Improving Cardiology-Rehospitalization Prediction Through the Synergy of Process Mining and Deep Learning: An Innovative Approach.

Nowadays, hospitals are facing the need for an accurate prediction of rehospitalizations. Rehospitalizations, indeed, represent both a high financial burden for the hospital and a proxy measure of care quality. The current work aims to address such a problem with an innovative approach, by building a Process Mining-Deep Learning model for the prediction of 6-months rehospitalization of patients hospitalized in a Cardiology specialty at San Raffaele Hospital, starting from their medical history contained in the Patients Hospital Records, with the double purpose of supporting resource planning and identifying at-risk patients.

Diurnal changes in the efficiency of information transmission at a sensory synapse.

Neuromodulators adapt sensory circuits to changes in the external world or the animal's internal state and synapses are key control sites for such plasticity. Less clear is how neuromodulation alters the amount of information transmitted through the circuit. We investigated this question in the context of the diurnal regulation of visual processing in the retina of zebrafish, focusing on ribbon synapses of bipolar cells. We demonstrate that contrast-sensitivity peaks in the afternoon accompanied by a four-fold increase in the average Shannon information transmitted from an active zone. This increase reflects higher synaptic gain, lower spontaneous "noise" and reduced variability of evoked responses. Simultaneously, an increase in the probability of multivesicular events with larger information content increases the efficiency of transmission (bits per vesicle) by factors of 1.5-2.7. This study demonstrates the multiplicity of mechanisms by which a neuromodulator can adjust the synaptic transfer of sensory information.

Fighting healthcare rocketing costs with value-based medicine: the case of stroke management.

Value-Based Medicine (VBM) is imposing itself as 'a new paradigm in healthcare management and medical practice.In this perspective paper, we discuss the role of VBM in dealing with the large productivity issue of the healthcare industry and examine some of the worldwide industrial and technological trends linked with VBM introduction. To clarify the points, we discuss examples of VBM management of stroke patients.In our conclusions, we support the idea of VBM as a strategic aid to manage rising costs in healthcare, and we explore the idea that VBM, by establishing value-generating networks among different healthcare stakeholders, can serve as the long sought-after redistributive mechanism that compensate patients for the industrial exploitation of their personal medical records.

Targeting neuronal and glial cell types with synthetic promoter AAVs in mice, non-human primates and humans.

Targeting genes to specific neuronal or glial cell types is valuable for both understanding and repairing brain circuits. Adeno-associated viruses (AAVs) are frequently used for gene delivery, but targeting expression to specific cell types is an unsolved problem. We created a library of 230 AAVs, each with a different synthetic promoter designed using four independent strategies. We show that a number of these AAVs specifically target expression to neuronal and glial cell types in the mouse and non-human primate retina in vivo and in the human retina in vitro. We demonstrate applications for recording and stimulation, as well as the intersectional and combinatorial labeling of cell types. These resources and approaches allow economic, fast and efficient cell-type targeting in a variety of species, both for fundamental science and for gene therapy.

Log files analysis and evaluation of circadian patterns for the early diagnosis of pump thrombosis with a centrifugal continuous-flow left ventricular assist device.

BACKGROUND: No clinical standardized methods exist to identify the early stage of the development of pump thrombosis in the setting of HVAD (Medtronic Inc., USA) implantation. We aimed at developing a clinically relevant tool to evaluate HVAD operation during long-term support and at identifying a new reliable marker for the early diagnosis of pump thrombosis reflecting altered patient-pump physiological interplay. METHODS: We developed a novel algorithm based on time-frequency analysis of the HVAD log files allowing the detection of the intrinsic circadian rhythmicity of the pump power consumption. With this tool, we retrospectively evaluated (1) post-operative restoration of circadian rhythm (n = 14 patients), (2) long-term stability of circadian rhythmicity in patients with no reported adverse events (n = 12), and (3) alteration of circadian fluctuations in patients who suffered from pump thrombosis (n = 19). RESULTS: We demonstrate (1) progressive development of circadian rhythm following post-operative recovery (93% of the patients, 23 ± 15 days after implantation), (2) long-term stability of circadian rhythmicity in patients with no thrombotic complications (92% of the patients; 962 (445-1447) days of support), and (3) severe instability and loss of circadian fluctuations before the thrombotic event (89% of the patients, 12 ± 6 days ahead of the clinical manifestation of overt pump thrombosis). Furthermore, we provide the first clinical evidence of recovery of circadian rhythmicity following non-surgical resolution of pump thrombosis. CONCLUSIONS: Time-frequency analysis of the HVAD log files provides a new tool for the early diagnosis of pump thrombosis. Loss of circadian rhythmicity might trigger medical evaluation, improving the results of medical management of pump thrombosis, and decreasing the need for pump exchange.

