Augmenting Scalable Communication-Based Role Allocation for a Three-Role Task

In evolutionary robotics role allocation studies, it is common that the role assumed by each robot is strongly associated with specific local conditions, which may compromise scalability and robustness because of the dependency on those conditions. To increase scalability, communication has been proposed as a means for robots to exchange signals that represent roles. This idea was successfully applied to evolve communication-based role allocation for a two-role task. However, it was necessary to reward signal differentiation in the fitness function, which is a serious limitation as it does not generalize to tasks where the number of roles is unknown a priori. In this paper, we show that rewarding signal differentiation is not necessary to evolve communication-based role allocation strategies for the given task, and we improve reported scalability, while requiring less a priori knowledge. Our approach for the two-role task puts fewer constrains on the evolutionary process and enhances the potential of evolving communication-based role allocation for more complex tasks. Furthermore, we conduct experiments for a three-role task where we compare two different cognitive architectures and several fitness functions and we show how scalable controllers might be evolved.

[Predictors of cesarean delivery in the Kinkanda general hospital in Matadi (Democratic Republic of Congo)]

OBJECTIVE

To identify predictors of cesarean delivery in the Kinkanda general hospital.

METHOD

Cross-sectional study conducted from 1st January 2012 to 31st December 2012, on women in labor who delivered at the maternity of Kinkanda. Predictors of cesarean delivery were identified by using binary logistic regression.

RESULTS

A uterine height of at least 33 cm at induction of labor (AOR 3.93 95% CI 2.04 to 7.59; p <0.001), primiparity (AOR 2.1 95%CI:1.06 to 4.15; p = 0.033), non-membership of the pregnant woman to this particular health center area (AOR 4.26 95%CI 2.16 to 8.42; p <0.001) and a weight of the newborn of at least 4000g (AOR 7.8 95%CI 2.3 to 26.44; p <0.001) were identified as risk factors for cesarean delivery.

CONCLUSION

Primiparity, uterine height of at least 33 cm, living in another health area than the Matadi area and fetal macrosomia are the main predictors of cesarean delivery during the year 2012 at the Kinkanda general hospital.

odNEAT: An Algorithm for Decentralised Online Evolution of Robotic Controllers

Online evolution gives robots the capacity to learn new tasks and to adapt to changing environmental conditions during task execution. Previous approaches to online evolution of neural controllers are typically limited to the optimisation of weights in networks with a prespecified, fixed topology. In this article, we propose a novel approach to online learning in groups of autonomous robots called odNEAT. odNEAT is a distributed and decentralised neuroevolution algorithm that evolves both weights and network topology. We demonstrate odNEAT in three multirobot tasks: aggregation, integrated navigation and obstacle avoidance, and phototaxis. Results show that odNEAT approximates the performance of rtNEAT, an efficient centralised method, and outperforms IM-(μ + 1), a decentralised neuroevolution algorithm. Compared with rtNEAT and IM-(μ + 1), odNEAT's evolutionary dynamics lead to the synthesis of less complex neural controllers with superior generalisation capabilities. We show that robots executing odNEAT can display a high degree of fault tolerance as they are able to adapt and learn new behaviours in the presence of faults. We conclude with a series of ablation studies to analyse the impact of each algorithmic component on performance.

Online Evolution of Adaptive Robot Behaviour

The authors propose and evaluate a novel approach to the online synthesis of neural controllers for autonomous robots. The authors combine online evolution of weights and network topology with neuromodulated learning. The authors demonstrate our method through a series of simulation-based experiments in which an e-puck-like robot must perform a dynamic concurrent foraging task. In this task, scattered food items periodically change their nutritive value or become poisonous. The authors demonstrate that the online evolutionary process, both with and without neuromodulation, is capable of generating controllers well adapted to the periodic task changes. The authors show that when neuromodulated learning is combined with evolution, neural controllers are synthesised faster than by evolution alone. An analysis of the evolved solutions reveals that neuromodulation allows for a more effective expression of a given topology's potential due to the active modification of internal dynamics. Neuromodulated networks learn abstractions of the task and different modes of operation that are triggered by external stimulus.

PMCNS

Novelty search is an evolutionary approach in which the population is driven towards behavioural innovation instead of towards a fixed objective. The use of behavioural novelty to score candidate solutions precludes convergence to local optima. However, in novelty search, significant effort may be spent on exploration of novel, but unfit behaviours. We propose progressive minimal criteria novelty search (PMCNS) to overcome this issue. In PMCNS, novelty search can freely explore the behaviour space as long as the solutions meet a progressively stricter fitness criterion. We evaluate the performance of our approach by evolving neurocontrollers for swarms of robots in two distinct tasks. Our results show that PMCNS outperforms fitness-based evolution and pure novelty search, and that PMCNS is superior to linear scalarisation of novelty and fitness scores. An analysis of behaviour space exploration shows that the benefits of novelty search are conserved in PMCNS despite the evolutionary pressure towards progressively fitter behaviours.

