Francisco del Pozo

Applying Watershed Algorithms to the Segmentation of Clustered Nuclei

Cluster division is a critical issue in fluorescence microscopy-based analytical cytology when preparation protocols do not provide appropriate separation of objects. Overlooking clustered nuclei and analyzing only isolated nuclei may dramatically increase analysis time or affect the statistical validation of the results. Automatic segmentation of clustered nuclei requires the implementation of specific image segmentation tools. Most algorithms are inspired by one of the two following strategies: 1) cluster division by the detection of internuclei gradients; or 2) division by definition of domains of influence (geometrical approach). Both strategies lead to completely different implementations, and usually algorithms based on a single view strategy fail to correctly segment most clustered nuclei, or perform well just for a specific type of sample. An algorithm based on morphological watersheds has been implemented and tested on the segmentation of microscopic nuclei clusters. This algorithm provides a tool that can be used for the implementation of both gradient- and domain-based algorithms, and, more importantly, for the implementation of mixed (gradient- and shape-based) algorithms. Using this algorithm, almost 90% of the test clusters were correctly segmented in peripheral blood and bone marrow preparations. The algorithm was valid for both types of samples, using the appropriate markers and transformations.

Emergence of network features from multiplexity.

Many biological and man-made networked systems are characterized by the simultaneous presence of different sub-networks organized in separate layers, with links and nodes of qualitatively different types. While during the past few years theoretical studies have examined a variety of structural features of complex networks, the outstanding question is whether such features are characterizing all single layers, or rather emerge as a result of coarse-graining, i.e. when going from the multilayered to the aggregate network representation. Here we address this issue with the help of real data. We analyze the structural properties of an intrinsically multilayered real network, the European Air Transportation Multiplex Network in which each commercial airline defines a network layer. We examine how several structural measures evolve as layers are progressively merged together. In particular, we discuss how the topology of each layer affects the emergence of structural properties in the aggregate network.

Automated FISH spot counting in interphase nuclei: Statistical validation and data correction

The evaluation of an automated system for Fluorescence In Situ Hybridization (FISH) spot counting in interphase nuclei is presented in this paper. Different types of experiments have been performed with samples from known populations. In all of them the goal is to detect mosaicism of chromosome X in leukocytes from mixtures in known proportions of healthy male and female blood. First the initial results from the automatic FISH analysis system were obtained and evaluated. Then the analysis was modified to reduce systematic errors, so that the results are closer to what an experienced human operator would have obtained (system calibration step). Finally, an additional control probe of chromosome Y was used to detect and discard cells where incorrect hybridization or other abnormal situations had occurred. In each step the system sensitivity was determined by the use of two statistical validation tests, so that the improvement brought about by the correction methods could be assessed. The results obtained in the study showed that, using both corrections, the system is able to detect 10% monosomies with a significance level alpha = 0.1%.

A study of a rural telemedicine system in the Amazon region of Peru

Voice and data communication facilities (email via VHF radio) were installed in 39 previously isolated health facilities in the province of Alto Amazonas in Peru. A baseline study was carried out in January 2001 and a follow-up evaluation in May 2002, after nine months of operation. We measured the reliability of the technology and the effect the system had on staff access to medical training and information. We also measured the indirect effects on the general population of access to better health-care. The experimental data were collected from 35 of the 39 sites in face-to-face questionnaire interviews. Before installation of the system, the mean consultation rate was 3 per month per facility (95% CI 1.5 to 4.5). At the end of the study, the mean consultation rate was 23 per month per facility (95% CI 14.7 to 31.5). There were 205 emergency transfers from the 39 health facilities. The system was employed in all these cases to alert the referral centre. The mean time required for evacuation was reduced from 8.6 h to 5.2 h. Health-care personnel reported that in 58 of the emergency cases (28%) the use of the system saved the life of the patient. The study shows that the use of communication technologies appropriate to local needs solves many problems in rural primary care, and that voice and email communication via VHF radio are feasible and useful for rural telemedicine.

Optimizing Functional Network Representation of Multivariate Time Series

By combining complex network theory and data mining techniques, we provide objective criteria for optimization of the functional network representation of generic multivariate time series. In particular, we propose a method for the principled selection of the threshold value for functional network reconstruction from raw data, and for proper identification of the networks indicators that unveil the most discriminative information on the system for classification purposes. We illustrate our method by analysing networks of functional brain activity of healthy subjects, and patients suffering from Mild Cognitive Impairment, an intermediate stage between the expected cognitive decline of normal aging and the more pronounced decline of dementia. We discuss extensions of the scope of the proposed methodology to network engineering purposes, and to other data mining tasks.

Magnetoencephalography as a putative biomarker for Alzheimer's disease

Alzheimers Disease (AD) is the most common dementia in the elderly and is estimated to affect tens of millions of people worldwide. AD is believed to have a prodromal stage lasting ten or more years. While amyloid deposits, tau filaments, and loss of brain cells are characteristics of the disease, the loss of dendritic spines and of synapses predate such changes. Popular preclinical detection strategies mainly involve cerebrospinal fluid biomarkers, magnetic resonance imaging, metabolic PET scans, and amyloid imaging. One strategy missing from this list involves neurophysiological measures, which might be more sensitive to detect alterations in brain function. The Magnetoencephalography International Consortium of Alzheimers Disease arose out of the need to advance the use of Magnetoencephalography (MEG), as a tool in AD and pre-AD research. This paper presents a framework for using MEG in dementia research, and for short-term research priorities.