Congenital Nystagmus Gene FRMD7 Is Necessary for Establishing a Neuronal Circuit Asymmetry for Direction Selectivity.

Neuronal circuit asymmetries are important components of brain circuits, but the molecular pathways leading to their establishment remain unknown. Here we found that the mutation of FRMD7, a gene that is defective in human congenital nystagmus, leads to the selective loss of the horizontal optokinetic reflex in mice, as it does in humans. This is accompanied by the selective loss of horizontal direction selectivity in retinal ganglion cells and the transition from asymmetric to symmetric inhibitory input to horizontal direction-selective ganglion cells. In wild-type retinas, we found FRMD7 specifically expressed in starburst amacrine cells, the interneuron type that provides asymmetric inhibition to direction-selective retinal ganglion cells. This work identifies FRMD7 as a key regulator in establishing a neuronal circuit asymmetry, and it suggests the involvement of a specific inhibitory neuron type in the pathophysiology of a neurological disease.

A synaptic mechanism for temporal filtering of visual signals.

The visual system transmits information about fast and slow changes in light intensity through separate neural pathways. We used in vivo imaging to investigate how bipolar cells transmit these signals to the inner retina. We found that the volume of the synaptic terminal is an intrinsic property that contributes to different temporal filters. Individual cells transmit through multiple terminals varying in size, but smaller terminals generate faster and larger calcium transients to trigger vesicle release with higher initial gain, followed by more profound adaptation. Smaller terminals transmitted higher stimulus frequencies more effectively. Modeling global calcium dynamics triggering vesicle release indicated that variations in the volume of presynaptic compartments contribute directly to all these differences in response dynamics. These results indicate how one neuron can transmit different temporal components in the visual signal through synaptic terminals of varying geometries with different adaptational properties.

Rapid mapping of visual receptive fields by filtered back projection: application to multi-neuronal electrophysiology and imaging.

Neurons in the visual system vary widely in the spatiotemporal properties of their receptive fields (RFs), and understanding these variations is key to elucidating how visual information is processed. We present a new approach for mapping RFs based on the filtered back projection (FBP), an algorithm used for tomographic reconstructions. To estimate RFs, a series of bars were flashed across the retina at pseudo-random positions and at a minimum of five orientations. We apply this method to retinal neurons and show that it can accurately recover the spatial RF and impulse response of ganglion cells recorded on a multi-electrode array. We also demonstrate its utility for in vivo imaging by mapping the RFs of an array of bipolar cell synapses expressing a genetically encoded Ca(2+) indicator. We find that FBP offers several advantages over the commonly used spike-triggered average (STA): (i) ON and OFF components of a RF can be separated; (ii) the impulse response can be reconstructed at sample rates of 125 Hz, rather than the refresh rate of a monitor; (iii) FBP reveals the response properties of neurons that are not evident using STA, including those that display orientation selectivity, or fire at low mean spike rates; and (iv) the FBP method is fast, allowing the RFs of all the bipolar cell synaptic terminals in a field of view to be reconstructed in under 4 min. Use of the FBP will benefit investigations of the visual system that employ electrophysiology or optical reporters to measure activity across populations of neurons.

Olfactory stimulation selectively modulates the OFF pathway in the retina of zebrafish.

Cross-modal regulation of visual performance by olfactory stimuli begins in the retina, where dopaminergic interneurons receive projections from the olfactory bulb. However, we do not understand how olfactory stimuli alter the processing of visual signals within the retina. We investigated this question by in vivo imaging activity in transgenic zebrafish expressing SyGCaMP2 in bipolar cell terminals and GCaMP3.5 in ganglion cells. The food-related amino acid methionine reduced the gain and increased sensitivity of responses to luminance and contrast transmitted through OFF bipolar cells but not ON. The effects of olfactory stimulus were blocked by inhibiting dopamine uptake and release. Activation of dopamine receptors increased the gain of synaptic transmission in vivo and potentiated synaptic calcium currents in isolated bipolar cells. These results indicate that olfactory stimuli alter the sensitivity of the retina through the dopaminergic regulation of presynaptic calcium channels that control the gain of synaptic transmission through OFF bipolar cells.

