Transformation and relational-structure schemes for visual pattern recognition

Two models for visual pattern recognition are described; the one based on application of internal compensatory transformations to pattern representations, the other based on encoding of patterns in terms of local features and spatial relations between these local features. These transformation and relational-structure models are each endowed with the same experimentally observed invariance properties, which include independence to pattern translation and pattern jitter, and, depending on the particular versions of the models, independence to pattern reflection and inversion (180° rotation). Each model is tested by comparing the predicted recognition performance with experimentally determined recognition performance using as stimuli random-dot patterns that were variously rotated in the plane. The level of visual recognition of such patterns is known to depend strongly on rotation angle. It is shown that the relational-structure model equipped with an invariance to pattern inversion gives responses which are in close agreement with the experimental data over all pattern rotation angles. In contrast, the transformation model equipped with the same invariances gives poor agreement to the experimental data. Some implications of these results are considered.     

Internal representations and operations in the visual comparison of transformed patterns: Effects of pattern point-inversion,positional symmetry,and separation

A scheme for visual pattern recognition is described. It is supposed, amongst other things, that patterns are internally represented by the visual system in terms of local features, spatial-order relations between local features, and global spatial relations specifying approximate pattern position with respect to the point of fixation. It is further supposed that there are two distinct types of internal operation that may be applied to the components of internal representations in the process of pattern comparison: typically a discrete spatial-order-reversal operation and a continuous position-shift operation. Some general predictions of the scheme are tested against data obtained in an experiment using random-dot patterns that were subjected to rigid transormations and presented at various locations along the horizontal meridian. Patterns were presented sequentially, in pairs, to subjects in a same-different comparison task. Pattern pairs were to be responded to as same if they were identical or related by point-inversion (planar rotation through 180°) or responded to as different. Extending earlier findings, the present results showed that same-detection performance for identical and point-inverted patterns depended differentially on the distance between the patterns and the symmetry of the pattern positions about the point of fixation in a manner consistent with the predictions of the scheme.Part of the work reported in this study was carried out while DHF was on leave from the University of Keele in 1983 as Visiting Professor in the Department of Psychology, Queen's University at Kingston, Ontario. Support by grants from Professor P.C. Dodwell and the Department of Psychology, Queen's University, is gratefully acknowledged. Support by an award from the Science and Engineering Research Council (SERC) to JIK is also gratefully acknowledged     

A spatial perturbation technique for the investigation of discrete internal representations of visual patterns

A technique is proposed for the investigation of discrete internal representations assumed to be formed by the visual system in response to pattern stimuli. The proposed technique involves applying a local 1-parameter group of spatial transformations to a pattern to generate a continuum of patterns. The visual discriminability of pairs of perturbed patterns corresponding to small fixed increments in the transformation parameter is determined at various points in the parameter range. By means of a simple rule, the characteristics of this discrimination performance are then related to the probability density functions assumed to underlie the hypothesized discrete internal representations. Two experimental applications of the proposed technique are described. The first is concerned with a discrete internal representation involving the specification of the collinearity or non-collinearity of the points in a pattern; the second is concerned with a discrete internal representation involving the specification of the acuteness or obtuseness of the angle between two lines in a pattern.     

Tibio-femoral movement in the living knee. A study of weight bearing and non-weight bearing knee kinematics using 'interventional' MRI

The aim of this study was to image tibio-femoral movement during flexion in the living knee. Ten loaded male Caucasian knees were initially studied using MRI, and the relative tibio-femoral motions, through the full flexion arc in neutral tibial rotation, were measured. On knee flexion from hyperextension to 120 degrees , the lateral femoral condyle moved posteriorly 22 mm. From 120 degrees to full squatting there was another 10 mm of posterior translation, with the lateral femoral condyle appearing almost to sublux posteriorly. The medial femoral condyle demonstrated minimal posterior translation until 120 degrees . Thereafter, it moved 9 mm posteriorly to lie on the superior surface of the medial meniscal posterior horn. Thus, during flexion of the knee to 120 degrees , the femur rotated externally through an angle of 20 degrees . However, on flexion beyond 120 degrees , both femoral condyles moved posteriorly to a similar degree. The second part of this study investigated the effect of gender, side, load and longitudinal rotation. The pattern of relative tibio-femoral movement during knee flexion appears to be independent of gender and side. Femoral external rotation (or tibial internal rotation) occurs with knee flexion under loaded and unloaded conditions, but the magnitude of rotation is greater and occurs earlier on weight bearing. With flexion plus tibial internal rotation, the pattern of movement follows that in neutral. With flexion in tibial external rotation, the lateral femoral condyle adopts a more anterior position relative to the tibia and, particularly in the non-weight bearing knee, much of the femoral external rotation that occurs with flexion is reversed.     