Hybrid stimulator for chronic experiments

A hybrid stimulator is described which combines constant current and constant voltage characteristics. Pulses are delivered with high output impedance to assure the passage of a predetermined amount of current. During the quiescent periods between pulses, the output changes to low impedance, diminishing the post-pulse charge. Electrode erosion and tissue damage are thus minimized, permitting safe, long term excitation.

Using surveillance data to estimate pandemic vaccine effectiveness against laboratory confirmed influenza A(H1N1)2009 infection: two case-control studies, Spain, season 2009-2010

BackgroundPhysicians of the Spanish Influenza Sentinel Surveillance System report and systematically swab patients attended to their practices for influenza-like illness (ILI). Within the surveillance system, some Spanish regions also participated in an observational study aiming at estimating influenza vaccine effectiveness (cycEVA study). During the season 2009-2010, we estimated pandemic influenza vaccine effectiveness using both the influenza surveillance data and the cycEVA study.MethodsWe conducted two case-control studies using the test-negative design, between weeks 48/2009 and 8/2010 of the pandemic season. The surveillance-based study included all swabbed patients in the sentinel surveillance system. The cycEVA study included swabbed patients from seven Spanish regions. Cases were laboratory-confirmed pandemic influenza A(H1N1)2009. Controls were ILI patients testing negative for any type of influenza. Variables collected in both studies included demographic data, vaccination status, laboratory results, chronic conditions, and pregnancy. Additionally, cycEVA questionnaire collected data on previous influenza vaccination, smoking, functional status, hospitalisations, visits to the general practitioners, and obesity. We used logistic regression to calculate adjusted odds ratios (OR), computing pandemic influenza vaccine effectiveness as (1-OR)*100.ResultsWe included 331 cases and 995 controls in the surveillance-based study and 85 cases and 351 controls in the cycEVA study. We detected nine (2.7%) and two (2.4%) vaccine failures in the surveillance-based and cycEVA studies, respectively. Adjusting for variables collected in surveillance database and swabbing month, pandemic influenza vaccine effectiveness was 62% (95% confidence interval (CI): -5; 87). The cycEVA vaccine effectiveness was 64% (95%CI: -225; 96) when adjusting for common variables with the surveillance system and 75% (95%CI: -293; 98) adjusting for all variables collected.ConclusionPoint estimates of the pandemic influenza vaccine effectiveness suggested a protective effect of the pandemic vaccine against laboratory-confirmed influenza A(H1N1)2009 in the season 2009-2010. Both studies were limited by the low vaccine coverage and the late start of the vaccination campaign. Routine influenza surveillance provides reliable estimates and could be used for influenza vaccine effectiveness studies in future seasons taken into account the surveillance system limitations.

Signal-to-noise ratio of the MEG signal after preprocessing

BACKGROUND Magnetoencephalography (MEG) provides a direct measure of brain activity with high combined spatiotemporal resolution. Preprocessing is necessary to reduce contributions from environmental interference and biological noise. NEW METHOD The effect on the signal-to-noise ratio of different preprocessing techniques is evaluated. The signal-to-noise ratio (SNR) was defined as the ratio between the mean signal amplitude (evoked field) and the standard error of the mean over trials. RESULTS Recordings from 26 subjects obtained during and event-related visual paradigm with an Elekta MEG scanner were employed. Two methods were considered as first-step noise reduction: Signal Space Separation and temporal Signal Space Separation, which decompose the signal into components with origin inside and outside the head. Both algorithm increased the SNR by approximately 100%. Epoch-based methods, aimed at identifying and rejecting epochs containing eye blinks, muscular artifacts and sensor jumps provided an SNR improvement of 5-10%. Decomposition methods evaluated were independent component analysis (ICA) and second-order blind identification (SOBI). The increase in SNR was of about 36% with ICA and 33% with SOBI. COMPARISON WITH EXISTING METHODS No previous systematic evaluation of the effect of the typical preprocessing steps in the SNR of the MEG signal has been performed. CONCLUSIONS The application of either SSS or tSSS is mandatory in Elekta systems. No significant differences were found between the two. While epoch-based methods have been routinely applied the less often considered decomposition methods were clearly superior and therefore their use seems advisable.

Intelligent alarms integrated in a multi-agent architecture for diabetes management

This paper describes the development of an intelligent agent that interprets blood glucose monitoring data received by a telemedicine-based diabetes management service. The agent generates automatic alarms when some deviations in the patient status are detected. The agent combines different methods to produce data summaries and automatic alarms based on statistics, rule-based techniques and model-based techniques. The rule-based analysis allows the detection of severe abnormalities using different time scales depending on the quality of the received information. The model-based analysis uses a physiological qualitative model implemented with a causal probabilistic network that detects deviations in the ‘insulin effectiveness’ along days. The KM (Knowledge Management) agent was tested with data from 11 patients with diabetes that used a telemedicine service during 1 year. The KM agent detected anomalous situations in the 100% of the cases where a therapy modification was decided by the healthcare professional; the agent detected abnormal data in 37% of transmissions, being able to decrease professionals’ workload due to telemedicine. The advantages of an automatic response integrated in a telemedicine service are that it focuses doctors’ and patients’ attention on abnormal data and gives instantaneous feedback to patients reinforcing the education and the motivation aspects of the therapy.