Computational pathology enables the automatic tissue analysis of Whole Slide Images (WSIs), offering unmatched possibilities to capture quantitative tumor microenvironment (TME) characteristics that are essential for patients' prognosis and therapy response. The existing clinical and digital biomarkers do not encompass the morphometric features and spatial interactions between vascular networks and immunological compartment in the TME. To address this challenge, this work presents a high- throughput quantitative framework for automatic segmentation and assessment of aberrant phenotypes of blood vessels and immune cell clusters in hematoxylin & eosin-stained WSIs, to construct the Vascular-Immune Pathomic (VIPath) biomarker. For our study, we utilized three public datasets of Colon Adenocarcinoma (COAD) and Stomach Adenocarcinoma (STAD) from The Cancer Genome Atlas (TCGA) and Clinical Proteomic Tumor Analysis Consortium (CPTAC) projects: TCGA-COAD, TCGA-STAD, and CPTAC-COAD. Additionally, we collected two in-house gastric cancer cohorts of 80 and 51 patients : DBGC-M0 and DBGC-M1. The VIPath biomarker was incorporated in a Cox proportional hazards model trained on the TCGA-COAD. Then, it was validated for predicting Overall Survival (OS) in TCGA-STAD, DBGC-M0, and DBGC-M1, Disease Free Survival in CPTAC-COAD and second-line therapy Progression-free Survival (PFS-2) in DBGC-M1. VIPath encompasses features from both vascular and immunological compartments interacting in the TME. Results proved that VIPath was capable to significantly stratify risk groups for OS TCGA-STAD (p=0.018), OS DBGC-M0 (p=0.029), OS DBGC-M1 (p=0.014), and PFS-2 DBGC-M1 (p$< $0.005). Furthermore, when inserted in a Cox model, it led to an improvement of C-index and R2over all other considered prognostic factors, i.e., p-TNM, ECOG, MSI, HER2.

Work-related musculoskeletal disorders (WMDs) affect a high percentage of operators performing repeated weight lifting and load carrying in industrial scenarios. Since upper limb muscles are affected in the process, the assistance provided by upper body exoskeletons is increasingly needed to prevent WMDs and their consequent cost to the health system. This paper presents the evaluation of Flexos, a portable, bilateral, shoulder exoskeleton prototype designed to assist logistic and industrial operators in performing occupational tasks. An in-lab assessment was conducted on twelve healthy subjects - 9 males, 3 females - to evaluate Flexos capability in assisting the user during the execution of isometric, dynamic, and carrying-load tasks. Different metrics were extracted from time-series signals to assess the effort related to five targeted muscles surrounding the shoulder complex. Despite the limited experimental size and the prototypal level of the device, Flexos managed to cover almost all the shoulders range of motion - 89.2% flexion/extension, and 88.4% internal/external rotation - and to globally decrease muscular activity in occupational activities, particularly when isometric contractions are required for a prolonged time, with average reductions of -27.2% for the static task, -18.6% for the dynamic task and -23.4% for the carrying-load task.

The identification of EEG biomarkers to discriminate Subjective Cognitive Decline (SCD) from Mild Cognitive Impairment (MCI) conditions is a complex task which requires great clinical effort and expertise. We exploit the self-attention component of the Transformer architecture to obtain physiological explanations of the model's decisions in the discrimination of 56 SCD and 45 MCI patients using resting-state EEG. Specifically, an interpretability workflow leveraging attention scores and time-frequency analysis of EEG epochs through Continuous Wavelet Transform is proposed. In the classification framework, models are trained and validated with 5-fold cross-validation and evaluated on a test set obtained by selecting 20% of the total subjects. Ablation studies and hyperparameter tuning tests are conducted to identify the optimal model configuration. Results show that the best performing model, which achieves acceptable results both on epochs' and patients' classification, is capable of finding specific EEG patterns that highlight changes in the brain activity between the two conditions. We demonstrate the potential of attention weights as tools to guide experts in understanding which disease-relevant EEG features could be discriminative of SCD and MCI.