Shoulder strength and range-of-motion characteristics in bodybuilders

The purpose of this study was to compare shoulder range-of-motion (ROM) and strength values between bodybuilders and nonbodybuilders. Fifty-four men (29 bodybuilders and 25 nonbodybuilders) between the ages of 21 and 34 years participated in the study. Goniometric measurements were used to assess shoulder flexion and internal and external rotation ROM. Isometric manual muscle tests were performed using a handheld dynamometer. Shoulder flexion, internal and external rotation, abduction, and prone shoulder retraction and elevation strength were tested. Independent t-tests were used to determine levels of statistical significance between the groups. Bodybuilders showed an overall loss of shoulder rotation ROM (166 degrees vs. 180 degrees ) and a significantly decreased internal rotation ROM (-11 degrees ) compared with the control group. Bodybuilders were significantly stronger on all isometric shoulder-strength tests than nonbodybuilders, except for the assessment of lower trapezius strength when expressed as a percentage of body weight. The results of this study indicate that bodybuilders have imbalances regarding strength and ROM at the shoulder that may make them susceptible to shoulder pathology.     

Preliminary crystallographic data for stearoyl-acyl carrier protein desaturase from castor seed.

Recombinant stearoyl-acyl carrier protein desaturase (EC 1.14.99.6) from castor seed has been crystallized with polyethylene glycol 8000 as precipitant. The crystals are orthorhombic, space group P2(1)2(1)2(1) with cell dimensions a = 81.3, b = 146.4 and c = 197.7 A. The observed diffraction pattern extends to at least 2.5 A resolution. Rotation function calculations indicate a non-crystallographic 3-fold rotation axis parallel to the crystallographic a-axis. Perpendicular to this axis, 2-fold rotation axes were found at 30 degrees intervals, i.e. maxima at kappa = 180 degrees, phi = 90 degrees and omega = 30 degrees and 60 degrees, respectively. Together with the packing density of the crystals (Vm = 2.4 A3/Da for n = 6), these results suggest, that the crystal asymmetric unit most likely contains a hexamer of desaturase subunits.     

Rotational flexibility of the human knee due to varus/valgus and axial moments in vivo

Knee ligamentous injuries persist in the sport of Alpine skiing. To better understand the load mechanisms which lead to injury, pure varus/valgus and pure axial moments were applied both singly and in combination to the right knees of six human test subjects. The corresponding relative knee rotations in three degrees of freedom were measured. Knee flexion angles for each test subject were 15 and 60 degrees for the individual moments and 60 degrees for the combination moments. For both knee flexion angles the hip flexion angle was 0 degrees. Leg muscles were quiescent and axial force was minimal during all tests. Tables of data include sample statistics for each of four flexibility parameters in each loading direction. Data were analyzed statistically to test for significant differences in flexibility parameters between the test conditions. In flexing the knee from 15 to 60 degrees, the resulting knee rotations under single moments depended upon flexion angle with varus, valgus, and internal rotations increasing significantly. Also, rotations were different depending on load direction; varus rotation was significantly different and greater than valgus rotation at both flexion angles. Also external rotation was significantly different and greater than internal at 15 degrees flexion, but not at 60 degrees flexion. Coupled rotations under single moments were also observed. Applying pure varus/valgus moments resulted in coupled external/internal rotations which were inconsistent and hence not significant. Applying pure axial moments resulted in consistent and hence significant varus/valgus rotations; an external axial moment induced varus rotation and an internal axial moment induced valgus rotation. For combination moments, varus/valgus rotations decreased significantly from those rotations at similar load levels in the single moment studies. Also, a varus moment significantly increased external rotation and a valgus moment significantly decreased internal rotation. These differences indicate significant interaction between corresponding load combinations. These results suggest that load interaction is a potentially important phenomenon in knee injury mechanics.     

