1,721,193 research outputs found
Spatial graphs and Convolutive Models
No full textIn the last two decades, many complex systems have benefited from the use of graph theory, and these approaches have shown robust applicability in the field of finance, computer circuits and in biological systems. Large scale models of brain systems make also a great use of random graph models. Graph theory can be instrumental in modeling the connectivity and spatial distribution of neurons, through a characterization of the relative topological properties. However, all approaches in studying brain function have been so far limited to use experimental constraints obtained at a macroscopic level (e.g. fMRI, EEG, MEG, DTI, DSI). In this contribution, we present a microscopic use (i.e. at the single neuron level) of graph theory to introduce a new model, which we call spatial convolutive model (SCM). Such a model is able to merge random graphs and Power Law models in such a way to quantitatively reproduce the topological and spatial connection distributions observed in real systems
The role of network connectivity on epileptiform activity
Full text linkA number of potentially important mechanisms have been identified as key players to generate epileptiform activity, such as genetic mutations, activity-dependent alteration of synaptic functions, and functional network reorganization at the macroscopic level. Here we study how network connectivity at cellular level can affect the onset of epileptiform activity, using computational model networks with different wiring properties. The model suggests that networks connected as in real brain circuits are more resistant to generate seizure-like activity. The results suggest new experimentally testable predictions on the cellular network connectivity in epileptic individuals, and highlight the importance of using the appropriate network connectivity to investigate epileptiform activity with computational models
Retinal image synthesis through the least action principle
No full textEye fundus image analysis is a fundamental approach in medical diagnosis and follow-up ophthalmic diagnostics. Manual annotation by experts needs hard work, thus only a small set of annotated vessel structures is available. Examples such as DRIVE and STARE include small sets for training images of fundus image benchmarks. Moreover, there is no vessel structure annotation for a number of fundus image datasets. Synthetic images have been generated by using appropriate parameters for the modeling of vascular networks or by methods developing deep learning techniques and supported by performance hardware. Our methodology aims to produce high-resolution synthetic fundus images alternative to the increasing use of generative adversarial networks, to overcome the problems that arise in producing slightly modified versions of the same real images, to simulate pathologies and for the prediction of eye-related diseases. Our approach is based on the principle of the least action to place vessels on the simulated eye fundus
Graph-theoretical derivation of brain structural connectivity
Full text linkBrain connectivity at the single neuron level can provide fundamental insights into how information is integrated and propagated within and between brain regions. However, it is almost impossible to adequately study this problem experimentally and, despite intense efforts in the field, no mathematical description has been obtained so far. Here, we present a mathematical framework based on a graph-theoretical approach that, starting from experimental data obtained from a few small subsets of neurons, can quantitatively explain and predict the corresponding full network properties. This model also changes the paradigm with which large-scale model networks can be built, from using probabilistic/empiric connections or limited data, to a process that can algorithmically generate neuronal networks connected as in the real system
A Non-Parametric Parallel Harris-Affine Detector
No full textThis paper describes a parallel version
of a new automatic Harris-based corner
detector. A simple but effective client-server based
scheduler has been implemented in order to
dynamically distribute the workload on heterogeneous
parallel architectures such as Grid systems.