In Italian universities, bioinformatics courses are increasingly being incorporated into different study paths. However, the content of bioinformatics courses is usually selected by the professor teaching the course, in the absence of national guidelines that identify the minimum indispensable knowledge in bioinformatics that undergraduate students from different scientific fields should achieve. The Training&Teaching group of the Bioinformatics Italian Society (BITS) proposed to university professors a survey aimed at portraying the current situation of bioinformatics courses within undergraduate curricula in Italy (i.e., bioinformatics courses activated within both bachelor's and master's degrees). Furthermore, the Training&Teaching group took a cue from the survey outcomes to develop recommendations for the design and the inclusion of bioinformatics courses in academic curricula. Here, we present the outcomes of the survey, as well as the BITS recommendations, with the hope that they may support BITS members in identifying learning outcomes and selecting content for their bioinformatics courses. As we share our effort with the broader international community involved in teaching bioinformatics at academic level, we seek feedback and thoughts on our proposal and hope to start a fruitful debate on the topic, including how to better fulfill the real bioinformatics knowledge needs of the research and the labor market at both the national and international level.

The papers in this special section were presented at the Fourteenth International Conference on Intelligent Computing (ICIC) held in Wuhan, China, on August 15-18, 2018.

The growing interest in wearable robots for assistance and rehabilitation purposes opens the challenge for developing intuitive and natural control strategies. Among several human-machine interaction approaches, myoelectric control consists in decoding the motor intention from muscular activity (or EMG signals) with the aim at moving the assistive robotic device accordingly, thus establishing an intimate human-machine connection. In this scenario, bio-inspired approaches, e.g. synergy-based controllers, are reveling to be the most robust.In this work, the authors presented an undercomplete autoencoder (AE) to extract muscles synergies for motion intention detection. The proposed AE topology has been validate with EMG signals acquired from the main upper limb muscles during planar isometric reaching tasks performed in a virtual environment while wearing an exoskeleton. The presented AE have shown promising results in muscle synergy extraction comparing its performance with the Non-Negative Matrix Factorization algorithm, i.e. the most used approach in literature. The synergy activations extracted with the AE have been then used for estimating the moment applied at the shoulder and elbow joints. Comparing such estimation with the results of other synergy-based techniques already proposed in literature, it emerged that the proposed method achieves comparable performance.

The papers presented in this special section were presented at the Thirteenth International Conference on Intelligent Computing (ICIC) that was held in Liverpool, UK, on August 7-10, 2017. ICIC was formed to provide an annual forum dedicated to the emerging and challenging topics in artificial intelligence, machine learning, bioinformatics, and computational biology, etc. It aims to bring together researchers and practitioners from both academia and industry to share ideas, problems, and solutions related to the multifaceted aspects of intelligent computing.

Osteoporosis is characterized by low Bone Mineral Density (BMD). This illness has a highcost impact in all developed countries. In this work, we propose a supervised approach to classify the BMD status of post-menopausal women based on an experimental non-invasive analysis of static and dynamic baropodometry. A questionnaire on nutritional habits and lifestyle and previous fractures was drawn up. Sixty women in amenorrhea > 12 months and age > 45 years were included and divided in 3 groups (Normal, Osteopenic and Osteoporotic) according to T-score values. Those with neurological or musculoskeletal disorders, history of vestibulopathies, uncorrected visual deficit or drug use were excluded. Static and Dynamic Baropodometry was performed to all the enrolled women and a preliminary processing was carried on to select the most relevant features via Principal Component Analysis (PCA). Subsequently, two supervised classifiers based on the selected 27 features were designed and tested and their results were discussed as a promising tool for screening subjects suffering from both bone and muscle functional decline.

Patients suffering from Parkinson's Disease (PD) are characterized by an abnormal handwriting activity since they have difficulties in motor coordination and a decline in cognition. In this paper, we propose a model-free technique for differentiating PD patients from healthy subjects by using a handwriting analysis tool based on biometric signals (e.g., surface ElectroMyoGraphy, pen pressure, etc.) and an Artificial Intelligence-based classifier. Experimental tests have been carried on with both healthy and PD subjects to identify the most representative features, the best writing patterns and the best AI-based classification approach between Artificial Neural Network (ANN) and Support Vector Machine (SVM) in terms of accuracy and repeatability. Finally, the obtained results are reported and discussed to infer some important properties on writing patterns, classification approaches and the role of muscular activities on the handwriting analysis applied to neurodegenerative disease research.

The eight papers included in this special section were presented at the 12th International Conference on Intelligent Computing (ICIC) held at Lanzhou, China, during August 2-5, 2016. ICIC was formed to provide an annual forum dedicated to the emerging and challenging topics in artificial intelligence, machine learning, bioinformatics, and computational biology, etc. It aims to bring together researchers and practitioners from both academia and industry to share ideas, problems, and solutions related to the multifaceted aspects of intelligent computing.

