Automatic eye blink detection using consumer web cameras

This research aims to advance blinking detection in the context of work activity. Rather than patients having to attend a clinic, blinking videos can be acquired in a work environment, and further automatically analyzed. Therefore, this paper presents a methodology to perform the automatic detection of eye blink using consumer videos acquired with low-cost web cameras. This methodology includes the detection of the face and eyes of the recorded person, and then it analyzes the low-level features of the eye region to create a quantitative vector. Finally, this vector is classified into one of the two categories considered —open and closed eyes— by using machine learning algorithms. The effectiveness of the proposed methodology was demonstrated since it provides unbiased results with classification errors under 5%

Choline-PET in prostate cancer management: The point of view of the radiation oncologist.

Among PET radiotracers, FDG seems to be quite accepted as an accurate oncology diagnostic tool, frequently helpful also in the evaluation of treatment response and in radiation therapy treatment planning for several cancer sites. To the contrary, the reliability of Choline as a tracer for prostate cancer (PC) still remains an object of debate for clinicians, including radiation oncologists. This review focuses on the available data about the potential impact of Choline-PET in the daily clinical practice of radiation oncologists managing PC patients. In summary, routine Choline-PET is not indicated for initial local T staging, but it seems better than conventional imaging for nodal staging and for all patients with suspected metastases. In these settings, Choline-PET showed the potential to change patient management. A critical limit remains spatial resolution, limiting the accuracy and reliability for small lesions. After a PSA rise, the problem of the trigger PSA value remains crucial. Indeed, the overall detection rate of Choline-PET is significantly increased when the trigger PSA, or the doubling time, increases, but higher PSA levels are often a sign of metastatic spread, a contraindication for potentially curable local treatments such as radiation therapy. Even if several published data seem to be promising, the current role of PET in treatment planning in PC patients to be irradiated still remains under investigation. Based on available literature data, all these issues are addressed and discussed in this review.

Routine use of point-of-care tests: usefulness and application in clinical microbiology.

Point-of-care (POC) tests offer potentially substantial benefits for the management of infectious diseases, mainly by shortening the time to result and by making the test available at the bedside or at remote care centres. Commercial POC tests are already widely available for the diagnosis of bacterial and viral infections and for parasitic diseases, including malaria. Infectious diseases specialists and clinical microbiologists should be aware of the indications and limitations of each rapid test, so that they can use them appropriately and correctly interpret their results. The clinical applications and performance of the most relevant and commonly used POC tests are reviewed. Some of these tests exhibit insufficient sensitivity, and should therefore be coupled to confirmatory tests when the results are negative (e.g. Streptococcus pyogenes rapid antigen detection test), whereas the results of others need to be confirmed when positive (e.g. malaria). New molecular-based tests exhibit better sensitivity and specificity than former immunochromatographic assays (e.g. Streptococcus agalactiae detection). In the coming years, further evolution of POC tests may lead to new diagnostic approaches, such as panel testing, targeting not just a single pathogen, but all possible agents suspected in a specific clinical setting. To reach this goal, the development of serology-based and/or molecular-based microarrays/multiplexed tests will be needed. The availability of modern technology and new microfluidic devices will provide clinical microbiologists with the opportunity to be back at the bedside, proposing a large variety of POC tests that will allow quicker diagnosis and improved patient care.

