Integration of heterogeneous data sources and automated reasoning in healthcare and domotic IoT systems

In recent years, IoT technology has radically transformed many crucial industrial and service sectors such as healthcare. The multi-facets heterogeneity of the devices and the collected information provides important opportunities to develop innovative systems and services. However, the ubiquitous presence of data silos and the poor semantic interoperability in the IoT landscape constitute a significant obstacle in the pursuit of this goal. Moreover, achieving actionable knowledge from the collected data requires IoT information sources to be analysed using appropriate artificial intelligence techniques such as automated reasoning. In this thesis work, Semantic Web technologies have been investigated as an approach to address both the data integration and reasoning aspect in modern IoT systems. In particular, the contributions presented in this thesis are the following: (1) the IoT Fitness Ontology, an OWL ontology that has been developed in order to overcome the issue of data silos and enable semantic interoperability in the IoT fitness domain; (2) a Linked Open Data web portal for collecting and sharing IoT health datasets with the research community; (3) a novel methodology for embedding knowledge in rule-defined IoT smart home scenarios; and (4) a knowledge-based IoT home automation system that supports a seamless integration of heterogeneous devices and data sources.

Design and characterization of automated agricultural vehicle systems

Nowadays, technological advancements have brought industry and research towards the automation of various processes. Automation brings a reduction in costs and an improvement in product quality. For this reason, companies are pushing research to investigate new technologies. The agriculture industry has always looked towards automating various processes, from product processing to storage. In the last years, the automation of harvest and cultivation phases also has become attractive, pushed by the advancement of autonomous driving. Nevertheless, ADAS systems are not enough. Merging different technologies will be the solution to obtain total automation of agriculture processes. For example, sensors that estimate products' physical and chemical properties can be used to evaluate the maturation level of fruit. Therefore, the fusion of these technologies has a key role in industrial process automation. In this dissertation, ADAS systems and sensors for precision agriculture will be both treated. Several measurement procedures for characterizing commercial 3D LiDARs will be proposed and tested to cope with the growing need for comparison tools. Axial errors and transversal errors have been investigated. Moreover, a measurement method and setup for evaluating the fog effect on 3D LiDARs will be proposed. Each presented measurement procedure has been tested. The obtained results highlight the versatility and the goodness of the proposed approaches. Regarding the precision agriculture sensors, a measurement approach for the Moisture Content and density estimation of crop directly on the field is presented. The approach regards the employment of a Near Infrared spectrometer jointly with Partial Least Square statistical analysis. The approach and the model will be described together with a first laboratory prototype used to evaluate the NIRS approach. Finally, a prototype for on the field analysis is realized and tested. The test results are promising, evidencing that the proposed approach is suitable for Moisture Content and density estimation.

Meta-learning for informed model searches in Automated Machine Learning

Day by day, machine learning is changing our lives in ways we could not have imagined just 5 years ago. ML expertise is more and more requested and needed, though just a limited number of ML engineers are available on the job market, and their knowledge is always limited by an inherent characteristic of theirs: they are humans. This thesis explores the possibilities offered by meta-learning, a new field in ML that takes learning a level higher: models are trained on other models' training data, starting from features of the dataset they were trained on, inference times, obtained performances, to try to understand the relationship between a good model and the way it was obtained. The so-called metamodel was trained on data collected by OpenML, the largest ML metadata platform that's publicly available today. Datasets were analyzed to obtain meta-features that describe them, which were then tied to model performances in a regression task. The obtained metamodel predicts the expected performances of a given model type (e.g., a random forest) on a given ML task (e.g., classification on the UCI census dataset). This research was then integrated into a custom-made AutoML framework, to show how meta-learning is not an end in itself, but it can be used to further progress our ML research. Encoding ML engineering expertise in a model allows better, faster, and more impactful ML applications across the whole world, while reducing the cost that is inevitably tied to human engineers.