Introduction and objective: Computer Aided Decision (CAD) systems based on Medical Imaging could support radiologists in grading Hepatocellular carcinoma (HCC) by means of Computed Tomography (CT) images, avoiding medical invasive procedures such as biopsies. The identification and characterization of Regions of Interest (ROIs) containing lesions is an important phase allowing an easier classification in two classes of HCCs. Two steps are needed for the detection of lesioned ROIs: a liver isolation in each CT slice and a lesion segmentation. Materials and methods: In our previous study, materials consisted in abdominal CT hepatic lesions of only three patients subjected to liver transplant, partial hepatectomy, or US-guided needle biopsy. In this paper, thanks to a more extensively phase of data collection, available materials impressively grew to 18 patients belonging to 2-balanced classes. Several approaches were implemented to segment the region of liver and, then, to detect the ROI of the lesions. At the end of these preprocessing phases, we extracted the same morphological features of the previous work and designed an evolutionary algorithm to optimize neural network classifiers based on different subsets of features. Results and conclusion: Tests conducted on the new ANN topologies showed a higher generalization of the average performance indices regardless of the applied training, validation and test sets, confirming both the validity and the robustness of the approach of previous study even though the limited number of patients.

Objectives: Nowadays breath test, using the Ghoss method for the calculation of gastric emptying of solids, is characterized by a very high number of expirations in a very long time (about 4 hours). In this work, a simplified model aiming to the reduction of the number of expirations during this time was designed, preserving high levels of accuracy, sensitivity and specificity in the classification between `normal' or `delayed' gastric emptying. Materials and Methods: Materials consist of 238 breath test exams from 66 different patients; for each exam, the relevance of each expiration was evaluated. Several models were designed and tested comparing their performance with a full model which took into account 17 expirations; among them, the model with the highest accuracy was selected: it consists of 7 expirations (baseline, 75, 135, 195, 210, 225 and 240 min). Results: Considering the previous model, the number of expirations was reduced by 62.5 %, still reaching high levels of accuracy, sensitivity and specificity (about 90 %) comparable with the full model which showed an accuracy of 98 %. Conclusion: The adoption of the proposed model led to a considerable reduction of required expirations, still having good performance in classifying `normal' and `delayed' gastric emptying. At the same time, since it requires fewer breaths, test was also simplified, allowing patients to do this exam at home. Furthermore, the reduction of breaths led to a considerable cost reduction of the entire examination.

The papers in this special section were presented at the 11th International Conference on Intelligent Computing (ICIC) held in Fuzhou, China, on August 20-23, 2015. This conference was formed to provide an annual forum dedicated to the emerging and challenging topics in artificial intelligence, machine learning, bioinformatics, etc. It aims to bring together researchers and practitioners from both academia and industry to share ideas, problems, and solutions related to the multifaceted aspects of intelligent computing.

This special section includes a selection of eight papers presented at the 10th International Conference on Intelligent Computing (ICIC) held in Taiyuan, China, on August 3–6, 2014.

This paper presents a myoelectric control of an arm exoskeleton designed for rehabilitation. A four-muscles-based NeuroMusculoSkeletal (NMS) model was implemented and optimized using genetic algorithms to adapt the model to different subjects. The NMS model is able to predict the shoulder and elbow torques which are used by the control algorithm to ensure a minimal force of interaction. The accuracy of the method is assessed through validation experiments conducted with two healthy subjects performing free movements along the pseudo-sagittal plane. The experiments show promising results for our approach showing its potential for being introduced in a rehabilitation protocol.

This paper presents the design and the implementation of a Computer Aided Diagnosis (CAD) system for the clinical analysis of Peripheral Blood Smears (PBS also called Blood Film). The proposed system is able to count and classify the five types of leucocytes located in the tail of a PBS for computing the leukocyte formula. Image processing and segmentation techniques were used to extract 33 leucocyte's features (morphological, chromatic and texture-based). Only 7 features, selected by using the Information Gain Ranking algorithm of Weka platform, were used to evaluate the classification performance of two different classifiers: Back Propagation Neural Network (BPNN) and Decision Tree (DT). From the comparison between the two proposed approaches we can argue that the BPNN performed better than the DT on the validation set. Finally, the Neural Network classifier was evaluated with a test set composed of 1274 leucocytes obtaining good results in terms of Precision (87.9%) and Sensitivity (97.4%).