NEW METHODS FOR DETECTION, SUSCEPTIBILITY TESTING AND BIOFILM ERADICATION OF DIFFICULT ORGANISMS

Aujourd'hui, les problèmes des maladies infectieuses concernent l'émergence d'infections difficiles à traiter, telles que les infections associées aux implants et les infections fongiques invasives chez les patients immunodéprimés. L'objectif de cette thèse était de développer des stratégies pour l'éradication des biofilms bactériens (partie 1), ainsi que d'étudier des méthodes innovantes pour la détection microbienne, pour l'établissement de nouveaux tests de sensibilité (partie 2). Le traitement des infections associées aux implants est difficile car les biofilms bactériens peuvent résister à des niveaux élevés d'antibiotiques. A ce jour, il n'y a pas de traitement optimal défini contre des infections causées par des bactéries de prévalence moindre telles que Enterococcus faecalis ou Propionibacterium acnés. Dans un premier temps, nous avons démontré une excellente activité in vitro de la gentamicine sur une souche de E. faecalis en phase stationnaire de croissance Nous avons ensuite confirmé l'activité de la gentamicine sur un biofilm précoce en modèle expérimental animal à corps étranger avec un taux de guérison de 50%. De plus, les courbes de bactéricidie ainsi que les résultats de calorimétrie ont prouvé que l'ajout de gentamicine améliorait l'activité in vitro de la daptomycine, ainsi que celle de la vancomycine. In vivo, le schéma thérapeutique le plus efficace était l'association daptomycine/gentamicine avec un taux de guérison de 55%. En établissant une nouvelle méthode pour l'évaluation de l'activité des antimicrobiens vis-à-vis de micro-organismes en biofilm, nous avons démontré que le meilleur antibiotique actif sur les biofilms à P. acnés était la rifampicine, suivi par la penicilline G, la daptomycine et la ceftriaxone. Les études conduites en modèle expérimental animal ont confirmé l'activité de la rifampicine seule avec un taux de guérison 36%. Le meilleur schéma thérapeutique était au final l'association rifampicine/daptomycine avec un taux de guérison 63%. Les associations de rifampicine avec la vancomycine ou la levofloxacine présentaient des taux de guérisons respectivement de 46% et 25%. Nous avons ensuite étudié l'émergence in vitro de la résistance à la rifampicine chez P. acnés. Nous avons observé un taux de mutations de 10"9. La caractérisation moléculaire de la résistance chez les mutant-résistants a mis en évidence l'implication de 5 mutations ponctuelles dans les domaines I et II du gène rpoB. Ce type de mutations a déjà été décrit au préalable chez d'autres espèces bactériennes, corroborant ainsi la validité de nos résultats. La deuxième partie de cette thèse décrit une nouvelle méthode d'évaluation de l'efficacité des antifongiques basée sur des mesures de microcalorimétrie isotherme. En utilisant un microcalorimètre, la chaleur produite par la croissance microbienne peut être-mesurée en temps réel, très précisément. Nous avons évalué l'activité de l'amphotéricine B, des triazolés et des échinocandines sur différentes souches de Aspergillus spp. par microcalorimétrie. La présence d'amphotéricine Β ou de triazole retardait la production de chaleur de manière concentration-dépendante. En revanche, pour les échinochandines, seule une diminution le pic de « flux de chaleur » a été observé. La concordance entre la concentration minimale inhibitrice de chaleur (CMIC) et la CMI ou CEM (définie par CLSI M38A), avec une marge de 2 dilutions, était de 90% pour l'amphotéricine B, 100% pour le voriconazole, 90% pour le pozoconazole et 70% pour la caspofongine. La méthode a été utilisée pour définir la sensibilité aux antifongiques pour d'autres types de champignons filamenteux. Par détermination microcalorimétrique, l'amphotéricine B s'est avéré être l'agent le plus actif contre les Mucorales et les Fusarium spp.. et le voriconazole le plus actif contre les Scedosporium spp. Finalement, nous avons évalué l'activité d'associations d'antifongiques vis-à-vis de Aspergillus spp. Une meilleure activité antifongique était retrouvée avec l'amphotéricine B ou le voriconazole lorsque ces derniers étaient associés aux échinocandines vis-à-vis de A. fumigatus. L'association échinocandine/amphotéricine B a démontré une activité antifongique synergique vis-à-vis de A. terreus, contrairement à l'association échinocandine/voriconazole qui ne démontrait aucune