Advancing Bag-of-visual-words Representations For Lesion Classification In Retinal Images.

Diabetic Retinopathy (DR) is a complication of diabetes that can lead to blindness if not readily discovered. Automated screening algorithms have the potential to improve identification of patients who need further medical attention. However, the identification of lesions must be accurate to be useful for clinical application. The bag-of-visual-words (BoVW) algorithm employs a maximum-margin classifier in a flexible framework that is able to detect the most common DR-related lesions such as microaneurysms, cotton-wool spots and hard exudates. BoVW allows to bypass the need for pre- and post-processing of the retinographic images, as well as the need of specific ad hoc techniques for identification of each type of lesion. An extensive evaluation of the BoVW model, using three large retinograph datasets (DR1, DR2 and Messidor) with different resolution and collected by different healthcare personnel, was performed. The results demonstrate that the BoVW classification approach can identify different lesions within an image without having to utilize different algorithms for each lesion reducing processing time and providing a more flexible diagnostic system. Our BoVW scheme is based on sparse low-level feature detection with a Speeded-Up Robust Features (SURF) local descriptor, and mid-level features based on semi-soft coding with max pooling. The best BoVW representation for retinal image classification was an area under the receiver operating characteristic curve (AUC-ROC) of 97.8% (exudates) and 93.5% (red lesions), applying a cross-dataset validation protocol. To assess the accuracy for detecting cases that require referral within one year, the sparse extraction technique associated with semi-soft coding and max pooling obtained an AUC of 94.2 ± 2.0%, outperforming current methods. Those results indicate that, for retinal image classification tasks in clinical practice, BoVW is equal and, in some instances, surpasses results obtained using dense detection (widely believed to be the best choice in many vision problems) for the low-level descriptors.

Nat2, Xrcc1 And Hogg1 Polymorphisms, Cigarette Smoking, Alcohol Consumption And Risk Of Upper Aerodigestive Tract Cancer.

To evaluate associations between polymorphisms of the N-acetyltransferase 2 (NAT2), human 8-oxoguanine glycosylase 1 (hOGG1) and X-ray repair cross-complementing protein 1 (XRCC1) genes and risk of upper aerodigestive tract (UADT) cancer. A case-control study involving 117 cases and 224 controls was undertaken. The NAT2 gene polymorphisms were genotyped by automated sequencing and XRCC1 Arg399Gln and hOGG1 Ser326Cys polymorphisms were determined by Polymerase Chain Reaction followed by Restriction Fragment Length Polymorphism (PCR-RFLP) methods. Slow metabolization phenotype was significantly associated as a risk factor for the development of UADT cancer (p=0.038). Furthermore, haplotype of slow metabolization was also associated with UADT cancer (p=0.014). The hOGG1 Ser326Cys polymorphism (CG or GG vs. CC genotypes) was shown as a protective factor against UADT cancer in moderate smokers (p=0.031). The XRCC1 Arg399Gln polymorphism (GA or AA vs. GG genotypes), in turn, was a protective factor against UADT cancer only among never-drinkers (p=0.048). Interactions involving NAT2, XRCC1 Arg399Gln and hOGG1 Ser326Cys polymorphisms may modulate the risk of UADT cancer in this population.

Yellowing And Bleaching Of Grey Hair Caused By Photo And Thermal Degradation.