In this study, we compared several classifiers for the supervised distinction between normal elderly and Alzheimer's disease individuals, based on resting state electroencephalographic markers, age, gender and education. Three main preliminary procedures served to perform features dimensionality reduction were used and discussed: a Support Vector Machines Recursive Features Elimination, a Principal Component Analysis and a novel method based on the correlation. In particular, five different classifiers were compared: two different configurations of SVM and three different optimal topologies of Error Back Propagation Multi Layer Perceptron Artificial Neural Networks (EBP MLP ANNs). Best result, in terms of classification (accuracy 86% and sensitivity 92%), was obtained by a Neural Network with 3 hidden layers that used as input: age, gender, education and 20 EEG features selected by the novel method based on the correlation.

A non-invasive Brain Computer Interface (BCI) based on a Convolutional Neural Network (CNN) is presented as a novel approach for navigation in Virtual Environment (VE). The developed navigation control interface relies on Steady State Visually Evoked Potentials (SSVEP), whose features are discriminated in real time in the electroencephalographic (EEG) data by means of the CNN. The proposed approach has been evaluated through navigation by walking in an immersive and plausible virtual environment (VE), thus enhancing the involvement of the participant and his perception of the VE. Results show that the BCI based on a CNN can be profitably applied for decoding SSVEP features in navigation scenarios, where a reduced number of commands needs to be reliably and rapidly selected. The participant was able to accomplish a waypoint walking task within the VE, by controlling navigation through of the only brain activity.

Introduction and objective: the purpose of this work is to design and implement an innovative tool to recognize 16 different human gestural actions and use them to predict 7 different emotional states. The solution proposed in this paper is based on RGB and depth information of 2D/3D images acquired from a commercial RGB-D sensor called Kinect. Materials: the dataset is a collection of several human actions made by different actors. Each action is performed by each actor for three times in each video. 20 actors perform 16 different actions, both seated and upright, totalling 40 videos per actor. Methods: human gestural actions are recognized by means feature extractions as angles and distances related to joints of human skeleton from RGB and depth images. Emotions are selected according to the state-of-the-art. Experimental results: despite truly similar actions, the overall-accuracy reached is approximately 80%. Conclusions and future works: the proposed work seems to be back-ground- and speed-independent, and it will be used in the future as part of a multimodal emotion recognition software based on facial expressions and speech analysis as well.

Falls are one of the major risks of injury for elderly living alone at home. Computer vision-based systems offer a new, low-cost and promising solution for fall detection. This paper presents a new fall-detection tool, based on a commercial RGB-D camera. The proposed system is capable of accurately detecting several types of falls, performing a real time algorithm in order to determine whether a fall has occurred. The proposed approach is based on evaluating the contraction and the expansion speed of the width, height and depth of the 3D human bounding box, as well as its position in the space. Our solution requires no pre-knowledge of the scene (i.e. the recognition of the floor in the virtual environment) with the only constraint about the knowledge of the RGB-D camera position in the room. Moreover, the proposed approach is able to avoid false positive as: sitting, lying down, retrieve something from the floor. Experimental results qualitatively and quantitatively show the quality of the proposed approach in terms of both robustness and background and speed independence.