amélioration significative de l'activité antifongique. - The diagnosis and treatment of infectious diseases are today increasingly challenged by the emergence of difficult-to-manage situations, such as infections associated with medical devices and invasive fungal infections, especially in immunocompromised patients. The aim of this thesis was to address these challenges by developing new strategies for eradication of biofilms of difficult-to-treat microorganisms (treatment, part 1) and investigating innovative methods for microbial detection and antimicrobial susceptibility testing (diagnosis, part 2). The first part of the thesis investigates antimicrobial treatment strategies for infections caused by two less investigated microorganisms, Enterococcus faecalis and Propionibacterium acnes, which are important pathogens causing implant-associated infections. The treatment of implant-associated infections is difficult in general due to reduced susceptibility of bacteria when present in biofilms. We demonstrated an excellent in vitro activity of gentamicin against E. faecalis in stationary growth- phase and were able to confirm the activity against "young" biofilms (3 hours) in an experimental foreign-body infection model (cure rate 50%). The addition of gentamicin improved the activity of daptomycin and vancomycin in vitro, as determined by time-kill curves and microcalorimetry. In vivo, the most efficient combination regimen was daptomycin plus gentamicin (cure rate 55%). Despite a short duration of infection, the cure rates were low, highlighting that enterococcal biofilms remain difficult to treat despite administration of newer antibiotics, such as daptomycin. By establishing a novel in vitro assay for evaluation of anti-biofilm activity (microcalorimetry), we demonstrated that rifampin was the most active antimicrobial against P. acnes biofilms, followed by penicillin G, daptomycin and ceftriaxone. In animal studies we confirmed the anti-biofilm activity of rifampin (cure rate 36% when administered alone), as well as in combination with daptomycin (cure rate 63%), whereas in combination with vancomycin or levofloxacin it showed lower cure rates (46% and 25%, respectively). We further investigated the emergence of rifampin resistance in P. acnes in vitro. Rifampin resistance progressively emerged during exposure to rifampin, if the bacterial concentration was high (108 cfu/ml) with a mutation rate of 10"9. In resistant isolates, five point mutations of the rpoB gene were found in cluster I and II, as previously described for staphylococci and other bacterial species. The second part of the thesis describes a novel real-time method for evaluation of antifungals against molds, based on measurements of the growth-related heat production by isothermal microcalorimetry. Current methods for evaluation of antifungal agents against molds, have several limitations, especially when combinations of antifungals are investigated. We evaluated the activity of amphotericin B, triazoles (voriconazole, posaconazole) and echinocandins (caspofungin and anidulafungin) against Aspergillus spp. by microcalorimetry. The presence of amphotericin Β or a triazole delayed the heat production in a concentration-dependent manner and the minimal heat inhibition concentration (MHIC) was determined as the lowest concentration inhibiting 50% of the heat produced at 48 h. Due to the different mechanism of action echinocandins, the MHIC for this antifungal class was determined as the lowest concentration lowering the heat-flow peak with 50%. Agreement within two 2-fold dilutions between MHIC and MIC or MEC (determined by CLSI M38A) was 90% for amphotericin B, 100% for voriconazole, 90% for posaconazole and 70% for caspofungin. We further evaluated our assay for antifungal susceptibility testing of non-Aspergillus molds. As determined by microcalorimetry, amphotericin Β was the most active agent against Mucorales and Fusarium spp., whereas voriconazole was the most active agent against Scedosporium spp. Finally, we evaluated the activity of antifungal combinations against Aspergillus spp. Against A. jumigatus, an improved activity of amphotericin Β and voriconazole was observed when combined with an echinocandin. Against A. terreus, an echinocandin showed a synergistic activity with amphotericin B, whereas in combination with voriconazole, no considerable improved activity was observed.