Yellowing is an undesirable phenomenon that is common in people with white and grey hair. Because white hair has no melanin, the pigment responsible for hair colour, the effects of photodegradation are more visible in this type of hair. The origin of yellowing and its relation to photodegradation processes are not properly established, and many questions remain open in this field. In this work, the photodegradation of grey hair was investigated as a function of the wavelength of incident radiation, and its ultrastructure was determined, always comparing the results obtained for the white and black fibres present in grey hair with the results of white wool. The results presented herein indicate that the photobehaviour of grey hair irradiated with a mercury lamp or with solar radiation is dependent on the wavelength range of the incident radiation and on the initial shade of yellow in the sample. Two types of grey hair were used: (1) blended grey hair (more yellow) and (2) grey hair from a single-donor (less yellow). After exposure to a full-spectrum mercury lamp for 200 h, the blended white hair turned less yellow (the yellow-blue difference, Db(*) becomes negative, Db(*)=-6), whereas the white hair from the single-donor turned slightly yellower (Db(*)=2). In contrast, VIS+IR irradiation resulted in bleaching in both types of hair, whereas a thermal treatment (at 81 °C) caused yellowing of both types of hair, resulting in a Db(*)=3 for blended white hair and Db(*)=9 for single-donor hair. The identity of the yellow chromophores was investigated by UV-Vis spectroscopy. The results obtained with this technique were contradictory, however, and it was not possible to obtain a simple correlation between the sample shade of yellow and the absorption spectra. In addition, the results are discussed in terms of the morphology differences between the pigmented and non-pigmented parts of grey hair, the yellowing and bleaching effects of grey hair, and the occurrence of dark-follow reactions.

Newborn Hearing Screening And Early Diagnostic In The Nicu.

The aim was to describe the outcome of neonatal hearing screening (NHS) and audiological diagnosis in neonates in the NICU. The sample was divided into Group I: neonates who underwent NHS in one step and Group II: neonates who underwent a test and retest NHS. NHS procedure was automated auditory brainstem response. NHS was performed in 82.1% of surviving neonates. For GI, referral rate was 18.6% and false-positive was 62.2% (normal hearing in the diagnostic stage). In GII, with retest, referral rate dropped to 4.1% and false-positive to 12.5%. Sensorineural hearing loss was found in 13.2% of infants and conductive in 26.4% of cases. There was one case of auditory neuropathy spectrum (1.9%). Dropout rate in whole process was 21.7% for GI and 24.03% for GII. We concluded that it was not possible to perform universal NHS in the studied sample or, in many cases, to apply it within the first month of life. Retest reduced failure and false-positive rate and did not increase evasion, indicating that it is a recommendable step in NHS programs in the NICU. The incidence of hearing loss was 2.9%, considering sensorineural hearing loss (0.91%), conductive (1.83%) and auditory neuropathy spectrum (0.19%).

Detection Of Explosives On The Surface Of Banknotes By Raman Hyperspectral Imaging And Independent Component Analysis.

The aim of this study was to develop a methodology using Raman hyperspectral imaging and chemometric methods for identification of pre- and post-blast explosive residues on banknote surfaces. The explosives studied were of military, commercial and propellant uses. After the acquisition of the hyperspectral imaging, independent component analysis (ICA) was applied to extract the pure spectra and the distribution of the corresponding image constituents. The performance of the methodology was evaluated by the explained variance and the lack of fit of the models, by comparing the ICA recovered spectra with the reference spectra using correlation coefficients and by the presence of rotational ambiguity in the ICA solutions. The methodology was applied to forensic samples to solve an automated teller machine explosion case. Independent component analysis proved to be a suitable method of resolving curves, achieving equivalent performance with the multivariate curve resolution with alternating least squares (MCR-ALS) method. At low concentrations, MCR-ALS presents some limitations, as it did not provide the correct solution. The detection limit of the methodology presented in this study was 50μgcm(-2).

Characterization Of Anfo Explosive By High Accuracy Esi(±)-ftms With Forensic Identification On Real Samples By Easi(-)-ms.

