Edge co-occurrence in natural images predicts contour grouping performance.

The human brain manages to correctly interpret almost every visual image it receives from the environment. Underlying this ability are contour grouping mechanisms that appropriately link local edge elements into global contours. Although a general view of how the brain achieves effective contour grouping has emerged, there have been a number of different specific proposals and few successes at quantitatively predicting performance. These previous proposals have been developed largely by intuition and computational trial and error. A more principled approach is to begin with an examination of the statistical properties of contours that exist in natural images, because it is these statistics that drove the evolution of the grouping mechanisms. Here we report measurements of both absolute and Bayesian edge co-occurrence statistics in natural images, as well as human performance for detecting natural-shaped contours in complex backgrounds. We find that contour detection performance is quantitatively predicted by a local grouping rule derived directly from the co-occurrence statistics, in combination with a very simple integration rule (a transitivity rule) that links the locally grouped contour elements into longer contours.

Telemetry of forces from proximal femoral replacements and relevance to fixation.

Two proximal femoral replacements were instrumented to enable axial forces to be determined at two sites within the prosthesis: in the main shaft and near the tip of the intramedullary stem. The goal was to measure the changes in force distribution over time, as indicated by the ratio of the two forces. Inductive coupling between a coil worn around the leg and a small implanted coil was used, both to supply power to electronic circuits sealed into a welded cavity in the prosthesis and to telemeter data from the prosthesis. Data from both subjects were recorded over the first two years following surgery. For the first subject, there was an increase in mean shaft force excursions (peak force minus resting force) during level walking from 0.53 x BW after 1 week 2.77 x BW after 23 months. The corresponding mean tip force excursions were 0.13 x BW and 1.74 x BW, respectively. The ratio of mean tip force excursions to shaft force excursions steadily increased over the same period from 25 to 63%. Similar increases over time in the tip/shaft ratio were found during treadmill walking, stair climbing and stair descending. Data from the second subject were obtained for the shaft forces only, and were consistent with those from the first subject. The progressive transfer of axial load from the proximal to the distal part of the IM stem recorded telemetrically, together with radiographic observations, suggested that bone remodelling had taken place together with a less stable interface around the proximal part of the stem. This process evidently began soon after implantation.

Migration of binucleate cells in the sheep placenta during normal pregnancy.

The maternofetal interface in the placentomes of the sheep placenta consists of a fetal cellular chorion layer whose apical microvilli interdigitate with those of a syncytial layer which borders the maternal connective tissue. Most of the granulated binucleate cells found in the chorion epithelium reach neither its basement membrane nor its microvillar junction apex. However, up to one fifth of the binucleate cells can be shown to be part of or push pseudopodia across the microvillar junction or are located within the syncytium. The syncytium and chorionic binucleate cells both have a nuclear chromatin structure and distinctive cytoplasmic granules which are not found in any other placental cells. This suggests that most, if not all, of the syncytium is derived by fusion of fetal binucleate cells which have migrated across the microvillar junction from the chorion, and that the immunological maternofetal junction does not lie between chorion and syncytial layers.

Ultrastructure of the uterine glands of the pig.

The ultrastructure of the uterine glands is described from material taken from nulliparous, pregnant and post-partum pigs. The glands are of a simple tubular type, lined by a single layer of glandular epithelium standing on a well developed basal lamina. Apart from those which are ciliate, the cells are of one type, with a large basal nucleus and a well developed supranuclear Golgi region. By the fifth week of pregnancy, saccules and secretion droplets, filled with material which appears to contain acid phosphatase, are associated with the still developing Golgi apparatus, and glycogen accumulates in the cytoplasm. By the ninth week the glandular cells are densely packed with parallel arrays of granular endoplasmic reticulum. As pregnancy continues the glandular epithelial cells enlarge further, the granular endoplasmic reticulum increases in amount, dense droplets collect in the cytoplasm and characteristic tubular organelles develop. In late pregnancy, small globular cells are found sparsely distributed within the basal lamina of the glandular epithelium, each almost completely enfolded within the basal portion of one of the large epithelial cells. They are referred to in this account as 'basal' cells. They are crowded with densely staining membrane-bound droplets but it is not clear whether, or in what way, they may contribute to the glandular secretions. They were not found elsewhere in the material studied and their possible origin is discussed. After parturition the uterine glands undergo very rapid involution; many of the lining cells atrophy and are shed into the uterine lumen.