Burden of unidentifiable mycobacteria in a reference laboratory

Modern identification techniques at the genomic level have greatly improved the taxonomic knowledge of mycobacteria. In adjunct to nucleic acid sequences, mycobacterial identification has been endorsed by investigation of the lipidic patterns of unique mycolic acids in such organisms. In the present investigation, the routine use of high-performance liquid chromatography (HPLC) of mycolic acids, followed by the sequencing of the 16S rRNA, allowed us to select 72 mycobacterial strains, out of 1,035 screened, that do not belong to any of the officially recognized mycobacterial species. Most strains (i.e., 47) were isolated from humans, 13 were from the environment, 3 were from animals, and 9 were from unknown sources. The majority of human isolates were grown from the respiratory tract and were therefore most likely not clinically significant. Some, however, were isolated from sterile sites (blood, pleural biopsy, central venous catheter, or pus). Many isolates, including several clusters of two or more strains, mostly slow growers and scotochromogenic, presented unique genetic and lipidic features. We hope the data reported here, including the results of major conventional identification tests, the HPLC profiles of strains isolated several times, and the whole sequences of the 16S rRNA hypervariable regions of all 72 mycobacteria, may encourage reporting of new cases. The taxonomy of the genus Mycobacterium is, in our opinion, still far from being fully elucidated, and the reporting of unusual strains provides the best background for the recognition of new species. Our report also shows the usefulness of the integration of novel technology to routine diagnosis, especially in cases involving slow-growing microorganisms such as mycobacteria.

The Reoviridae family

The members of the Reoviridae family are extremely varied in host ranges and have such diverse natural histories that it is compelling to conclude that their structural asset and replication strategy are uniquely successful in evolutionary terms. It follows that their study addresses fundamental aspects of virology, besides the ones which are customary with important pathogens affecting humans, animals and plants. We deal here with the present taxonomy of the family Reoviridae and of its genera, two of which (Orbivirus and Rotavirus) are treated separately in this issue. Along with a cursory presentation of the remaining genera, we give a concise update of recent findings on the genus Orthoreovirus.

Tentative evidence of AIDS-associated biotype of Mycobacterium kansasii

Previous studies revealed heterogeneous behavior within the species Mycobacterium kansasii against commercially available DNA probes (Accuprobe M. kansasii culture identification test; Gen-Probe); several isolates, conventionally identified as M. kansasii, failed in fact to hybridize. Looking for a possible association with phenotypic features, we tested a fully characterized panel of 69 clinical isolates of M. kansasii (19 of which were Accuprobe negative) with a semiquantitative micromethod which tests for 19 enzymatic activities (Api Zym; BioMérieux). The strains were from 25 hospitals in 18 Italian towns; 20 isolates came from human immunodeficiency virus type 1-positive patients who fulfilled the Centers for Disease Control criteria for AIDS diagnosis. On the basis of the whole set of phenotypic traits, our strains clustered in two groups, allowing the differentiation of biotypes within the species. There was a perfect association between biotype 2 and hybridization failures with Accuprobe and a very significant association between this novel biotype 2 and AIDS status, which suggests that it differs in virulence.

Nosocomial diarrhoeas in a surgical division hyperendemic for Clostridium difficile: epidemiologic aspects emerging from an analysis of clinical records

Having previously shown that Clostridium difficile was responsible for an intense and protracted endemic of nosocomial diarrhoeas in the surgical division of a Tuscan hospital, we started a retrospective analysis on all records from the affected division, to cover a period of 15 months. A statistical description is given of a large series of nosocomial diarrhoeas, as well as direct estimates of their incidence rates in selected high risk subgroups. The situation described is epidemiologically unique, and its study tells how an endemic may reach a steady state of high nosocomial morbidity.

Epidemiological studies of Pseudomonas aeruginosa infections in leg ulcers

A 6-month prospective study was carried out in an angiology ward on patients with leg ulcers. Pseudomonas aeruginosa was isolated from about half of the samples from the patients. Of the isolates, 85 were serotyped, fingerprinted as to passive pyocin sensitivity and characterized for antibiotic susceptibility pattern. According to all criteria, the isolates belonged to six strains, two of which were clearly epidemic: one infected 11 patients yielding 52 identical isolates, the other infected 4 patients (19 isolates).

Age-related changes in the prevalence of precipitating antibodies to BK virus in infants and children

Nearly 1000 sera from children were tested by immunoelectro-osmophoresis against BK virus, and age-specific prevalence rates were estimated from birth until the age of 12 years. Declining rates during the first 12 months showed the waning of passive immunity, which at birth reflects the mother's immune status. The changes of prevalence suggested that the peak incidence of primary infections occurred at about 2 years, with an estimated peak annual rate of 24.6%.