Ammonium nitrate fuel oil (ANFO) is an explosive used in many civil applications. In Brazil, ANFO has unfortunately also been used in criminal attacks, mainly in automated teller machine (ATM) explosions. In this paper, we describe a detailed characterization of the ANFO composition and its two main constituents (diesel and a nitrate explosive) using high resolution and accuracy mass spectrometry performed on an FT-ICR-mass spectrometer with electrospray ionization (ESI(±)-FTMS) in both the positive and negative ion modes. Via ESI(-)-MS, an ion marker for ANFO was characterized. Using a direct and simple ambient desorption/ionization technique, i.e., easy ambient sonic-spray ionization mass spectrometry (EASI-MS), in a simpler, lower accuracy but robust single quadrupole mass spectrometer, the ANFO ion marker was directly detected from the surface of banknotes collected from ATM explosion theft.

Perfis temporais NDVI MODIS, na cana-soca, de maturação tardia

This paper describes the development of a relational database and a tool for viewing MODIS NDVI temporal profile, using data from MOD09Q1 product, specifically the surface bidirectional reflectance factor relative to the RED and NIR wavelength, mosaic of 8-day temporal composition, and the quality band, in sugarcane fields in the state of São Paulo, for analysis of the late stubble-cane maturation. From sugarcane farms were obtained the historical data about yield, soil, variety, location of the each pixel for each subregion monitored. All data were integrated in a database developed in PostgreSQL. The tool was implemented using Java language and allowed a fast and automatic way of analyzing sugarcane phenological patterns. It concluded that the MODIS NDVI temporal profile using data from MOD09Q1 product is able to subsidize the monitoring of phenological changes in the sugarcane.

Efeitos da constrição pupilar na perimetria de freqüência dupla em olhos normais

Purpose: To study the effects of pupillary constriction on frequency doubling perimetry in a group of normal subjects. Methods: Eighteen healthy volunteers participated in the study. Only one eye per patient (right eye) underwent frequency doubling perimetry (Full Threshold C-20 strategy). For the second session, one drop of 2% pilocarpine was administered to the volunteers' right eye and the examination was repeated after 60 minutes. Results: Sixty minutes after administration of 2% pilocarpine, there was a significant reduction of the pupillary diameter from 4.22 ± 0.17 mm to 1.55 ± 0.51 mm (p<0.05). There was a significant reduction of the mean retinal sensibility after pupillary constriction. The threshold sensitivity of the central 5º worsened by 5.67 ± 2.49 dB; the area between 2.5º and 10º worsened by 4.49 ± 2.73 dB; and the area between 10º and 20º worsened by 5.10 ± 3.55 dB (p<0.01). A reduction of 4.06 ± 2.67 dB was observed in the mean deviation, as well as an increase of 0.64 ± 0.94 dB in the pattern standard deviation (p<0.01). No differences were observed regarding the number of fixation losses, false-positive and false-negatives responses, and duration of the examination. Conclusion: Changes in pupillary diameter may produce significant declines in threshold sensitivities of the 20º visual field tested by frequency doubling perimetry. These results suggest that is important to maintain a constant pupillary diameter in seriate examinations.

Curva de aprendizado e flutuações a curto e longo prazos com perimetria de freqüência dupla (FDT)

OBJECTIVE: To evaluate the learning effect, short-term fluctuation and long-term fluctuation in healthy subjects undergoing frequency doubling perimetry (FDP). METHODS: Twenty healthy young subjects underwent FDT (program N30, full threshold) in one eye (right). Each subject was tested once in the first three sessions and three times in the fourth session. Both short- and long-term fluctuations were studied as the average fluctuation of all the tested points or as a point-to-point fluctuation. To study the learning effect, the MDs values of the first session were compared to the second, third and fourth sessions. RESULTS: In the short-term analysis (3 examination done in the last session), the total mean sensitivity was 31.91 ± 1.20 dB and the mean MD and PSD were 0.84 ± 1.85 and 3.73 ± 1.55 dB, respectively. The average short-term fluctuation was 1.72 ± 0.38 dB. When the four examination, performed at different visits, were compared, the average mean sensitivity of all sessions and the average long-term fluctuation were 31.75 ± 1.11 and 2.16 ± 0.26 dB, respectively. The MD averages of the first, second, third and fourth tests were 0.11 ± 2.14 dB, 0.47 ± 1.64 dB, 1.16 ± 1.62 dB and 0.98 ± 1.92 dB respectively. The MD difference between the first and the third and between the first and the fourth examinations were statistically significant (p<0.05). CONCLUSION: The threshold sensitivity detected by FDP is influenced by both short- and long-term fluctuations. We observed a mild learning effect that shoud be taken into account whenever a patient undergoes this test for the first time.

