina. Foram utilizadas 72 secções de tecido provenientes de seis de casos de biópsias renais com diagnóstico de glomerulonefrite membranosa, fixados em formalina a 10% e incluídos em parafina. A recolha de dados foi realizada por observação microscópica com preenchimento de uma grelha de classificação dos parâmetros: preservação da morfologia, intensidade da marcação específica, quantidade relativa de estruturas marcadas, marcação inespecífica/fundo, contraste e padrão da marcação, que permitiu a classificação dos binómios estudados num score quantitativo de 0-100 pontos. O binômio que apresentou melhores resultados foi 3-Amino-9-ethylcarbazole + Hematoxilina (score 71,81) e o binômio 3,3›- Diaminobenzidine Tetrahydrochloride+Periodic Acid Methenamine Silver (score 7,81), apresentou os piores resultados. O resultado do teste Kruskal-Wallis indica-nos a presença de diferenças estatísticas entre os binómios em estudo (p=0,000). A Hematoxilina pode ser considerada a coloração mais eficaz, pois cumpriu a sua função de auxiliar e facilitar a observação do tipo de padrão com os dois cromogênios utilizados. O cromogênio 3-Amino-9-ethylcarbazole apresentou resultados semelhantes aos produzidos pelo 3,3›-Diaminobenzidine Tetrahydrochloride, no entanto, permitiu identificar em todos os casos o padrão granular de imunomarcação, ao contrário do que aconteceu com este último.

Diffuse liver disease classification from ultrasound surface characterization, clinical and laboratorial data

In this work liver contour is semi-automatically segmented and quantified in order to help the identification and diagnosis of diffuse liver disease. The features extracted from the liver contour are jointly used with clinical and laboratorial data in the staging process. The classification results of a support vector machine, a Bayesian and a k-nearest neighbor classifier are compared. A population of 88 patients at five different stages of diffuse liver disease and a leave-one-out cross-validation strategy are used in the classification process. The best results are obtained using the k-nearest neighbor classifier, with an overall accuracy of 80.68%. The good performance of the proposed method shows a reliable indicator that can improve the information in the staging of diffuse liver disease.

Risk factors for metabolic bone disease in Crohn's disease patients

Background: The aim was to evaluate the presence of metabolic bone disease (MBD) in patients with Crohn’s disease (CD) and to identify potential etiologic factors. Methods: The case–control study included 99 patients with CD and 56 controls with a similar age and gender distribution. Both groups had dual-energy x-ray absorptionmetry and a nutritional evaluation. Single nucleotide polymorphisms at the IL1, TNF-a, LTa, and IL-6 genes were analyzed in patients only. Statistical analysis was performed using SPSS software. Results: The prevalence of MBD was significantly higher in patients (P ¼ 0.006). CD patients with osteoporosis were older (P < 0.005), small bowel involvement and surgical resections were more frequent (P < 0.005), they more often exhibited a penetrating or stricturing phenotype (P < 0.05), duration of disease over 15 years (P < 0.005), and body mass index (BMI) under 18.5 kg/m2 (P < 0.01) were more often found. No association was found with steroid use. Patients with a Z-score < 2.0 more frequently had chronic active disease (P < 0.05). With regard to diet, low vitamin K intake was more frequent (P ¼ 0.03) and intake of total, monounsaturated, and polyunsaturated fat was higher in patients with Z-score < 2.0 (P < 0.05). With respect to genetics, carriage of the polymorphic allele for LTa252 A/G was associated with a higher risk of osteoporosis (P ¼ 0.02). Regression analysis showed that age over 40 years, chronic active disease, and previous colonic resections were independently associated with the risk of developing MBD. Conclusions: The prevalence of MBD was significantly higher in CD patients. Besides the usual risk factors, we observed that factors related to chronic active and long-lasting disease increased the risk of MBD.