BACKGROUND: Expressed sequences (e.g. ESTs) are a strong source of evidence to improve gene structures and predict reliable alternative splicing events. When a genome assembly is available, ESTs are suitable to generate gene-oriented clusters through the well-established EasyCluster software. Nowadays, EST-like sequences can be massively produced using Next Generation Sequencing (NGS) technologies. In order to handle genome-scale transcriptome data, we present here EasyCluster2, a reimplementation of EasyCluster able to speed up the creation of gene-oriented clusters and facilitate downstream analyses as the assembly of full-length transcripts and the detection of splicing isoforms. RESULTS: EasyCluster2 has been developed to facilitate the genome-based clustering of EST-like sequences generated through the NGS 454 technology. Reads mapped onto the reference genome can be uploaded using the standard GFF3 file format. Alignment parsing is initially performed to produce a first collection of pseudo-clusters by grouping reads according to the overlap of their genomic coordinates on the same strand. EasyCluster2 then refines read grouping by including in each cluster only reads sharing at least one splice site and optionally performs a Smith-Waterman alignment in the region surrounding splice sites in order to correct for potential alignment errors. In addition, EasyCluster2 can include unspliced reads, which generally account for >50% of 454 datasets, and collapses overlapping clusters. Finally, EasyCluster2 can assemble full-length transcripts using a Directed-Acyclic-Graph-based strategy, simplifying the identification of alternative splicing isoforms, thanks also to the implementation of the widespread AStalavista methodology. Accuracy and performances have been tested on real as well as simulated datasets. CONCLUSIONS: EasyCluster2 represents a unique tool to cluster and assemble transcriptome reads produced with 454 technology, as well as ESTs and full-length transcripts. The clustering procedure is enhanced with the employment of genome annotations and unspliced reads. Overall, EasyCluster2 is able to perform an effective detection of splicing isoforms, since it can refine exon-exon junctions and explore alternative splicing without known reference transcripts. Results in GFF3 format can be browsed in the UCSC Genome Browser. Therefore, EasyCluster2 is a powerful tool to generate reliable clusters for gene expression studies, facilitating the analysis also to researchers not skilled in bioinformatics.

BACKGROUND: Discovering the molecular targets of compounds or the cause of physiological conditions, among the multitude of known genes, is one of the major challenges of bioinformatics. One of the most common approaches to this problem is finding sets of differentially expressed, and more recently differentially co-expressed, genes. Other approaches require libraries of genetic mutants or require to perform a large number of assays. Another elegant approach is the filtering of mRNA expression profiles using reverse-engineered gene network models of the target cell. This approach has the advantage of not needing control samples, libraries or numerous assays. Nevertheless, the impementations of this strategy proposed so far are computationally demanding. Moreover the user has to arbitrarily choose a threshold on the number of potentially relevant genes from the algorithm output. RESULTS: Our solution, while performing comparably to state of the art algorithms in terms of discovered targets, is more efficient in terms of memory and time consumption. The proposed algorithm computes the likelihood associated to each gene and outputs to the user only the list of likely perturbed genes. CONCLUSIONS: The proposed algorithm is a valid alternative to existing algorithms and is particularly suited to contemporary gene expression microarrays, given the number of probe sets in each chip, also when executed on common desktop computers.

BACKGROUND: DNA microarray data are used to identify genes which could be considered prognostic markers. However, due to the limited sample size of each study, the signatures are unstable in terms of the composing genes and may be limited in terms of performances. It is therefore of great interest to integrate different studies, thus increasing sample size. RESULTS: In the past, several studies explored the issue of microarray data merging, but the arrival of new techniques and a focus on SVM based classification needed further investigation. We used distant metastasis prediction based on SVM attribute selection and classification to three breast cancer data sets. CONCLUSIONS: The results showed that breast cancer classification does not benefit from data merging, confirming the results found by other studies with different techniques.

BACKGROUND: Mechanistic models are becoming more and more popular in Systems Biology; identification and control of models underlying biochemical pathways of interest in oncology is a primary goal in this field. Unfortunately the scarce availability of data still limits our understanding of the intrinsic characteristics of complex pathologies like cancer: acquiring information for a system understanding of complex reaction networks is time consuming and expensive. Stimulus response experiments (SRE) have been used to gain a deeper insight into the details of biochemical mechanisms underlying cell life and functioning. Optimisation of the input time-profile, however, still remains a major area of research due to the complexity of the problem and its relevance for the task of information retrieval in systems biology-related experiments. RESULTS: We have addressed the problem of quantifying the information associated to an experiment using the Fisher Information Matrix and we have proposed an optimal experimental design strategy based on evolutionary algorithm to cope with the problem of information gathering in Systems Biology. On the basis of the theoretical results obtained in the field of control systems theory, we have studied the dynamical properties of the signals to be used in cell stimulation. The results of this study have been used to develop a microfluidic device for the automation of the process of cell stimulation for system identification. CONCLUSION: We have applied the proposed approach to the Epidermal Growth Factor Receptor pathway and we observed that it minimises the amount of parametric uncertainty associated to the identified model. A statistical framework based on Monte-Carlo estimations of the uncertainty ellipsoid confirmed the superiority of optimally designed experiments over canonical inputs. The proposed approach can be easily extended to multiobjective formulations that can also take advantage of identifiability analysis. Moreover, the availability of fully automated microfluidic platforms explicitly developed for the task of biochemical model identification will hopefully reduce the effects of the 'data rich--data poor' paradox in Systems Biology.

