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.

Perinatal Ischemic Stroke: a Five-Year Retrospective Study in a Level-III Maternity

Objective To study the incidence, clinical presentation, risk factors, imaging diagnosis, and clinical outcome of perinatal stroke. Methods Data was retrospectively collected from full-term newborns admitted to the neonatal unit of a level III maternity in Lisbon with cerebral stroke, from January 2007 to December 2011. Results There were 11 cases of stroke: nine were arterial ischemic stroke and two were cerebral venous sinus thrombosis. We estimated an incidence of arterial ischemic stroke of 1.6/5,000 births and of cerebral venous sinus thrombosis of 7.2/100,000 births. There were two cases of recurrent stroke. Eight patients presented with symptoms while the remaining three were asymptomatic and incidentally diagnosed. The most frequently registered symptoms (8/11) were seizures; in that, generalized clonic (3/8) and focal clonic (5/8). Strokes were more commonly left-sided (9/11), and the most affected artery was the left middle cerebral artery (8/11). Transfontanelle ultrasound was positive in most of the patients (10/11), and stroke was confirmed by cerebral magnetic resonance in all patients. Electroencephalographic recordings were carried out in five patients and were abnormal in three (focal abnormalities n=2, burst-suppression pattern n=1). Eight patients had previously identified risk factors for neonatal stroke which included obstetric and neonatal causes. Ten patients were followed up at outpatients setting; four patients developed motor deficits and one presented with epilepsy. Conclusions Although a modest and heterogeneous sample, this study emphasizes the need for a high level of suspicion when it comes to neonatal stroke, primarily in the presence of risk factors. The prevalence of neurological sequelae in our series supports the need of long-term follow-up and early intervention strategies.