Automatic Small Bowel Tumor Diagnosis by Using Multi-Scale Wavelet-Based Analysis in Wireless Capsule Endoscopy Images

BACKGROUND: Wireless capsule endoscopy has been introduced as an innovative, non-invasive diagnostic technique for evaluation of the gastrointestinal tract, reaching places where conventional endoscopy is unable to. However, the output of this technique is an 8 hours video, whose analysis by the expert physician is very time consuming. Thus, a computer assisted diagnosis tool to help the physicians to evaluate CE exams faster and more accurately is an important technical challenge and an excellent economical opportunity. METHOD: The set of features proposed in this paper to code textural information is based on statistical modeling of second order textural measures extracted from co-occurrence matrices. To cope with both joint and marginal non-Gaussianity of second order textural measures, higher order moments are used. These statistical moments are taken from the two-dimensional color-scale feature space, where two different scales are considered. Second and higher order moments of textural measures are computed from the co-occurrence matrices computed from images synthesized by the inverse wavelet transform of the wavelet transform containing only the selected scales for the three color channels. The dimensionality of the data is reduced by using Principal Component Analysis. RESULTS: The proposed textural features are then used as the input of a classifier based on artificial neural networks. Classification performances of 93.1% specificity and 93.9% sensitivity are achieved on real data. These promising results open the path towards a deeper study regarding the applicability of this algorithm in computer aided diagnosis systems to assist physicians in their clinical practice.

Blood Donation in the Hospital de S. José. A 11-Year Retrospective Analysis of Deferrals and Exclusions

Background: Approximately 5% of the population donates blood each year in developed countries. Recruiting and maintaining a pool of altruistic and healthy blood donors is a challenging task. Blood donation as a dynamic process must naturally co-exist with the arguably essential deferrals. Aims: To analyse a 11-year cohort of donors and blood donations in order to determine the profile of the average donor and the typical donation. Characterize the donor’s population in terms of gender, age, number of donations, most common causes for deferral and exclusion and the possible relationships between them. Establish the tendency flow of donations per year. Methods: Analysis of 95861 blood donations from 31550 donors collected between 2000 and 2010 (11 years) in the Immunohemotherapy Department of the ‘‘Centro Hospitalar Lisboa Central - Hospital de Sa˜o Jose´’’ (Lisboa, Portugal). Prior to blood donation, donors were required to fill out a form of informed consent, a questionnaire of 21 ‘‘yes or no’’ questions and were submitted to a clinical assessment and physical examination including: measurement of weight, blood pressure, pulse and capillary hemoglobin levels. Post-donation, the collected blood was tested for ALT elevation and blood-borne viral agents (HBV, HCV, HIV 1 and 2 and HTLV 1 and 2) and other infections (Treponema pallidum). Blood donors and donations were registered in a database and statistically studied in terms of: gender and age distribution, number of donations, most common causes for deferral and exclusion. The frequency of blood donations throughout the period of observation was analyzed and statistically significant relationships between the collected variables were investigated. Results: From the population of 31550 donors 61% were male and a mean age of 41.5 years (± 12.5 years) was found. From the total of 95682 blood donations collected 78% were successful while the most common causes for deferral were: donation incompatible hemoglobin levels (5% of the blood donations and 22% of deferrals), ALT elevation (3% and 14%), positive blood screening test for Treponema pallidum (1% and 6%), medication (1% and 4%), positive serological blood markers for HBV (1% and 4%), endoscopy in the previous 12 months (1% and 3%), arterial hypertension (1% and 3%), infectious conditions (1% and 3%), influenza or influenza-like symptoms (1% and 2%) and positive serological blood markers for HCV (1% and 2%). Summary/Conclusions: Several factors may have contributed to a limited number of new regular donors in the population, namely: ageing population, the alienation of the individual from the community induced by modern lifestyles and job precariousness. It is of the utmost importance to refine our blood donation campaigns according to the existing population of donors. The optimization of the blood donation potential of a population of donors must be achieved through the development of reliable and consistent screening methods. In order to appeal to new donors it is important to promote blood donations considering the profile of the regular and healthy blood donor of the existing population.

Comparison of Methods and Co-Registration Maps of EEG and fMRI in Occipital Lobe Epilepsy

Clinically childhood occipital lobe epilepsy (OLE) manifests itself with distinct syndromes. The traditional EEG recordings have not been able to overcome the difficulty in correlating the ictal clinical symptoms to the onset in particular areas of the occipital lobes. To understand these syndromes it is important to map with more precision the epileptogenic cortical regions in OLE. Experimentally, we studied three idiopathic childhood OLE patients with EEG source analysis and with the simultaneous acquisition of EEG and fMRI, to map the BOLD effect associated with EEG spikes. The spatial overlap between the EEG and BOLD results was not very good, but the fMRI suggested localizations more consistent with the ictal clinical manifestations of each type of epileptic syndrome. Since our first results show that by associating the BOLD effect with interictal spikes the epileptogenic areas are mapped to localizations different from those calculated from EEG sources and that by using different EEG/fMRI processing methods our results differ to some extent, it is very important to compare the different methods of processing the localization of activation and develop a good methodology for obtaining co-registration maps of high resolution EEG with BOLD localizations.