Intestinal content detection in capsule endoscopy using robust features

This work covers two aspects. First, it generally compares and summarizes the similarities and differences of state of the art feature detector and descriptor and second it presents a novel approach of detecting intestinal content (in particular bubbles) in capsule endoscopy images. Feature detectors and descriptors providing invariance to change of perspective, scale, signal-noise-ratio and lighting conditions are important and interesting topics in current research and the number of possible applications seems to be numberless. After analysing a selection of in the literature presented approaches, this work investigates in their suitability for applications information extraction in capsule endoscopy images. Eventually, a very good performing detector of intestinal content in capsule endoscopy images is presented. A accurate detection of intestinal content is crucial for all kinds of machine learning approaches and other analysis on capsule endoscopy studies because they occlude the field of view of the capsule camera and therefore those frames need to be excluded from analysis. As a so called “byproduct” of this investigation a graphical user interface supported Feature Analysis Tool is presented to execute and compare the discussed feature detectors and descriptor on arbitrary images, with configurable parameters and visualized their output. As well the presented bubble classifier is part of this tool and if a ground truth is available (or can also be generated using this tool) a detailed visualization of the validation result will be performed.

Detection of suspected placental invasion by MRI: Do the results depend on observer' experience?

PURPOSE: To evaluate the diagnostic value of previously described MR features used for detecting suspected placental invasion according to observers' experience. MATERIALS AND METHODS: Our population included 25 pregnant women (mean age 35.16) investigated by prenatal MRI (1.5T, T1- and T2-weighted MR-sequences without i.v. contrast), among them 12 with histopathologically proven placental invasion and 13 women (52%) without placental invasion used as control group. Two senior and two junior radiologists blindly and independently reviewed MR-examinations in view of 6 previously defined MR-features indicating presence and degree of placental invasion (placenta increta, accreta or percreta). For each reader the sensibility, specificity, and receiver operating curve (ROC) were calculated. Interobserver agreements between senior and junior readers were determined. Stepwise logistic regression was performed including the 6 MR-features predictive of placental invasion. RESULTS: Demographics between both groups were statistically equivalent. Overall sensitivity and specificity for placental invasion was 90.9% and 75.0% for seniors and 81.8% and 61.8% for juniors, respectively. The best single MR-feature indicating placental invasion was T2-hypointense placental bands (r(2)=0.28), followed by focally interrupted myometrial border, infiltration of pelvic organs and tenting of the bladder (r(2)=0.36). Interobserver agreement for detecting placental invasion was 0.64 for seniors and 0.41 for juniors, thus substantial and moderate, respectively. Seniors detected placental invasion and depth of infiltration with significantly higher diagnostic certitude than juniors (p=0.0002 and p=0.0282, respectively). CONCLUSION: MRI can be a reliable and reproducible tool for the detection of suspected placental invasion, but the diagnostic value significantly depends on observers' experience.

Laboratory detection of respiratory viruses by automated techniques.

Advances in clinical virology for detecting respiratory viruses have been focused on nucleic acids amplification techniques, which have converted in the reference method for the diagnosis of acute respiratory infections of viral aetiology. Improvements of current commercial molecular assays to reduce hands-on-time rely on two strategies, a stepwise automation (semi-automation) and the complete automation of the whole procedure. Contributions to the former strategy have been the use of automated nucleic acids extractors, multiplex PCR, real-time PCR and/or DNA arrays for detection of amplicons. Commercial fully-automated molecular systems are now available for the detection of respiratory viruses. Some of them could convert in point-of-care methods substituting antigen tests for detection of respiratory syncytial virus and influenza A and B viruses. This article describes laboratory methods for detection of respiratory viruses. A cost-effective and rational diagnostic algorithm is proposed, considering technical aspects of the available assays, infrastructure possibilities of each laboratory and clinic-epidemiologic factors of the infection.

Utility of the mini-cog for detection of cognitive impairment in primary care: data from two spanish studies.