Results obtained on the COMETA Grid show the
effectiveness and the robustness of the proposed
approach
A visual framework to create photorealistic retinal vessels for diagnosis purposes
No full textThe methods developed in recent years for synthesising an ocular fundus can be been divided into two main categories. The first category of methods involves the development of an anatomical model of the eye, where artificial images are generated using appropriate parameters for modelling the vascular networks and fundus. The second type of method has been made possible by the development of deep learning techniques and improvements in the performance of hardware (especially graphics cards equipped with a large number of cores). The methodology proposed here to produce high-resolution synthetic fundus images is intended to be an alternative to the increasingly widespread use of generative adversarial networks to overcome the problems that arise in producing slightly modified versions of the same real images. This will allow the simulation of pathologies and the prediction of eye-related diseases. The proposed approach is based on the principle of least action and correctly places the vessels on the simulated eye fundus without using real morphometric information. An a posteriori analysis of the average characteristics such as the size, length, bifurcations, and endpoint positioning confirmed the substantial accuracy of the proposed approach compared to real data. A graphical user interface allows the user to make any changes in real time by controlling the positions of control points
Exudates as landmarks identified through fcm clustering in retinal images
Full text linkThe aim of this work was to develop a method for the automatic identification of exudates, using an unsupervised clustering approach. The ability to classify each pixel as belonging to an eventual exudate, as a warning of disease, allows for the tracking of a patient’s status through a noninvasive approach. In the field of diabetic retinopathy detection, we considered four public domain datasets (DIARETDB0/1, IDRID, and e-optha) as benchmarks. In order to refine the final results, a specialist ophthalmologist manually segmented a random selection of DIARETDB0/1 fundus images that presented exudates. An innovative pipeline of morphological procedures and fuzzy C-means clustering was integrated in order to extract exudates with a pixel-wise approach. Our methodology was optimized, and verified and the parameters were fine-tuned in order to define both suitable values and to produce a more accurate segmentation. The method was used on 100 tested images, resulting in averages of sensitivity, specificity, and accuracy equal to 83.3%, 99.2%, and 99.1%, respectively
Global Archaelogical Mosaicing for Underwater Scenes
No full textThis contribution regards the mosaicing of seabed landscapes, in order to represent higher resolution photos of whole sites with wrecks in a fast and safe fashion. A stereo vision system has been arranged by adding two cameras to the payload aboard a Remotely Operated Vehicle. A number of problems arise due to poor luminosity, cloudy water, water distortion and presence of artifacts. A robust algorithm has been defined to reduce the radial distortion of the camera lenses and to enhance the results
Transfer Learning Approach for High-Imbalance and Multi-class Classification of Fluorescence Images
No full textRecent advances in deep learning have often surpassed human performance in image classification. Among the most renowned cases, just think of the ImageNet Large Scale Visual Recognition Challenge competition. However, challenges persist in complex fields such as medical imaging. An example is the Human Protein Atlas which maps all human proteins in more than 171,000 images that makes a computation challenge due to high class imbalance. To address these challenges from a green perspective, we propose a transfer learning approach using Convolutional Neural Networks (CNNs) pre-trained on the ImageNet dataset. We use CNN layers as feature extractors, feeding the extracted features into a Support Vector Machine with a linear kernel. Our method combines both image-level and cell-level perspectives. Furthermore, at the cell level, we segment nuclei and extract the surrounding nuclear membrane area. The combination of the two perspectives shows promising classification performance with limited computational effort
A Novel Approach for Leveraging Agent-Based Experts on Large Language Models to Enable Data Sharing Among Heterogeneous IoT Devices in Agriculture
Full text linkThe rapid adoption of Internet of Things (IoT) devices in agriculture has led to the generation of diverse data types, creating challenges in data sharing and integration across heterogeneous platforms. This paper presents a novel approach to facilitate data sharing among heterogeneous IoT devices in agriculture using agent-based experts built on large language models (LLMs). Background: Traditional methods of data sharing in agriculture face limitations due to the lack of standardization and interoperability among IoT devices. Previous approaches, such as model fine-tuning and prompt engineering, have shown promise but struggle with open-ended agricultural queries and context comprehension. The proposed Agent-based Data Sharing (ADS) framework combines semantic web technologies with agent-based design and LLMs to enable seamless information exchange, decentralized data sharing, and knowledge transfer through intelligent expert agents. This approach leverages the strengths of LLMs in understanding text and their extensive training data while addressing the challenges of data interoperability and context-aware decision-making in agriculture. Using synthetic agricultural data, we evaluated the framework's performance in disease diagnosis and precision farming recommendations. The results demonstrate significant improvements in data integration, interoperability, and decision-making efficiency. With extensive data sharing, mean performance scores increased by 16% for disease diagnosis and 25% for precision farming compared to baseline scenarios. The framework's ability to manage diverse devices and handle natural language queries through agent-based experts highlights its potential for real-world agricultural applications. This approach could support the advancement of smart farming through IoT applications and pave the way for improved efficiency in sustainable agriculture. However, challenges such as data privacy, standardization, and incentive structures need to be addressed in future research
- …