Optimization of a GCaMP calcium indicator for neural activity imaging.

Genetically encoded calcium indicators (GECIs) are powerful tools for systems neuroscience. Recent efforts in protein engineering have significantly increased the performance of GECIs. The state-of-the art single-wavelength GECI, GCaMP3, has been deployed in a number of model organisms and can reliably detect three or more action potentials in short bursts in several systems in vivo. Through protein structure determination, targeted mutagenesis, high-throughput screening, and a battery of in vitro assays, we have increased the dynamic range of GCaMP3 by severalfold, creating a family of "GCaMP5" sensors. We tested GCaMP5s in several systems: cultured neurons and astrocytes, mouse retina, and in vivo in Caenorhabditis chemosensory neurons, Drosophila larval neuromuscular junction and adult antennal lobe, zebrafish retina and tectum, and mouse visual cortex. Signal-to-noise ratio was improved by at least 2- to 3-fold. In the visual cortex, two GCaMP5 variants detected twice as many visual stimulus-responsive cells as GCaMP3. By combining in vivo imaging with electrophysiology we show that GCaMP5 fluorescence provides a more reliable measure of neuronal activity than its predecessor GCaMP3. GCaMP5 allows more sensitive detection of neural activity in vivo and may find widespread applications for cellular imaging in general.

Color vision: retinal blues.

Two complementary studies have resolved the circuitry underlying green-blue color discrimination in the retina. A blue-sensitive interneuron provides the inhibitory signal required for computing green-blue color opponency.

Spikes in retinal bipolar cells phase-lock to visual stimuli with millisecond precision.

BACKGROUND: The conversion of an analog stimulus into the digital form of spikes is a fundamental step in encoding sensory information. Here, we investigate this transformation in the visual system of fish by in vivo calcium imaging and electrophysiology of retinal bipolar cells, which have been assumed to be purely graded neurons. RESULTS: Synapses of all major classes of retinal bipolar cell encode visual information by using a combination of spikes and graded signals. Spikes are triggered within the synaptic terminal and, although sparse, phase-lock to a stimulus with a jitter as low as 2-3 ms. Spikes in bipolar cells encode a visual stimulus less reliably than spikes in ganglion cells but with similar temporal precision. The spike-generating mechanism does not alter the temporal filtering of a stimulus compared with the generator potential. The amplitude of the graded component of the presynaptic calcium signal can vary in time, and small fluctuations in resting membrane potential alter spike frequency and even switch spiking on and off. CONCLUSIONS: In the retina of fish, the millisecond precision of spike coding begins in the synaptic terminal of bipolar cells. This neural compartment regulates the frequency of digital signals transmitted to the inner retina as well as the strength of graded signals.

In vivo evidence that retinal bipolar cells generate spikes modulated by light.

Retinal bipolar cells have been assumed to generate purely graded responses to light. To test this idea we imaged the presynaptic calcium transient in live zebrafish. We found that ON, OFF, transient and sustained bipolar cells are all capable of generating fast 'all-or-none' calcium transients modulated by visual stimulation.

Novel image processing methods for the analysis of calcium dynamics in glial cells.

Calcium (Ca(2+)) waves and Ca(2+) oscillations within cells initiate a wide range of physiological processes including control of cell signaling, gene expression, secretion, and cell migration. A thorough analysis of Ca(2+) waves in glial cells provides information not only about the subcellular location of signaling processing events but also about nonneuronal or intercellular signaling pathways, their timing, routes, spatial domains, and coordination. In this study, three novel image processing methods have been applied to the study of Ca(2+) dynamics in cells. These bring additional information to the methods already available in the literature, providing insight into the analysis of calcium dynamics in fluorescence recordings and defining bidimensional maps that give a complete and detailed description of calcium intracellular behavior. The application of these processing methods to glial cells highlighted the complex 2-D Ca(2+) dynamics phenomena, the location of calcium uptake and release microdomains on the endoplasmic reticulum, and the correlation between different calcium signals inside the cell. A perinuclear zone acting as a filter and regulator of intracellular calcium waves was detected: it acts as a controller of calcium fluxes between the cytoplasm and the nucleus.

Fractal behavior of spontaneous neurotransmitter release: from single-synapse to whole-cell recordings.

