Maria Rita Massai

Nuclear morphometry as an important prognostic factor in stage I renal cell carcinoma

Although 60% of Stage I renal carcinoma patients die from tumor within 5 years postoperatively, a considerable percentage survive that period. Nuclear grading can help to predict the outcome, but many of the patients are Grade 2, and the prognosis of this subclass is uncertain. Therefore, nuclear morphometry was carried out in 41 patients with Stage I renal cell carcinoma. Of these, 24 died within 5 years and 17 have survived that period. Using a mean nuclear area of 32 μm2 as the decision threshold, none of the 24 short‐term survivors are below that threshold and three of the long‐term survivors exceed that value (18% false‐positives) (99% confidence limit). Separate analysis with sets for learning and testing and Grade 2 patients gave similar results. The results show the essential prognostic value of morphometry in this set of patients with Stage I renal cell carcinoma.

Expression of the ALK protein by anaplastic large-cell lymphomas correlates with high proliferative activity

A variable fraction of anaplastic large‐cell lymphomas (ALCLs) exhibits a t(2;5)(p23;q35) translocation that results in expression of the chimeric hyperphosphorylated protein NPM‐ALK (p80). Tumor cells expressing NPM‐ALK exhibit markedly enhanced proliferative activity, but comparative cellular kinetic studies on ALK+ (ALK lymphomas) and ALK– lymphomas are lacking. The present study showed that ALK+ lymphomas, detected with the monoclonal antibody ALKc (n = 17), had significantly higher average values for the proliferation‐associated parameters mitotic index, ana/telophase index, growth index (x × mitotic index – apoptotic index, assuming x = 3), percentages of Ki‐67+ cells and fraction of cells expressing cyclin A or B or the cell cycle–regulatory protein p34cdc2 than did ALK– ALCLs (n = 15). Whether this intense proliferative activity contributes to the good response to chemotherapy and favorable outcome of ALK+ ALCLs remains to be assessed in a larger series of patients. Our findings support the notion that ALK+ and ALK– ALCLs are 2 distinct disease entities. Int. J. Cancer 86:777–781, 2000.

Effect of compliant intermediate airways on total respiratory resistance and elastance in mechanical ventilation

Total respiratory resistance and elastance were estimated off-line in a sample of 60 patients undergoing mechanical ventilation by means of two regression models in order to analyse and understand a possible physiological mechanism determining differences in inspiration and expiration. The first model considered a single value for resistance and elastance over a whole breathing cycle, whereas the second model considered separate values for inspiratory and expiratory resistance and a single value for elastance. Inspiratory resistance was found to be lower than expiratory resistance, and intermediate values were obtained for resistance estimated over the whole breathing cycle. Students t-test showed a highly significant difference between these resistance estimates, and principal components analysis demonstrated a significant increase in information when both inspiratory and expiratory resistances were used. Minor differences were found between values of elastance calculated with the two approaches. In an attempt to interpret these experimental results, a lung model incorporating the non-linear viscoelastic properties of the intermediate airways was considered. This model suggested that changes in intermediate airway volume play a significant role in breathing mechanics during artificial ventilation and indicated that inspiratory and expiratory resistance could be useful parameters for locating airway obstruction.

A neural network improves the classification of high-risk intensive care patients

A two-layer feedforward neural network for classifying intensive care patients as normal or at risk for severe cardiorespiratory disorders was designed and compared with the best, previously investigated, Bayesian classifier of similar complexity. At three distinct observation times soon after heart surgery, the three variables most effective in separating the two classes were measured and used to test the performances of classifiers by the leave-one-out method. The results showed that the ability of the well-known neural network to describe input-output nonlinear behaviour made it possible to obtain a lower level of misclassification of new data without loss of generalization.

A naïve Bayes classifier for planning transfusion requirements in heart surgery

RATIONALE, AIMS AND OBJECTIVES Transfusion of allogeneic blood products is a key issue in cardiac surgery. Although blood conservation and standard transfusion guidelines have been published by different medical groups, actual transfusion practices after cardiac surgery vary widely among institutions. Models can be a useful support for decision making and may reduce the total cost of care. The objective of this study was to propose and evaluate a procedure to develop a simple locally customized decision-support system. METHODS We analysed 3182 consecutive patients undergoing cardiac surgery at the University Hospital of Siena, Italy. Univariate statistical tests were performed to identify a set of preoperative and intraoperative variables as likely independent features for planning transfusion quantities. These features were utilized to design a naïve Bayes classifier. Model performance was evaluated using the leave-one-out cross-validation approach. All computations were done using spss and matlab code. RESULTS The overall correct classification percentage was not particularly high if several classes of patients were to be identified. Model performance improved appreciably when the patient sample was divided into two classes (transfused and non-transfused patients). In this case the naïve Bayes model correctly classified about three quarters of patients with 71.2% sensitivity and 78.4% specificity, thus providing useful information for recognizing patients with transfusion requirements in the specific scenario considered. CONCLUSIONS Although the classifier is customized to a particular setting and cannot be generalized to other scenarios, the simplicity of its development and the results obtained make it a promising approach for designing a simple model for different heart surgery centres needing a customized decision-support system for planning transfusion requirements in intensive care unit.

Modified Karhunen-Loéve expansion for evaluating skin-colour-associated melanoma risk factors

An approach based on the modified Karhunen-Loéve expansion (MKLE) of constitutive and facultative skin colour data acquired by colorimeters in melanoma patients and healthy control subjects, was used to identify two colour features defining skin-colour-associated risk of melanoma. None of four common statistical classifiers trained on colour features were sufficiently accurate for allowing skin colour alone to be used for classification purposes, though a Bayesian quadratic classifier matched the transformed data well. This study supersedes the indeterminate character of most common clinical criteria based on qualitative factors and, irrespective of the results of classification, provides objective skin colour information for the prevention of melanoma.

A stepwise regression approach applied to evaluation of nonlinear patterns in breathing mechanics

A multiple linear stepwise regression approach was applied to respiratory data from intensive care patients. The influence of different nonlinear viscoelastic properties on the prediction of pressure at the airway opening taken as dependent variable was studied by entering and removing potential predictive variables. These variables were derived from measured ventilatory flow data so as to quantitatively describe elastic and resistive nonlinearities attributed to typical pathophysiological mechanisms. It was found that for a better characterization of breathing mechanics, volume-dependent lung elastance must be introduced and that turbulence effects can sometimes arise. The resulting regression model still corresponds to the classical one-compartment model of breathing mechanics, but two parameters are: airflow-dependent resistance and lung volume-dependent elastance.