Interaction between staining and degradation of a composite resin in contact with colored foods

Composite resins might be susceptible to degradation and staining when in contact with some foods and drinks. This study evaluated color alteration and changes in microhardness of a microhybrid composite after immersion in different colored foods and determined whether there was a correlation between these two variables. Eighty composite disks were randomly divided into 8 experimental groups (n = 10): kept dry; deionized water; orange juice; passion fruit juice; grape juice; ketchup; mustard and soy sauce. The disks were individually immersed in their respective test substance at 37 ºC, for a period of 28 days. Superficial analysis of the disk specimens was performed by taking microhardness measurements (Vickers, 50 g load for 45 seconds) and color alterations were determined with a spectrophotometer (CINTRA 10- using a CIEL*a*b* system, 400-700 nm wavelength, illuminant d65 and standard observer of 2º) at the following times: baseline (before immersion), 1, 7, 14, 21 and 28 days. Results were analyzed by ANOVA and Tukey's test (p < 0.05). Both variables were also submitted to Pearson's correlation test (p < 0.05). The passion fruit group underwent the greatest microhardness change, while the mustard group suffered the greatest color alteration. Significant positive correlation was found between the two variables for the groups deionized water, grape juice, soy sauce and ketchup. Not all color alteration could be associated with surface degradation.

A method to determine volatile contaminants in polyethylene terephthalate (PET) packages by HDC-GC-FID and its application to post-consumer materials

A simple and low cost method to determine volatile contaminants in post-consumer recycled PET flakes was developed and validated by Headspace Dynamic Concentration and Gas Chromatography-Flame Ionization Detection (HDC-GC-FID). The analytical parameters evaluated by using surrogates include: correlation coefficient, detection limit, quantification limit, accuracy, intra-assay precision, and inter-assay precision. In order to compare the efficiency of the proposed method to recognized automated techniques, post-consumer PET packaging samples collected in Brazil were used. GC-MS was used to confirm the identity of the substances identified in the PET packaging. Some of the identified contaminants were estimated in the post-consumer material at concentrations higher than 220 ng.g-1. The findings in this work corroborate data available in the scientific literature pointing out the suitability of the proposed analytical method.

Numerical investigation of the three-dimensional secondary instabilities in the time-developing compressible mixing layer

Mixing layers are present in very different types of physical situations such as atmospheric flows, aerodynamics and combustion. It is, therefore, a well researched subject, but there are aspects that require further studies. Here the instability of two-and three-dimensional perturbations in the compressible mixing layer was investigated by numerical simulations. In the numerical code, the derivatives were discretized using high-order compact finite-difference schemes. A stretching in the normal direction was implemented with both the objective of reducing the sound waves generated by the shear region and improving the resolution near the center. The compact schemes were modified to work with non-uniform grids. Numerical tests started with an analysis of the growth rate in the linear regime to verify the code implementation. Tests were also performed in the non-linear regime and it was possible to reproduce the vortex roll-up and pairing, both in two-and three-dimensional situations. Amplification rate analysis was also performed for the secondary instability of this flow. It was found that, for essentially incompressible flow, maximum growth rates occurred for a spanwise wavelength of approximately 2/3 of the streamwise spacing of the vortices. The result demonstrated the applicability of the theory developed by Pierrehumbet and Widnall. Compressibility effects were then considered and the maximum growth rates obtained for relatively high Mach numbers (typically under 0.8) were also presented.