Action recognition via bio-inspired features: The richness of center–surround interaction

Motion is a key feature for a wide class of computer vision approaches to recognize actions. In this article, we show how to define bio-inspired features for action recognition. To do so, we start from a well-established bio-inspired motion model of cortical areas V1 and MT. The primary visual cortex, designated as V1, is the first cortical area encountered in the visual stream processing and early responses of V1 cells consist in tiled sets of selective spatiotemporal filters. The second cortical area of interest in this article is area MT where MT cells pool incoming information from V1 according to the shape and characteristic of their receptive field. To go beyond the classical models and following the observations from Xiao et al. [61], we propose here to model different surround geometries for MT cells receptive fields. Then, we define the so-called bio-inspired features associated to an input video, based on the average activity of MT cells. Finally, we show how these features can be used in a standard classification method to perform action recognition. Results are given for the Weizmann and KTH databases. Interestingly, we show that the diversity of motion representation at the MT level (different surround geometries), is a major advantage for action recognition. On the Weizmann database, the inclusion of different MT surround geometries improved the recognition rate from 63.01 ± 2.07% up to 99.26 ± 1.66% in the best case. Similarly, on the KTH database, the recognition rate was significantly improved with the inclusion of MT different surround geometries (from 47.82 ± 2.71% up to 92.44 ± 0.01% in the best case). We also discussed the limitations of the current approach which are closely related to the input video duration. These promising results encourage us to further develop bio-inspired models incorporating other brain mechanisms and cortical areas in order to deal with more complex videos.     

Personalities and alimentary behaviors in obese patients

The ability of lonidamine (LND), an energolytic derivative of indazole-carboxylic acid, to modulate the cytotoxic activity of cisplatin (CDDP) and epidoxorubicin (EPI), singly or in combination, was investigated in two human breast cancer cell lines (MCF7 and T47D). A 72-hr post-incubation with a non-cytotoxic concentration of LND (75 microM) increased the activity of a 1-hr CDDP treatment as well as that of a 1 to 16-hr EPI treatment. A different pattern of interaction among the drugs and modulator was observed as a function of the sequence of drug treatment. Specifically, supra-additive or additive effects of the combination were obtained in the two cell lines according to the different treatment schemes. In particular, the maximum potentiation was observed in MCF7 cells simultaneously exposed to CDDP, EPI and LND for 1 hr and then post-incubated with LND for 72 hr, and in T47 first exposed to EPI and LND, then to CDDP and LND, and finally post-incubated with LND. Flow cytometric analysis of MCF7 cell distribution in the different cycle phases showed that combined treatment with EPI/CDDP/LND was able to stabilize cell cycle perturbations (mainly G2M accumulation) induced by individual agents. The ability of LND to potentiate CDDP and EPI cytotoxicity, and the consideration that LND causes side effects different from those caused by alkylating agents and anthracyclines, make this compound an attractive candidate for multidrug combination therapy in breast cancer.     

Structural features determining the antibiotic potencies of natural and synthetic hop bitter resins, their precursors and derivatives

Twenty-six hop bitter resins, some hitherto not investigated, were tested for antimicrobial activities. Gram-positive bacteria were much more sensitive than Gram-negative ones. The inhibitory effect against Bacillus subtilis 168 was measured by several methods and the general rule could be established that the antibiotic properties are mainly dependent on the hydrophobic parts of the molecules. Thus the acyl-lupuphenones (2-acyl-3,5-4,4',6-tri(3-methyl-2-butenyl)-cyclohexane-triones (1, 3, 5) having three prenyl and one acyl side chain are the most active substances. Their minimum inhibitory concentration (MIC) increases from the capro (0.5 muM) to the aceto derivative (11 muM). Any substitution with hydrophilic functions or loss of hydrophobic groups causes reductions in biological activity. This is most evident with the corresponding acyl-phloroglucine precursors (2-acyl-1,3,5-trihydroxybenzenes) which lack the three prenyl side chains (MIC, 110 to 5050 muM respectively). Conversion of the central six-membered ring structure into a five-membered one results in additional losses of antimicrobial activity. These findings support the proposal that the lipophilic region of the cell membrane represents the target site for the hop bitter resins.     