In this paper we present a comparative study among well established data mining algorithm (namely J48 and naive Bayes tree) and novel machine learning paradigms like ant miner and gene expression programming. The aim of this study was to discover significant rules discriminating ER+ and ER-cases of breast cancer. We compared both statistical accuracy and biological validity of the results using common statistical methods and gene ontology. Some worth noting characteristics of these systems have been observed and analysed even giving some possible interpretations of findings. With this study we tried to show how intelligent systems can be employed in the design of experimental pipeline in disease processes investigation and how deriving high-throughput results can be validated using new computational tools. Results returned by this approach seem to encourage new efforts in this field

Genomic DNA copy number aberrations are frequent in solid tumours although their underlying causes remain obscure. In this paper we show how Artificial Immune System (AIS) paradigm can be successfully employed in the elucidation of biological dynamics of cancerous processes using a novel fuzzy rule induction system for data mining (IFRAIS). Competitive results have been obtained using IFRAIS. A biological interpretation of the results, carried out using Gene Ontology, followed the statistical assessment and put in evidence interesting patterns that are currently under investigation.

This paper presents a novel approach to artificial neural network (ANN) topology optimisation that uses multi-objective genetic algorithm in order to find the best network configuration for the Wisconsin breast cancer database (WBCD) classification problem. The WBCD [Mangasarian, OL., et al., 1995][Mangasarian, OL., et al.][Wolberg, WH., et al., 1995] is a publicly available database composed by 699 cases, each of which is defined by 11 parameters. The former first 10 values of each record account for geometrical features of cells extracted with FNA biopsy. The last parameter represents the nature of the tumour; two classes of tumour are considered in this database: benignant and malignant tumours. An Intelligent System, IDEST, was designed and implemented. At the core of this system there's an Artificial Neural Network that is able to classify cases. The design of such an ANN is a non trivial task and choices incoherent with the problem could lead to instability of the network. For these reasons a fixed topology genetic algorithm (GA) approach was used to find an optimal topology for the given problem. In a second step a multi-objective GA (MOGA) was developed and employed in order to refine the search in the "topology space". Results shown by the IDEST demonstrate the great potentialities of similar approaches.

A unified evolutionary approach to coplanar radiotherapy inverse planning is proposed. It consists of a genetic algorithm-based framework that solves with little modification treatment planning for three different kinds of radiation therapy: conformal, so-called aperture-based and intensity modulated. Thanks to evolutionary optimisation techniques we have been able to search for full beam configurations, that is, beam intensity, beam shape and especially beam orientation. Unlike some previous works found in literature, our proposed solution automatically determines exact beam angles not relaying solely on a geometrical basis but involving beam intensity profiles, thus considering the effective delivered dose. Our dose distribution model has been validated through comparison with commercial system: fixed the same beam configuration, both calculated beam shapes and the DVH have been compared. Then we have tested the optimisation algorithm with real clinical cases: these involved both simple (convex target, far OARs) and complex (concave target, close OARs) ones. As stated by physician and by simulation with the same commercial system, our tools found good solutions in both cases using corresponding correct therapy.

This paper presents a new hybrid range image segmentation approach. Two separate techniques are applied consecutively. First, an edge based segmentation technique extracts the edge points-creases and jumps-contained in the given range image. Then, by using only the edge point position information, the boundaries are computed. Secondly, the points clustered into each region are approximated by single surfaces through a genetic algorithm (GA). The GA takes advantage of previous edge representation finding the surface parameters that best fit each region. It works in a local way, according to the boundary information, reducing considerably the required CPU time. Experimental results with different range images are presented; moreover a comparison using either the edge detection stage or not is given.

In this paper two different approaches to provide information from events by high energy physics experiments are shown. Usually the representations produced in such experiments are spot-composed and the classical algorithms to be needed for data analysis are time consuming. For this reason the possibility to speed up pattern recognition tasks by soft computing approach with parallel algorithms has been investigated. The first scheme shown in the following is a two-layer neural network with forward connections, the second one consists of an evolutionary algorithm with elitistic strategy and mutation and cross-over adaptive probability. Test results of these approaches have been carried out analysing a set of images produced by an optical ring imaging Cherenkov (RICH) detector at CERN.