Recognition of spatially transformed objects in men and women: Analysis of behavior and evoked potentials

The accuracy, reaction time, and evoked potentials have been analyzed in 16 men and 15 women during recognition of familiar objects at different levels of spatial transformation. Three levels of transformation have been used: in a fixed position relative to each other, all details were shifted in the radial direction (level 1) followed by analogous displacement in combination with the rotation of all details of the figure by ±0?C45 and ±45?C90 degrees (levels 3 and 4). The task performance did not depend on gender: the image transformation resulted in identification impairment, which was the worst in the case of the maximum rotation of the details. The changes in evoked potentials (ERP) are different for men and women. The amplitude of component P1 in the parietal cortex of men is associated with the level of figure transformation: a greater rotation angle leads to a higher response. In women, figure transformation evokes ERP changes in the time window of negativity N150, and they are associated with ungrouping of the figure, but not with rotation of details, and are located in other visual areas: occipital and temporal. The data obtained are discussed in terms of the theory of gender specificity of the visual representation of space.     

Geometrical and conformational preferences of the 9-fluorenylmethoxycarbonyl-amino moiety.

Structural parameters, originating from x-ray crystallographic data, have been compiled for 13 derivatives of amino acids, peptides and related compounds, which contain a total of 14 Fmoc-NH- moieties. For these moieties, molecular geometries and conformations--described by the omegao, theta1, theta2 and theta3' torsion angles--were analysed and compared with the corresponding parameters for the Z-NH- and Boc-NH-moieties (290 and 553, respectively). To gain a deeper insight into the conformational features of the Fmoc-NH- moiety, ab initio free molecule calculations were performed for fully relaxed minima. Also the potential energy surface as a function of the torsion angles (theta3', theta2) was generated. The conformational features of the Fmoc-NH- moiety: (i) two possible values for the angle omegao (approximately 180 degrees or, rarely, approximately theta degrees) and (ii) the angle theta1 = 180 degrees +/- 15 degrees, are common to the Z-NH- and Boc-NH- systems. By contrast, the theta2 and theta3 angles in the Fmoc, Z and Boc groups differ essentially. In the Fmoc groups theta2 mostly has values of 180 degrees +/- 30 degrees and values up [115 degrees] seem to be forbidden, whereas fewer than half of the Z groups adopt theta2 approximately 180 degrees and the remainder have theta2 in the range of [90 degrees +/- 20 degrees]. On the other hand, the Boc methyl groups are staggered. The theta3 values observed for Fmoc are limited to the regions of 180 degrees +/- 20 degrees and 160 degrees +/- 20 degrees], while for the Z group a variety of theta3 occurs. The orientation of the fluorenyl vs the urethane function is mostly trans. Our results suggest a lower conformational flexibility for the Fmoc group compared with that of the Z group. Our calculations confirm that the observed conformational features for the Fmoc-NH- moiety are inherent properties. The Fmoc-NH-moiety in crystals involves the participation of its O=C-NH functionality in hydrogen bonds.     

Selective recruitment of the triceps surae muscles with changes in knee angle

The muscles of the triceps surae group are important for performance in most sports and in the performance of activities of daily life. In addition, hypertrophy and balance among these muscles are integral to success in bodybuilding. The purpose of this study was to compare the muscle utilization patterns of the 2 major muscles of the triceps surae group, the soleus (SOL) and gastrocnemius (lateral head = LG and medial head = MG), and the tibialis anterior (TA) as an antagonist muscle to the group. Their electromyographic (EMG) signals were compared during 50 constant external resistance contractions at a level established before the testing session. Eleven experienced subjects contributed data during plantar flexion at 3 different knee angles (90, 135, and 180 degrees ). Both root mean square amplitude and integrated signal analyses of the EMGs revealed that the MG produced significantly greater activity than either the SOL or TA at 180 degrees, whereas the LG was not different from the SOL at any knee angle measured. Data also revealed that the SOL produced less electrical activity at 180 degrees than at the other knee angles, whereas the MG produced greater electrical activity. As would be expected, the TA produced lower EMG values than any of the triceps surae muscles at all angles tested. These data indicate that selective targeting of the SOL and MG is possible through the manipulation of knee angle. This targeting appears to be controlled by the biarticular and monoarticular structures of the MG and SOL, respectively. The LG appears less affected by knee position than the MG. Results suggest that the SOL can be targeted most effectively with the knee flexed at 90 degrees and the MG with the leg fully extended. The LG appears to also be more active at 180 degrees; however, it is not as affected as the MG or SOL by knee angle.     