Objectives. To study the utility of the Mini-Cog test for detection of patients with cognitive impairment (CI) in primary care (PC). Methods. We pooled data from two phase III studies conducted in Spain. Patients with complaints or suspicion of CI were consecutively recruited by PC physicians. The cognitive diagnosis was performed by an expert neurologist, after formal neuropsychological evaluation. The Mini-Cog score was calculated post hoc, and its diagnostic utility was evaluated and compared with the utility of the Mini-Mental State (MMS), the Clock Drawing Test (CDT), and the sum of the MMS and the CDT (MMS + CDT) using the area under the receiver operating characteristic curve (AUC). The best cut points were obtained on the basis of diagnostic accuracy (DA) and kappa index. Results. A total sample of 307 subjects (176 CI) was analyzed. The Mini-Cog displayed an AUC (±SE) of 0.78 ± 0.02, which was significantly inferior to the AUC of the CDT (0.84 ± 0.02), the MMS (0.84 ± 0.02), and the MMS + CDT (0.86 ± 0.02). The best cut point of the Mini-Cog was 1/2 (sensitivity 0.60, specificity 0.90, DA 0.73, and kappa index 0.48 ± 0.05). Conclusions. The utility of the Mini-Cog for detection of CI in PC was very modest, clearly inferior to the MMS or the CDT. These results do not permit recommendation of the Mini-Cog in PC.

A survey strategy for human respiratory syncytial virus detection among haematopoietic stem cell transplant patients: epidemiological and methodological analysis

Human respiratory syncytial virus (HRSV) causes severe infections among children and immunocompromised patients. We compared HRSV infections among Haematopoietic Stem Cell Transplant program (HSCT) patients and children using direct immunofluorescence (DFA), point-of-care RSV Bio Easy® and a polymerase chain reaction (PCR) assay. Overall, 102 samples from HSCT patients and 128 from children obtained positivity rate of 18.6% and 14.1% respectively. PCR sensitivity was highest mainly on samples collected after five days of symptoms onset. A combination of both DFA and reverse transcriptase-PCR methods for HSCT high-risk patients is the best diagnostic flow for HRSV diagnosis among these patients.

Detection of the B"-GWGR variant in the southernmost region of Brazil: unveiling the complexity of the human immunodeficiency virus-1 subtype B epidemic

Typical human immunodeficiency virus-1 subtype B (HIV-1B) sequences present a GPGR signature at the tip of the variable region 3 (V3) loop; however, unusual motifs harbouring a GWGR signature have also been isolated. Although epidemiological studies have detected this variant in approximately 17-50% of the total infections in Brazil, the prevalence of B"-GWGR in the southernmost region of Brazil is not yet clear. This study aimed to investigate the C2-V3 molecular diversity of the HIV-1B epidemic in southernmost Brazil. HIV-1 seropositive patients were ana-lysed at two distinct time points in the state of Rio Grande do Sul (RS98 and RS08) and at one time point in the state of Santa Catarina (SC08). Phylogenetic analysis classified 46 individuals in the RS98 group as HIV-1B and their molecular signatures were as follows: 26% B"-GWGR, 54% B-GPGR and 20% other motifs. In the RS08 group, HIV-1B was present in 32 samples: 22% B"-GWGR, 59% B-GPGR and 19% other motifs. In the SC08 group, 32 HIV-1B samples were found: 28% B"-GWGR, 59% B-GPGR and 13% other motifs. No association could be established between the HIV-1B V3 signatures and exposure categories in the HIV-1B epidemic in RS. However, B-GPGR seemed to be related to heterosexual individuals in the SC08 group. Our results suggest that the established B"-GWGR epidemics in both cities have similar patterns, which is likely due to their geographical proximity and cultural relationship.