Spontaneous release of neurotransmitter vesicles at brain chemical synapses has been deeply investigated in the last decades at several levels. First and second order statistics have been widely adopted as a tool for assessing, inter-alia, dependence of spontaneous release on the concentration of ionic species in the intra- or extra-cellular environment. Furthermore, several studies demonstrated that spontaneous release exhibits fractal, and generally non purely random, behavior. Most experimental work on this topic exploits population whole-cell patch-clamp recordings in order to acquire post-synaptic currents elicited by neurotransmitter release into the synaptic cleft. Since several synapses merge on the dendritic arbor of a single neuronal cell, whole-cell recordings of miniature excitatory post-synaptic currents (mEPSCs) implies the temporal superimposition of releasing events from all active synapses on the arbor. This limitation can be overcome by exploiting the loose-patch clamp technique on single synapses, thus acquiring spontaneous release events from individual synapses. Here, we present results obtained by applying well-established methods for the quantification of fractal behavior in the series of mEPSCs acquired through the use of both whole-cell and single-synapse loose-patch recording techniques on hippocampal neurons and synapses. Our long-term aim is to get a better understanding of the release process and of the mechanisms of neuronal integration when the information is coming from several simultaneously active synaptic sites.

Study of neuronal networks development from in-vitro recordings: A Granger causality based approach.

This methodological work is aimed at providing a Granger causality based approach to the study of neuronal networks development in vitro. The analysis procedure makes use of tools derived from statistics and network theory for accessing network development of in-vitro neuronal cultures from their electrical activity, recorded through Multi Electrode Arrays (MEAs). The preliminary results that will be presented here show the potential of this approach for characterizing in a quantitative way the developmental stages of neuronal networks and provide some evidences which are consistent with direct in-vitro and in-vivo observations reported by other authors.

Exploiting the multiplicative nature of fluoroscopic image stochastic noise to enhance calcium imaging recording quality.

One of the main problems that affect fluoroscopic imaging is the difficulty in coupling the recorded activity with the morphological information. The comprehension of fluorescence events in relationship with the internal structure of the cell can be very difficult. At this purpose, we developed a new method able to maximize the fluoroscopic movie quality. The method (Maximum Intensity Enhancement, MIE) works as follow: considering all the frames that compose the fluoroscopic movie, the algorithm extracts, for each pixel of the matrix, the maximal brightness value assumed along all the frames. Such values are collected in a maximum intensity matrix. Then, the method provides the projection of the target molecule oscillations which are present in the DeltaF/F(0) movie onto the maximum intensity matrix. This is done by creating a RGB movie and by assigning to the normalized (DeltaF/F(0)) activity a single channel and by reproducing the maximum intensity matrix on all the frames by using the remaining color channels. The application of such a method to fluoroscopic calcium imaging of astrocyte cultures demonstrated a meaningful enhancement in the possibility to discern the internal and external structure of cells.

Statistical long-term correlations in dissociated cortical neuron recordings.

The study of nonlinear long-term correlations in neuronal signals is a central topic for advanced neural signal processing. In particular, the existence of long-term correlations in neural signals recorded via multielectrode array (MEA) could provide interesting information about changes in interneuron communications. In this study we propose a new method for long-term correlation analysis of neuronal burst activity based on the periodogram alpha slope estimation of the MEA signal. We applied our method to recordings taken from cultured networks of dissociated rat cortical neurons. We show the effectiveness of the method in analyzing the activity changes as well as the temporal dynamics that take place during the development of such cultures. Results demonstrate that the alpha parameter is able to divide the network development in three well-defined stages, showing pronounced variations in the long-term correlation among bursts.

A blind method for the estimation of the Hurst exponent in time series: theory and application.

Nowadays many methods for the estimation of self-similarity (Hurst coefficient, H) in time series are available. Most of them, even if very effective, need some a priori information to be applied. We analyzed the eight most used methods for H estimation (working both in time and in frequency). We tested these methods on data generated with four kinds of time series models (fBm and fGn generated iteratively with Feder algorithm, 1f(alpha), and the fractional autoregressive integrated moving-average) in the range 0.1

The estimation of long-term memory characteristics in MEA neuronal culture recordings.

The nonlinear analysis of multichannel MEA recordings from neuronal networks is becoming a central topic in Neuroengineering. Up-to-date these kind of analyses required complex ad hoc methods. In this paper we introduce a new approach that allows the analysis of the whole-neuronal-network-activity with well-established nonlinear signal processing methods. In particular, we show here the estimation of long-term-memory behaviors through the Periodogram method in the bursting activity of cortical neuron cultures during development. Moreover, we show how this method is able to highlight structural activity changes of the network.