Coupling of rotation and catalysis in F(1)-ATPase revealed by single-molecule imaging and manipulation

F(1)-ATPase is a rotary molecular motor that proceeds in 120 degrees steps, each driven by ATP hydrolysis. How the chemical reactions that occur in three catalytic sites are coupled to mechanical rotation is the central question. Here, we show by high-speed imaging of rotation in single molecules of F(1) that phosphate release drives the last 40 degrees of the 120 degrees step, and that the 40 degrees rotation accompanies reduction of the affinity for phosphate. We also show, by single-molecule imaging of a fluorescent ATP analog Cy3-ATP while F(1) is forced to rotate slowly, that release of Cy3-ADP occurs at approximately 240 degrees after it is bound as Cy3-ATP at 0 degrees . This and other results suggest that the affinity for ADP also decreases with rotation, and thus ADP release contributes part of energy for rotation. Together with previous results, the coupling scheme is now basically complete.     

Temperature-sensitive reaction intermediate of F1-ATPase

F(1)-ATPase is a rotary molecular motor that makes 120 degrees stepping rotations, with each step being driven by a single-ATP hydrolysis. In this study, a new reaction intermediate of F(1)-ATPase was discovered at a temperature below 4 degrees C, which makes a pause at the same angle in its rotation as when ATP binds. The rate constant of the intermediate reaction was strongly dependent on temperature with a Q(10) factor of 19, implying that the intermediate reaction accompanies a large conformational change. Kinetic analyses showed that the intermediate state does not correspond to ATP binding or hydrolysis. The addition of ADP to the reaction mixture did not alter the angular position of the intermediate state, but specifically lengthened the time constant of this state. Conversely, the addition of inorganic phosphate caused a pause at an angle of +80 degrees from that of the intermediate state. These observations strongly suggest that the newly found reaction intermediate is an ADP-releasing step.     

A mathematical model of conformational changes in membrane protein part of F0F1-ATP synthase

The transformation of the electrochemical membrane proton gradient into the energy of chemical bond in adenosinetriphosphate is one of the most important biological processes occurring in the living cell. The main enzyme that directly catalyzes the synthesis of ATP from ADP and Pi in aerobic conditions is F0F1-ATP synthase. In the present study, conformational changes in membrane protein part of ATP synthase during its catalytic cycle were described in terms of dynamic equations of solid state rotation. It was shown that this process should be considered in the framework of classical mechanics. The time dependences of the angle rotation for the protein complex rotates around its central axis were oobtained. Possible types of rotation were analyzed. It was proved that, considering the modern level of knowledge and understanding of ATP synthase structure and function are taken into account, the minimal period of rotor turnover cannot be less than 10(-9) s. With regard for viscous friction and elastic tension in the central axis, the calculated time of whole turnover varies from 10(-6) to 10(-3) and depends on the set of parameters used. These results indicate the essential contribution of friction and elastic tension to the rotation dynamics of the rotor. It is suggested that the proposed model can be used as a part of the entire algorithm for computer simulation of ATP synthase.     

Structure of DNA hexamer sequence d-CGATCG by two-dimensional nuclear magnetic resonance spectroscopy and restrained molecular dynamics