Forest fires cluster detection with space-time scan statistics

Forest fires are defined as uncontrolled fires often occurring in wildland areas, but that can also affect houses or agricultural resources. Causes are both natural (e.g.,lightning phenomena) and anthropogenic (human negligence or arsons).Major environmental factors influencing the fire ignition and propagation are climate and vegetation. Wildfires are most common and severe during drought period and on windy days. Moreover, under water-stress conditions, which occur after a long hot and dry period, the vegetation is more vulnerable to fire. These conditions are common in the United State and Canada, where forest fires represent a big problem. We focused our analysis on the state of Florida, for which a big dataset on forest fires detection is readily available. USDA Forest Service Remote Sensing Application Center, in collaboration with NASA-Goddard Space Flight Center and the University of Maryland, has compiled daily MODIS Thermal Anomalies (fires and biomass burning images) produced by NASA using a contextual algorithm that exploits the strong emission of mid-infrared radiation from fires. Fire classes were converted in GIS format: daily MODIS fire detections are provided as the centroids of the 1 kilometer pixels and compiled into daily Arc/INFO point coverage.

Multisensor change detection with nonlinear canonical correlation

The analysis of multi-modal and multi-sensor images is nowadays of paramount importance for Earth Observation (EO) applications. There exist a variety of methods that aim at fusing the different sources of information to obtain a compact representation of such datasets. However, for change detection existing methods are often unable to deal with heterogeneous image sources and very few consider possible nonlinearities in the data. Additionally, the availability of labeled information is very limited in change detection applications. For these reasons, we present the use of a semi-supervised kernel-based feature extraction technique. It incorporates a manifold regularization accounting for the geometric distribution and jointly addressing the small sample problem. An exhaustive example using Landsat 5 data illustrates the potential of the method for multi-sensor change detection.

Multi-scale label interaction to improve object detection and segmentation

The project aims at advancing the state of the art in the use of context information for classification of image and video data. The use of context in the classification of images has been showed of great importance to improve the performance of actual object recognition systems. In our project we proposed the concept of Multi-scale Feature Labels as a general and compact method to exploit the local and global context. The feature extraction from the discriminative probability or classification confidence label field is of great novelty. Moreover the use of a multi-scale representation of the feature labels lead to a compact and efficient description of the context. The goal of the project has been also to provide a general-purpose method and prove its suitability in different image/video analysis problem. The two-year project generated 5 journal publications (plus 2 under submission), 10 conference publications (plus 2 under submission) and one patent (plus 1 pending). Of these publications, a relevant number make use of the main result of this project to improve the results in detection and/or segmentation of objects.

Automatic Detection of Non-deverbal Event Nouns for Quick Lexicon Production

In this work we present the results of experimental work on the development of lexical class-based lexica by automatic means. Our purpose is to assess the use of linguistic lexical-class based information as a feature selection methodology for the use of classifiers in quick lexical development. The results show that the approach can help reduce the human effort required in the development of language resources significantly.

Faint arc-minute extended radio emission around Cygnus X-3

Aims.We revisit the vicinity of the microquasar Cygnus X-3 at radio wavelengths. We aim to improve our previous search for possible associated extended radio features/hot spots in the position angle of the Cygnus X-3 relativistic jets focusing on shorter angular scales than previously explored. Methods.Our work is mostly based on analyzing modern survey and archive radio data, mainly including observations carried out with the Very Large Array and the Ryle Telescopes. We also used deep near-infrared images that we obtained in 2005. Results.We present new radio maps of the Cygnus X-3 field computed after combining multi-configuration Very Large Array archive data at 6 cm and different observing runs at 2 cm with the Ryle Telescope. These are probably among the deepest radio images of Cygnus X-3 reported to date at cm wavelengths. Both interferometers reveal an extended radio feature within a few arc-minutes of the microquasar position, thus making our detection more credible. Moreover, this extended emission is possibly non-thermal, although this point still needs confirmation. Its physical connection with the microquasar is tentatively considered under different physical scenarios. We also report on the serendipitous discovery of a likely Fanaroff-Riley type II radio galaxy only away from Cygnus X-3.