A gene for autosomal recessive symmetrical spastic cerebral palsy maps to chromosome 2q24-25

Cerebral palsy has an incidence of approximately 1/500 births, although this varies between different ethnic groups. Genetic forms of the disease account for approximately 1%-2% of cases in most countries but contribute a larger proportion in populations with extensive inbreeding. We have clinically characterized consanguineous families with multiple children affected by symmetrical spastic cerebral palsy, to locate recessive genes responsible for this condition. The eight families studied were identified from databases of patients in different regions of the United Kingdom. After ascertainment and clinical assessment, we performed a genomewide search for linkage, using 290 polymorphic DNA markers. In three families, a region of homozygosity at chromosome 2q24-q25 was identified between the markers D2S124 and D2S148. The largest family gave a maximum LOD score of 3.0, by multipoint analysis (HOMOZ). The maximum combined multipoint LOD score for the three families was 5.75. The minimum region of homozygosity is approximately 5 cM between the markers D2S124 and D2S2284. We have shown that a proportion of autosomal recessive symmetrical spastic cerebral palsy maps to chromosome 2q24-25. The identification of genes involved in the etiology of cerebral palsy may lead to improved management of this clinically intractable condition.     

Percutaneous large-core biopsy of papillary breast lesions

Neisseria meningitidis, Haemophilus influenzae, and Streptococcus pneumoniae possess the ability to cleave human IgA1 antibodies, and all successfully colonize and occasionally invade the human upper respiratory tract. N. meningitidis invades the bloodstream after a period of nasopharyngeal colonization. We directly compared levels of IgA1 protease activity in strains (n=52) derived from the cerebrospinal fluid or blood of patients with meningococcal disease with strains of N. meningitidis obtained from asymptomatic carriers (n=25). IgA1 protease activity was determined by a sensitive semiquantitative ELISA assay. Levels of IgA1 protease activity were significantly higher (P<0.0001) in strains associated with invasive meningococcal disease (98% with detectable activity, mean = 580 mU) than with those obtained from asymptomatic carriers (76% with detectable activity, mean = 280 mU). Despite marked variation in enzyme activity, almost all strains (96%) possessed the gene for IgA1 protease. Given the panmictic population structure of the bacterial isolates investigated, these data, obtained from two groups infected with N. meningitidis, but with markedly different clinical outcomes, provide the first quantitative evidence that IgA1 protease activity is a virulence determinant that contributes to the pathogenic phenotype, and suggest IgA1 protease as a potential target for prophylaxis.     

Cochlear ablation alters acoustically induced c-fos mRNA expression in the adult rat auditory brainstem

Expression of c-fos mRNA was studied in the adult rat brain following cochlear ablations by using in situ hybridization. In normal animals, expression was produced by acoustic stimulation and was found to be tonotopically distributed in many auditory nuclei. Following unilateral cochlear ablation, acoustically driven expression was eliminated or decreased in areas normally activated by the ablated ear, e.g., the ipsilateral dorsal and ventral cochlear nuclei, dorsal periolivary nuclei, and lateral nucleus of the trapezoid body and the contralateral medial and ventral nuclei of the trapezoid body, lateral lemniscal nuclei, and inferior colliculus. These deficits did not recover, even after long survivals up to 6 months. Results also indicated that neurons in the dorsal cochlear nucleus could be activated by contralateral stimulation in the absence of ipsilateral cochlear input and that the influence of the contralateral ear was tonotopically organized. Results also indicated that c-fos expression rose rapidly and persisted for up to 6 months in neurons in the rostral part of the contralateral medial nucleus of the trapezoid body following a cochlear ablation, even in the absence of acoustic stimulation. This response may reflect a release of constitutive excitatory inputs normally suppressed by missing afferent input or changes in homeostatic gene expression related to sensory deprivation. Instances of transient, surgery-dependent increases in c-fos mRNA expression in the absence of acoustic stimulation were observed in the superficial dorsal cochlear nucleus and the cochlear nerve root on the ablated side.     

The CD4+ T lymphocyte is a site of steroid resistance in asthma

Phytohaemagglutinin (PHA)-induced T-cell proliferation is suppressed completely in steroid-sensitive asthma (SSA) by fluticasone propionate (FP). By contrast, in patients with steroid-resistant asthma (SRA), this proliferative response is only partially attenuated by steroids, which suggests that the T lymphocyte may harbour a key molecular defect in these patients. Both CD4+ and CD8+ T cells may be involved in orchestrating the inflammation underlying asthma. We examined whether CD4+ or CD8+ T cells isolated from SRA and SSA patients are equally susceptible to steroid suppression of PHA-induced proliferation. Complete suppression of CD4+ T-lymphocyte proliferation was seen in both SSA and control subjects at concentrations of 10(-9) M FP. In contrast, proliferation of CD4+ T cells from SRA patients was only partially inhibited, even at 10(-6) M FP. CD8+ responses from SRA, SSA and controls were all similar, with only a partial suppression of proliferation at 10(-6) M FP. Differential suppression by FP of CD4+ T cells has thus been demonstrated between SRA and SSA patients.