A neural network seismic detector

This experimental study focuses on a detection system at the seismic station level that should have a similar role to the detection algorithms based on the ratio STA/LTA. We tested two types of neural network: Multi-Layer Perceptrons and Support Vector Machines, trained in supervised mode. The universe of data consisted of 2903 patterns extracted from records of the PVAQ station, of the seismography network of the Institute of Meteorology of Portugal. The spectral characteristics of the records and its variation in time were reflected in the input patterns, consisting in a set of values of power spectral density in selected frequencies, extracted from a spectro gram calculated over a segment of record of pre-determined duration. The universe of data was divided, with about 60% for the training and the remainder reserved for testing and validation. To ensure that all patterns in the universe of data were within the range of variation of the training set, we used an algorithm to separate the universe of data by hyper-convex polyhedrons, determining in this manner a set of patterns that have a mandatory part of the training set. Additionally, an active learning strategy was conducted, by iteratively incorporating poorly classified cases in the training set. The best results, in terms of sensitivity and selectivity in the whole data ranged between 98% and 100%. These results compare very favorably with the ones obtained by the existing detection system, 50%.

Insights into molecular and cellular events involved in normal and pathological vertebral development and fusion using zebrafish as model organism

The vertebral column and its units, the vertebrae, are fundamental features, characteristic of all vertebrates. Developmental segregation of the vertebral bodies as articulated units is an intrinsic requirement to guarantee the proper function of the spine. Whenever these units become fused either during development or postsegmentation, movement is affected in a more or less severe manner, depending on the number of vertebrae affected. Nevertheless, fusion may occur as part of regular development and as a physiological requirement, like in the tetrapod sacrum or in fish posterior vertebrae forming the urostyle. In order to meet the main objective of this PhD project, which aimed to better understand the molecular and cellular events underlying vertebral fusion under physiological and pathological conditions, a detailed characterization of the vertebral fusion occurring in zebrafish caudal fin region was conducted. This showed that fusion in the caudal fin region comprised 5 vertebral bodies, from which, only fusion between [PU1++U1] and ural2 [U2+] was still traceable during development. This involved bone deposition around the notochord sheath while fusion within the remaining vertebral bodies occur at the level of the notochord sheath, as during the early establishment of the vertebral bodies. A comparison approach between the caudal fin vertebrae and the remaining vertebral column showed conserved features such as the presence of mineralization related proteins as Osteocalcin were identified throughout the vertebral column, independently on the mineralization patterns. This unexpected presence of Osteocalcin in notochord sheath, here identified as Oc1, suggested that this gene, opposing to Oc2, generally associated with bone formation and mature osteoblast activity, is potentially associated with early mineralization events including chordacentrum formation. Nevertheless, major differences between caudal fin region and anterior vertebral bodies considering arch histology and mineralization patterns, led us to use RA as an inductive factor for vertebral fusion, allowing a direct comparison of equivalent structures under normal and fusion events. This fusion phenotype was associated with notochord sheath ectopic mineralization instead of ectopic perichordal bone formation related with increased osteoblast activity, as suggested in previous reports. Additionally, alterations in ECM content, cell adhesion and blood coagulation were discussed as potentially related with the fusion phenotype. Finally, Matrix gla protein, upregulated upon RA treatment and shown to be associated with chordacentrum mineralization sites in regular development, was further described considering its potential function in vertebral formation and pathological fusion. Therefore with this work we propose zebrafish caudal fin vertebral fusion as a potential model to study both congenital and postsegmentation fusion and we present candidate factors and genes that may be further explored in order to clarify whether we can prevent vertebral fusion.

On-line operation of an intelligent seismic detector

This study describes the on-line operation of a seismic detection system to act at the level of a seismic station providing similar role to that of a STA /LTA ratio-based detection algorithms. The intelligent detector is a Support Vector Machine (SVM), trained with data consisting of 2903 patterns extracted from records of the PVAQ station, one of the seismographic network's stations of the Institute of Meteorology of Portugal (IM). Records' spectral variations in time and characteristics were reflected in the SVM input patterns, as a set of values of power spectral density at selected frequencies. To ensure that all patterns of the sample data were within the range of variation of the training set, we used an algorithm to separate the universe of data by hyper-convex polyhedrons, determining in this manner a set of patterns that have a mandatory part of the training set. Additionally, an active learning strategy was conducted, by iteratively incorporating poorly classified cases in the training set. After having been trained, the proposed system was experimented in continuous operation for unseen (out of sample) data, and the SVM detector obtained 97.7% and 98.7% of sensitivity and selectivity, respectively. The same type of ANN presented 88.4 % and 99.4% of sensitivity and selectivity when applied to data of a different seismic station of IM. © 2013 Springer-Verlag Berlin Heidelberg.

LifeSpeeder: a web and mobile platform for events location

The use of smartphones and tablets as become almost banal in these days. Smartphones, besides serving their main purpose of making and receiving calls, come to be one of the main equipments to obtain information from the Internet, using the commonly installed browsers or through the use of dedicated applications. Furthermore, several other devices are also very frequent to the majority of the modern smartphones and tablets in the market (e.g., GPS - Global Positioning System). This devices give the current systems a very high potential of usage. One example of applicability, comes from the wish to find and navigate to events or activities which are or will soon be occurring near the user. The LifeSpeeder platform is one of the first applications in the mobile equipment market of applications which take into consideration exactly what we have just outlined, i.e., a mobile and desktop application which allows the users to locate events according with their preferences and to get help navigating to them. In this paper we briefly describe the LifeSpeeder's front and back-end. © 2014 Springer International Publishing.