Solution conformation of self-complementary DNA duplex d-CGATCG, containing 5' d-CpG 3' site for intercalation of anticancer drug, daunomycin and adriamycin, has been investigated by nuclear magnetic resonance (NMR) spectroscopy. Complete resonance assignments of all the protons (except some H5'/H5" protons) have been obtained following standard procedures based on double quantum filtered correlation spectroscopy (dQF COSY) and two-dimensional nuclear Overhauser effect (NOE) spectra. Analysis of sums of coupling constants in one-dimensional NMR spectra, cross peak patterns in dQF COSY spectra and inter proton distances shows that the DNA sequence assumes a conformation close to the B-DNA family. The deoxyribose sugar conformation is in dynamic equilibrium with predominantly S-type conformer and a minor N-type conformer with N<-->S equilibrium varying with temperature. At 325 K, the mole fraction of the N-conformer increases for some of the residues by approximately 9%. Using a total of 10 spin-spin coupling constants and 112 NOE intensities, structural refinement has been carried out using Restrained Molecular Dynamics (rMD) with different starting structures, potential functions and rMD protocols. It is observed that pseudorotation phase angle of deoxyribose sugar for A3 and T4 residues is approximately 180 degrees and approximately 120 degrees, respectively while all other residues are close to C2'endo-conformation. A large propeller twist (approximately -18 degrees) and smallest twist angle (approximately 31 degrees) at A3pT4 step, in the middle of the sequence, a wider (12 A) and shallower (3.0 A) major groove with glycosidic bond rotation as high anti at both the ends of hexanucleotide are observed. The structure shows base-sequence dependent variations and hence strong local structural heterogeneity, which may have implications in ligand binding.     

Solution 1H NMR study of the active site structure for the double mutant H64Q/V68F cyanide complex from mouse neuroglobin

Solution (1)H NMR spectroscopy has been carried out to investigate the molecular and electronic structures of the active site in H64Q/V68F double mutant mouse neuroglobin in the cyanomet form. Two heme orientations resulting from a 180 degrees rotation about the alpha-gamma-meso axis were observed with a population ratio about 1:1, and the clearly distinguished B isomer was used to perform the study. Based on the analysis of the dipolar shifts and paramagnetic relaxation constants, the distal Gln(64)(E7) side chain is obtained to adopt an orientation that may produce hydrogen bond between the N(epsilon)H(1) and the Fe-bound cyanide. The side chain of Phe(68)(E11) is oriented out of the heme pocket just like that in triple mutant of cyanide complex of sperm whale myoglobin. A 15 degrees rotation of the imidazole ring in axial His(96) is observed, which is close to the varphi angle determined from the crystal structure of NgbCO. The quantitative determinations of the orientation and anisotropies of the paramagnetic susceptibility tensor reveal that cyanide is tilted by 8 degrees from the heme normal which allows for contact to the Gln(64)(E7) N(epsilon)H(1). The E7 and E11 residues appear to control the direction and the extent of tilt of the bound ligand. Furthermore, the tilt of the ligand has no obvious influence on the heme heterogeneity of cyanide ligation for isomer A/B of the wild type and mutant protein, indicating that factors other than steric effects, such as polarity of heme pocket, impacts on ligand binding affinity.     

Orientation of a beta-hairpin antimicrobial peptide in lipid bilayers from two-dimensional dipolar chemical-shift correlation NMR

The orientation of a beta-sheet membrane peptide in lipid bilayers is determined, for the first time, using two-dimensional (2D) (15)N solid-state NMR. Retrocyclin-2 is a disulfide-stabilized cyclic beta-hairpin peptide with antibacterial and antiviral activities. We used 2D separated local field spectroscopy correlating (15)N-(1)H dipolar coupling with (15)N chemical shift to determine the orientation of multiply (15)N-labeled retrocyclin-2 in uniaxially aligned phosphocholine bilayers. Calculated 2D spectra exhibit characteristic resonance patterns that are sensitive to both the tilt of the beta-strand axis and the rotation of the beta-sheet plane from the bilayer normal and that yield resonance assignment without the need for singly labeled samples. Retrocyclin-2 adopts a transmembrane orientation in dilauroylphosphatidylcholine bilayers, with the strand axis tilted at 20 degrees +/- 10 degrees from the bilayer normal, but changes to a more in-plane orientation in thicker 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphatidyl-choline (POPC) bilayers with a tilt angle of 65 degrees +/- 15 degrees . These indicate that hydrophobic mismatch regulates the peptide orientation. The 2D spectra are sensitive not only to the peptide orientation but also to its backbone (phi, psi) angles. Neither a bent hairpin conformation, which is populated in solution, nor an ideal beta-hairpin with uniform (phi, psi) angles and coplanar strands, agrees with the experimental spectrum. Thus, membrane binding orders the retrocyclin conformation by reducing the beta-sheet curvature but does not make it ideal. (31)P NMR spectra of lipid bilayers with different compositions indicate that retrocyclin-2 selectively disrupts the orientational order of anionic membranes while leaving zwitteronic membranes intact. These structural results provide insights into the mechanism of action of this beta-hairpin antimicrobial peptide.     