[Role of viruses in acute gastroenteritis]

In a two years screening carried out on Florentine children we confirmed that Rotaviruses are the most important ethiological agents of acute gastroenteritis in infants less than two years old, irrespective of sex. We had evidence that Rotavirus diarrhea does not occur with the same incidence every year. Rotavirus gastroenteritis is characterized by profuse diarrhea, vomiting, fever and often by respiratory symptoms. Severe complications have not been observed.

Immunoelectroosmophoresis for the human papovavirus BK

The reaction between BK virus and its antibodies may well be visualized with immunoelectroosmophoresis (IEOP). Serum antibody levels may be compared or graded by reacting them against a continuous antigen front. In seroepidemiologic surveys IEOP compares favourably with hemagglutination inhibition in terms of sensitivity, practicability and reproducibility. Technical conditions are described which are suitable for antigen detection, quantification and identification.

[Serodiagnosis of rotavirus infections in gastroenteritis of infants]

We report the use of the line immunoelectroosmophoresis (LIEOP), a modification of the immunoelectroosmophoresis (IEOP), for the serodiagnosis of rotavirus infection, in comparison with the complement fixation test. We tested 17 paired and 29 single sera from hospitalized children affected by acute gastroenteritis; rotavirus had been searched in the stools both by electron microscopy and IEOP. The LIEOP showed seroconversion in all the 11 cases with positive stools and also in 5 out of 5 patients whose stool sample had been negative. Either we had not been able to detect the virus, or there was an intervening (nosocomial) rotavirus infection. All sera which were reactive by complement fixation with bovine rotavirus were also reactive by LIEOP. Some of the sera which reacted by LIEOP, both with the homologous and the heterologous antigen, were negative by complement fixation.

[Pyocin typing of 111 strains of Pseudomonas aeruginosa isolated in Tuscany, using the Gillies and Govan method]

Pseudomonas aeruginosa strains from Tuscany have been typed for pyocine production against Govan and Gillies' extended set of indicator strains; results have been expressed following a binary code. Among 111 strains only 6 could not be typed; the remaining 105 strains (94.6%) presented 59 different patterns, 44 of which were represented by one isolate only. The most commonly encountered pattern (n. 8045, corresponding to Govan and Gillies' subtype 1e) was shared by 13 strains. Isolates from outpatients or from the environment tended to a uniform distribution of pyocine patterns; those from two different hospitals tended to group into distinct subtypes.

Surface glycosaminoglycans and calcium distribution in 3T3 cells

Resting 3T3 cells have relatively more sulphated glycosaminoglycans and Ca2+ in their cell coat than do growing or SV40-virus-transformed cells. It is suggested that the different Ca2+-binding capacity controls cellular activities by affecting the distribution of Ca2+ between the intra- and ecto-cellular compartments.

Biometrical implications of factorial experiments for the study of lymphocyte mitogenic response

Lymphocyte responsiveness to mitogens is affected by a number of experimental variables; the latter may be conveniently studied several at a time, following a factorial design that allows the complete evaluation of the effects of each variable and of their interplays. Statistical inferences may be drawn through an analysis of variance, provided that the data conform to the underlying model, i.e. that the effects are additive and that the errors are normally distributed and homoscedastic. Thymidine incorporation data should first be transformed to log cpm; the effects on the metameter may then be assumed to be additive, and their errors approach normality; the error variances may be affected by the experimental variables involved, so that the homogeneity of the error must be checked before relying on the pooled estimate of the error for parametric tests of significance. The above considerations have been illustrated with factorial experiments on [3H]TdR incorporation by rabbit spleen cells stimulated with varying amounts of PHA-P or PHA-M, under a variety of culture conditions.

[Multivariate analysis of aminoacyl-tRNA synthetase activities of liver and hepatomas (author's transl)]

Data on the aminoacyl-tRNA synthetase activities in normal liver and in three transplantable rat hepatomas, viz Yoshida's AH 130 and Morris's 5123C and 7793, were studied by means of factor analysis, a powerful technique of multivariate analysis particularly suitable foridentifying factors which theoretically may account for the pattern of interrelations between several variables. The analysis has been performed on 51 patterns of enzyme activities, covering 17 out of the 20 aminoacyl-tRNA synthetases; 80% on average of the total variability of the 17 enzyme activities may be accounted for by the first 4 factors extracted. The enzymes seem to fall into different groups, depending on their relationship with the factors thus identified. The results suggest that enzymes belonging to the same group share a common control mechanism, and are independent of the enzymes belonging to different groups, both in normal liver and in hepatomas.