Fabio Stroppa

Fall detection in indoor environment with kinect sensor

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.

A Robot-Assisted Neuro-Rehabilitation System for Post-Stroke Patients’ Motor Skill Evaluation with ALEx Exoskeleton

Robotic Neuro Rehabilitation has been proved to be effective for motor recovery and less demanding for therapists. During a therapy, the aim of the task is to maximize the patient’s effort with respect to his/her clinical status and motor abilities, improving at the same time his/her impaired movements. In this paper, we propose a framework for performance evaluation of post-stroke subjects able to provide assistance-as-needed based on their motor skills during a therapy session with an upper-limb robotic exoskeleton.

RELIVE: A Markerless Assistant for CPR Training

Cardiopulmonary resuscitation (CPR) is a first-aid key survival technique used to stimulate breathing and keep blood flowing to the heart. Its effective administration can significantly increase the chances of survival in victims of cardiac arrest. In this paper, we propose a markerless system for quality CPR training based on RGB-D (RGB + Depth) sensors, called RELIVE. Then, we report the results of a series of experimental tests conducted to evaluate RELIVE tracking performance. The proposed system is able to accurately track the 3-D position of the hands performing CPR by means of RGB-D sensors to estimate the chest compression rate and depth, providing a real-time visual/audio feedback about the rescuers performance. Finally, the system usability has been assessed by both healthcare professionals and lay people.

RELIVE Tracking for quality cardiopulmonary resuscitation training: An experimental comparison with a standard CPR training mannequin

The most widely used and recognized approach to train high uality basic life support (BLS) manoeuvres and automated exteral defibrillation (AED) remains the classic instructor-led training ourse. A recent review, however, introduced new evidence in upport to alternative methods of training, including the use of selfirected learning and CPR feedback/prompt devices.1 For the Viva! ampaign 2014, the Italian Resuscitation Council developed a new nd more ambitious project called “Relive” game.2 In this article, we ropose a new markerless solution3,4 called RELIVE Tracking still eveloped within the Mini-VREM project,5 which is able to accuately estimate chest compressions depth and rate during chest ompression. In addition, RELIVE Tracking has been tested with two ifferent RGB-D (Red Green Blue-Depth) sensors based on differnt technologies featuring different prizes (Kinect® v1, Microsoft, edmond, WA, USA and Creative Senz3D®, Creative Technology, ingapore, Republic of Singapore) and has been provided of a gameike realistic interface used for conveying a 3D visual feedback to he user/rescuer. The RELIVE Tracking software (the engineer’s heart of Relive ame) was specifically developed, to guide the training and to mprove the quality of chest compression (CC) by tracking the ands of the user, without the need of any marker. RELIVE Trackng features a game-like Graphical User Interface (GUI) (Fig. 1) that llows non-experts to intuitively access all the application. RELIVE racking was tested with both RGB-D sensors, on a sample of ten ealthy subjects to evaluate the effect of the proposed software on C performance. This study was carried out at the PERCRO Laboraory in Pisa in August 2014. Ten male participants were recruited rom students and researchers (non-CPR experts) at the PERCRO aboratory. For each participant, the experiment consisted of a roup of three trials of CC each lasting 30 s and characterized by different depth CC (4–6 cm). Each group of trials was repeated for ach of the two RGB-D sensors, 60 trials in total. For each of the 60 rials, the data were simultaneously acquired with RELIVE Trackng and with a traditional training mannequin (Resusci Anne – RA, aerdal Medical, Stavanger, Norway) that was used for a quantitaive evaluation of the accuracy of CC depth measured with RELIVE racking. The best RELIVE Tracking performance was obtained with ELIVE Tracking using Microsoft Kinect® v1, with an average square uadratic error equal to 4.3 ± 0.3 mm, whereas the worst was with he Creative Senz3d® with a mean square quadratic error equal to .5 ± 0.3 mm. Considering RELIVE Tracking as a low-cost training

A Robust Real-Time 3D Tracking Approach for Assisted Object Grasping

Robotic exoskeletons are being increasingly and successfully used in neuro-rehabilitation therapy scenarios. Indeed, they allow patients to perform movements requiring more complex inter-joint coordination and gravity counterbalancing, including assisted object grasping. We propose a robust RGB-D camera-based approach for automated tracking of both still and moving objects that can be used for assisting the reaching/grasping tasks in the aforementioned scenarios. The proposed approach allows to work with non pre-processed objects, giving the possibility to propose a flexible therapy. Moreover, our system is specialized to estimate the pose of cylinder-like shaped objects to allow cylinder grasps with the help of a robotic hand orthosis.

An Improved Adaptive Robotic Assistance Methodology for Upper-Limb Rehabilitation

In this work, we propose an improved version of our algorithm for real-time robotic assistance tuning in robot-based therapy with any kind of active device for upper-limb rehabilitation. In particular, the work describes in detail how to extract accurate performance indices from the subject’s execution, and how to correlate them with the amount of assistance to be correspondingly provided over time. The algorithm also aims at enhancing subject’s efforts for a more effective recovery, tailoring the therapy to the patient without prior knowledge of his/her clinical status. Finally, an assessment phase illustrates the effectiveness of the procedure, showing how the system tunes the assistance required by the subjects to perform specific tasks.

Real-Time 3D Tracker in Robot-Based Neurorehabilitation

Abstract The chapter describes a computer vision-based robot-assisted system used in neurorehabilitation of post-stroke patients that allows the subjects to reach for and grasp objects in a defined workspace. The proposed computer vision technique is used to model objects that have not been preprocessed in a real setting, track them in real time, and provide their actual pose to the robotic device in order to accomplish grasping tasks. The robotic device is composed of three integrated modules: (i) a 4-DOF arm exoskeleton that supports the patients impaired arm when reaching for the objects; (ii) a 3-DOF actuated wrist exoskeleton for optimizing the hand pose in the grasping task; and (iii) a 2-DOF (flexion/extension) underactuated hand exoskeleton designed to be automatically adjusted for different grasping tasks based on contact forces. The conducted tests have demonstrated the robustness of the proposed approach, and its performance in the neurorehabilitation scenario through reaching and grasping task experiments.

Convex polygon fitting in robot-based neurorehabilitation

Abstract Fitting a polygon to a set of points is a task that finds application in many scientific fields. In particular, in robot-based neurorehabilitation, it would be interesting to retrieve the shape that best fits with the path followed by a patient, and evaluate the performance based on the accuracy of the drawing. However, when dealing with a dataset sampled by a drawn trajectory, the methods proposed by the literature may not be exhaustive. In this work we propose a method to define the distance between a set of points and a polygon, which is used as cost function of a Genetic Algorithm to solve the polygon fitting problem. This method involves a novel space separation metrics to retrieve the correct polygon edge to be compared with each point of the set, featuring linear time complexity. We compared the proposed approach with the metrics known in the literature, finding that our method performs significantly better in retrieving the original polygon. Finally, we present a robot-based rehabilitation application in which the proposed method is used to evaluate the performance of a group of subjects. The achievements of twenty healthy subjects were compared with three stroke patients. Results emphasize significant differences between the two categories of subjects, proving that the proposed algorithm can quantitatively determine the degree of impairment of a stroke survivor and be used in the future as reference for monitoring and enhancing the efficiency of robot-based therapies.

EasyCluster2: an improved tool for clustering and assembling long transcriptome reads

BackgroundExpressed 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.ResultsEasyCluster2 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.ConclusionsEasyCluster2 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.