Transesophageal echocardiography (TEE) provides high quality 3D echocardiographic recordings in close proximity to the heart which is beneficial for interventions in the catheterization laboratory (cath lab) for both preprocedural planning and for monitoring the intervention. However, additional training and a certain level of operator expertise are required to master the maneuvering of the TEE probe in order to easily obtain consistent anatomical views. To ease these requirements, an anatomical view stabilization method is proposed. The stabilization method helps to visualize consistent TEE views without the need to repeatably maneuver the probe to the exact same position and orientation. The proposed method was tested on in-vivo data sets acquired from 5 patients pre, during or post intervention in the cath lab. Two validation approaches were performed offline, namely semi-quantitative scoring by two independent trained observers on the consistency of the anatomical views and quantitative measurements of the location and orientation of 3 manually annotated landmark points at the aortic root. The results from both validation approaches showed significant improvement in the user scores and significant reduction in the landmark distances after stabilization. These results show the potential of the proposed method in increasing reproducibility of particular anatomical views in TEE imaging.
Abstract Fifty double-muscled Belgian Blue bulls were divided into five groups. The first group (control, CG) was given, on an ad libitum basis, a fattening diet based on sugar-beet pulp. In G2 and G3, fattening was interrupted 103 and 187 days, respectively after the beginning of the experiment, by approximately a 2-month period of food restriction during which the animals received a maintenance ration. The last two groups, namely G4 and G5, received for about 4 months a limited amount of the restriction diet to support growth of 0·5 and 0 kg/day, respectively, before being fattened as CG. The average daily gain was 1·48 kg/day in CG v. 1·33, 1·30, 1·43 and 1·61 kg/day during the period of ad libitum feeding in G2, G3, G4 and G5. Higher nitrogen (N) balance during compensatory growth in G2, G4 and G5 (78·8, 81·0 and 74·6 v. 53·3 g N per day in CG, P < 0·001) resulted mainly from a higher efficiency of N retention. In G3, however, compensatory growth was characterized by lower N digestibility and efficiency of N retention. Muscle proportion in the carcass was higher in G3 while fat plus connective tissue proportion was lower (P < 0·05). Muscle pH values in CG were lower than in the other groups (P < 0·01) but ether extract values were higher. The intramuscular fat of G2 and G3 was characterized by higher proportions of polyunsaturated fatty acids.
The aim of this article is to evaluate the accuracy and reproducibility of two-dimensional speckle tracking echocardiography (2D-STE) for the estimation of left ventricular (LV) twist, using tagged cardiac magnetic resonance (cMR) as the reference standard, and to assess how much 2D-STE rotational parameters are affected by the level at which measurements are made within the LV.Forty-three patients with various heart diseases and 10 healthy volunteers underwent cMR and 2D-STE on the same day. With both methods, basal and apical time-rotation curves were generated at endocardial, midwall, and epicardial levels. By using the most apical cMR short-axis cross-section as a comparator, apical rotation was significantly underestimated by 2D-STE. When 2D-STE and cMR short-axis cross-sections were matched for their internal dimensions, measurements of endocardial, midwall, and epicardial twists no longer differ between cMR and 2D-STE (12.6 +/- 5.9 vs. 12.5 +/- 5.7 degrees , 10.5 +/- 4.6 vs. 9.7 +/- 4.1 degrees , and 8.9 +/- 4.0 vs. 8.4 +/- 3.7 degrees , respectively, all P = ns).Compared with tagged cMR, 2D-STE underestimates apical rotation and LV twist. This is related to the inability of 2D-STE to image the real LV apex in most of the patients. However, when 2D-STE and cMR data are compared at similar acquisition levels, both techniques provide similar values.
This article presents an original method to accurately predict the end of discharge of rechargeable batteries inserted in portable electronic equipments. The proposed method is based on two neural networks organized in a master-slave relation. A prediction accuracy of 3% (18 minutes) is reached. A further improvement of the system is introduced by adapting on-line another neural network to the actual battery currently in use. This adaptive method reduces the average error to 10 minutes. Results are promising and implementation, carried out in a portable multimeter prototype, only requires a small amount of the computing power already available inside most portable equipments.
Abstract In Belgium, both dairy and beef cattle are managed intensively. Three commercial farms owning both a Belgian Blue —double muscled type- suckler beef herd and a Holstein-Friesian dairy herd were surveyed between 1994 and 1998. The aim of the study was to compare growth and reproduction performances of both breeds in a similar farm environment. The replacement heifers of both breeds were kept together until advanced pregnancy. Reproduction data were collected monthly from heifers older than fourteen months. The live weight of the replacement heifers was recorded quarterly until 20 months of age. Breed influenced the growth of replacement heifers at 6, 9 and 12 months (P < 0.001). The growth rate of Belgian Blue heifers was reduced after the age of one year and the live weight difference between breeds was not significant anymore. Belgian Blue heifers were always heavier than Holstein-Friesian heifers. Successfully bred heifers had a greater growth rate than non successfully bred heifers within the Belgian Blue breed. This was significant for the weight at 6 months (P < 0.05) and at 12 months (P < 0.05). Such an effect was not detected for the Holstein-Friesian breed (P > 0.05). Weight was not a limiting factor for fecundity of the Holstein-Friesian heifers. Management policies other than growth performance explained the difference between optimal fecundity and the fecundity index met in this study. Non successfully bred Belgian Blue heifers were lighter than successfully bred animals. Unsufficient weight performance had an impact on fecundity in that breed.
Nous proposons de considerer le probleme de l'extraction de contours comme la restauration d'une sequence de points. L'extracteur propose est un systeme hybride reseau de neurones - chaine de Markov qui permet l'introduction dans le modele de connaissances a priori de haut niveau. Une application a la detection automatique du ventricule gauche du cœur dans des radiographies numeriques est entierement decrite