The existence during gestation of an immunological buffer zone at the interface between maternal and foetal tissues.

In mammalian pregnancy the trophoblast normally constitutes an uninterrupted boundary of foetal tissue in immediate contact with maternal tissue, including blood in some species, and is the decisive immunological barrier to rejection of the foetus as an allograft. The ability of the trophoblast to function as a barrier evidently results from its capacity to resist immunological attack by either alloantibody or alloimmune cells and to prevent immunocompetent cells from reaching and damaging the foetus but, as yet, there is no general agreement regarding the means by which it exercises these functions. In view of the dramatic hormonal changes that occur during pregnancy and the undisputed involvement of trophoblast in these endocrine events, the possibility exists of an interaction between the hormones of pregnancy and the immunological phenomena. The present account furnishes evidence that endocrine activity at the maternal surface of the trophoblast, the presumptive site of the immunological frontier between foetus and mother, may be a factor in its local survival at implantation. The placental hormones so far known that are capable of blocking the antigen receptor sites of the mother's lymphocytes and thus preventing the latter from reacting with the foetal antigens are the glycoprotein, human chorionic gonadotrophin (HCG) and the polypeptide hormone, human chorionic somatomammotrophin (HCS) or human placental lactogen (HPL), both of which are specific to the human placenta. The origin of these hormones, their spatial distribution and their probable interaction with placental steroid hormones are discussed. It is argued that the place of highest concentration of these hormones is on the surface of the syncytial microvilli and the adjacent caviolae of the apical plasma membrane, as well as on the surfaces of the persisting cytotrophoblastic cells of the basal plate (cytotrophoblastic shell), the cell islands and the septa-precisely where the immunological challenge of the foetal allograft to the maternal host occurs. An explanation is offered for the continuing production of the voluminous quantities of these hormones during human pregnancy.

The ultrastructure of the sheep placenta around the time of parturition.

Placental tissue, taken from seven sheep before, during and after delivery, was fixed and sectioned for study by transmission electron microscopy. Daily blood samples from the same sheep were taken from about 135 days p.c. and were analysed by radioimmunoassay for total unconjugated oestrogens. Sections of placentomal and interplacentomal areas were examined for ultrastructural variations that might be correlated with the endocrine changes observed over this period, and to determine whether placental separation preceded delivery of the fetus and whether separation occurred at the fetal-maternal junction or elsewhere. Ultrastructural correlates of the endocrine pattern were not detected. Placental separation occurred after delivery and involved disruption of the fetal tissue (trophoblast), while its microvillous attachment to the maternal symplasma remained intact.

The forces in the distal femur and the knee during walking and other activities measured by telemetry.

The forces and moments in the shaft of a distal femoral replacement were measured by telemetry for a subject during different activities, and calculations were then made of the forces at the knee. The axial force showed a small peak at heel-strike followed by two main peaks during stance. In level walking, the peak axial force was between 1,487 and 1,718 N (2.2-2.5 BW), the peak shear force was 269-368 N (0.4-0.54 BW) directed anteriorly on the tibia, the peak axial torque was 7 Nm internal, while the patellofemoral force was 466-571 N. The highest axial force was recorded for descending stairs (2.8 BW). Standing on one leg produced 2.4 BW, while lying supine and raising the leg produced 1.7 BW. The data produced may resemble that of a normal subject, and has application to basic joint mechanics, to joint reconstruction, and to total knee replacement design and evaluation.

Progesterone and oestrogen in pregnancy and parturition: comparative aspects and hierarchical control.

The different ways in which the progesterone requirements of pregnancy are met in various species are reviewed here. Progesterone production expressed in terms of metabolic body weight lies within about one order of magnitude in several species (but not in the rat), whether the hormone is predominantly ovarian or placental in origin. Parturition is usually preceded by a decrease in the plasma concentration of progesterone and the evidence which suggests that a decrease in secretion involves enzyme induction is summarized. In the sheep a sequence of well-marked hormonal changes can be discerned--an increase in fetal cortisol secretion followed by a fall in placental progesterone and a rise in oestrogen and prostaglandin F2alpha secretion. This sequence has been interpreted in terms of hierarchical control theory which envisages a series of levels placed in a descending order of importance and in which there is a downward transmission of a signal from a higher level that modifies activity at a lower level. A dominant level in the hierarchy in the sheep and pig seems to be the fetal hypothalamus and pituitary, but in the rabbit the maternal hypothalamus, pituitary and ovary play a more dominant role.