Structure of dimeric ATP synthase from mitochondria: an angular association of monomers induces the strong curvature of the inner membrane

Respiration in all cells depends upon synthesis of ATP by the ATP synthase complex, a rotary motor enzyme. The structure of the catalytic moiety of ATP synthase, the so-called F(1) headpiece, is well established. F(1) is connected to the membrane-bound and ion translocating F(0) subcomplex by a central stalk. A peripheral stalk, or stator, prevents futile rotation of the headpiece during catalysis. Although the enzyme functions as a monomer, several lines of evidence have recently suggested that monomeric ATP synthase complexes might interact to form a dimeric supercomplex in mitochondria. However, due to its fragility, the structure of ATP synthase dimers has so far not been precisely defined for any organism. Here we report the purification of a stable dimeric ATP synthase supercomplex, using mitochondria of the alga Polytomella. Structural analysis by electron microscopy and single particle analysis revealed that dimer formation is based on specific interaction of the F(0) parts, not the F(1) headpieces which are not at all in close proximity. Remarkably, the angle between the two F(0) part is about 70 degrees, which induces a strong local bending of the membrane. Hence, the function of ATP synthase dimerisation is to control the unique architecture of the mitochondrial inner membrane.     

Generalization of Dexterous Manipulation Is Sensitive to the Frame of Reference in Which It Is Learned

Studies have shown that internal representations of manipulations of objects with asymmetric mass distributions that are generated within a specific orientation are not generalizable to novel orientations, i.e., subjects fail to prevent object roll on their first grasp-lift attempt of the object following 180° object rotation. This suggests that representations of these manipulations are specific to the reference frame in which they are formed. However, it is unknown whether that reference frame is specific to the hand, the body, or both, because rotating the object 180° modifies the relation between object and body as well as object and hand. An alternative, untested explanation for the above failure to generalize learned manipulations is that any rotation will disrupt grasp performance, regardless if the reference frame in which the manipulation was learned is maintained or modified. We examined the effect of rotations that (1) maintain and (2) modify relations between object and body, and object and hand, on the generalizability of learned two-digit manipulation of an object with an asymmetric mass distribution. Following rotations that maintained the relation between object and body and object and hand (e.g., rotating the object and subject 180°), subjects continued to use appropriate digit placement and load force distributions, thus generating sufficient compensatory moments to minimize object roll. In contrast, following rotations that modified the relation between (1) object and hand (e.g. rotating the hand around to the opposite object side), (2) object and body (e.g. rotating subject and hand 180°), or (3) both (e.g. rotating the subject 180°), subjects used the same, yet inappropriate digit placement and load force distribution, as those used prior to the rotation. Consequently, the compensatory moments were insufficient to prevent large object rolls. These findings suggest that representations of learned manipulation of objects with asymmetric mass distributions are specific to the body- and hand-reference frames in which they were learned.     

Interpreting multiscale domains of tree cover disturbance patterns in North America

Spatial patterns at multiple observation scales provide a framework to improve understanding of pattern-related phenomena. However, the metrics that are most sensitive to local patterns are least likely to exhibit consistent scaling relations with increasing extent (observation scale). A conceptual framework based on multiscale domains (i.e., geographic locations exhibiting similar scaling relations) allows the use of sensitive pattern metrics, but more work is needed to understand the actual patterns represented by multiscale domains. The objective of this study was to improve the interpretation of scale-dependent patterns represented by multiscale domains. Using maps of tree cover disturbance covering North American forest biomes from 2000 to 2012, each 0.09-ha location was described by the proportion and contagion of disturbance in its neighborhood, for 10 neighborhood extents from 0.81 ha to 180 km². A k-means analysis identified 13 disturbance profiles based on the similarity of disturbance proportion and contagion across neighborhood extent. A wall to wall map of multiscale domains was produced by assigning each location (disturbed and undisturbed) to its nearest disturbance profile in multiscale pattern space. The multiscale domains were interpreted as representing two aspects of local patterns – the proximity of a location to disturbance, and the interior-exterior relationship of a location relative to nearby disturbed areas.