Pasar al contenido principal

Resultados relacionados

Resultados totales (Incluyendo duplicados): 6
Encontrada(s) 1 página(s)

RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Impact of Parkinson&apos, s Disease on Functional Mobility at Different Stages

  • Mollà-Casanova, Sara
  • PEDRERO, J.F.
  • Inglés, Marta
  • Lopez Pascual, Juan
  • Muñoz-Gómez, Elena
  • Aguilar-Rodríguez, Marta
  • Sempere-Rubio, Nuria
  • Serra-Añó, Pilar
[EN] Introduction: Specific functional assessments to determine the progression of Parkinson¿s Disease (PD) are important to slow down such progression and better plan rehabilitation. This study aimed to explore possible differences in the performance of different functional tasks included in a mobility test using sensors embedded in an Android device, in people at different PD stages. Materials and Methods: Eighty-seven participants with PD agreed to participate in this cross-sectional study. They were assessed once using an inertial sensor and variables related to functional status were recorded (i.e., MLDisp, APDisp, DispA, Vrange, MLRange, PTurnSit, PStand, TTime, and RTime). Results: There was significant impairment of the vertical range during gait between stages I and II. Further, when stages II and III were compared, the sit-to-stand power was significantly impaired, and the total time required to complete the test increased significantly (p < 0.05). Even more significant differences were obtained when stages I and III were compared, in particular, dysfunction in postural control, vertical range, sit to stand power and total time. Finally, there were no significant differences between stages in the medial-lateral displacements and reaction time (p > 0.05). Conclusion: Functional mobility becomes more significantly impaired in the PD population as the PD stages progress. This implies impaired postural control, decreased ability to sit down or stand up from a chair, increased metabolic cost during walking, and overall slowing-down of motor function., This work was supported by the Universitat de Valencia (INV19-01-13-07, 2019) and developed within the framework of the IBERUS project. Technological Network of Biomedical Engineering applied to degenerative pathologies of the neuromusculoskeletal system in clinical and outpatient settings (CER-20211003), CERVERA Network financed by the Ministry of Science and Innovation through the Centre for Industrial Technological Development (CDTI), charged to the General State Budgets 2021, and the Recovery, Transformation, and Resilience Plan.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Protocol for Forearm Pronosupination Strength Measuring in Different Postures: Reliability and Influence of Relevant Factors

  • Herrera-Ligero, Cristina
  • Sánchez Zuriaga, Daniel
  • Martínez-Iranzo, Ursula
  • Pitarch Corresa, Salvador
  • De Rosario Martínez, Helios
[EN] Importance: Despite the importance of pronosupination strength, which is key in daily activities, current evaluation protocols are disparate, and results differ regarding the influence of factors such as gender, age, limb dominance, or posture. Objective: To design a reliable device to measure pronation and supination torques in different forearm positions, considering the influence of gender, age, and limb dominance. Design: Reliability study with analysis of intrarater and interrater repeatability and examination of the influence of different factors on the strength of pronation and supination. Setting: Biomechanical analysis laboratory of the Instituto de Biomecánica de Valencia. Participants: Convenience sample of 39 healthy male and female participants, ages 18 to 65 yr. Outcomes and measures: Pronation strength and supination strength were studied in five forearm positions: 30° and 60° of supination, neutral rotation, and 30° and 60° of pronation. The influence of gender, age, and limb dominance was studied using repeated-measures analysis of variance. Intrarater and interrater reliabilities were studied in 17 participants, and the intraclass correlation coefficients (ICCs) were calculated from three measurement sessions. Results: Except for 60° of pronation, all ICCs ranged from .72 to .97. Gender, limb dominance, and posture significantly affected pronosupination strength (p < .05), with the highest supination strength at 60° of pronation and the highest pronation strength at 60° of supination. Conclusions and relevance: In using this reliable pronosupination torque assessment device and standardized protocol in the clinical setting, clinicians must consider the influence of gender, limb dominance, and forearm posture. What This Article Adds: The results of this study will allow occupational therapists to establish differences between patients' functionality in comparison with healthy individuals and to design recovery treatments and facilitate performance of manual tasks, considering which are the most advantageous positions to exert force. Furthermore, the developed device can be used to monitor the evolution of these torques in an objective and reliable manner., This article was developed within the framework of the IBERUS project, a technological network of biomedical engineering applied to degenerative pathologies of the neuromusculoskeletal system in clinical and outpatient settings (CER-20211003) , and the CERVERA Network, financed by the Ministry of Science and Innovation through the Center for Industrial Technological Development, charged to the General State Budgets 2021 and the Recovery, Transformation and Resilience Plan.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Classification of Parkinson&apos, s disease stages with a two-stage deep neural network

  • PEDRERO, J.F.
  • Serra Añó, Pilar
  • Mollà-Casanova, Sara
  • Lopez Pascual, Juan
  • Belda-Lois, Juan-Manuel|||0000-0002-7648-799X
Research Activity (IMAMCA/2022/7) supported by Instituto Valenciano de Competitividad Empresarial (IVACE) and Valencian Regional Government (GVA), and supported by the Universitat de València [INV19-01-13-07, 2019] funding and developed within the framework of the IBERUS project. Technological Network of Biomedical Engineering applied to degenerative pathologies of the neuromusculoskeletal system in clinical and outpatient settings (CER-20211003), CERVERA Network financed by the Ministry of Science and Innovation through the Center for Industrial Technological Development (CDTI), charged to the General State Budgets 2021, and the Recovery, Transformation, and Resilience Plan.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Tesis doctoral (DoctoralThesis)

Enhancing RGB-Based Markerless Motion Capture with 3D Temporal Scans. Study and Validation of a New Method

  • Ruescas Nicolau, Ana Virginia
[ES] Esta tesis explora la integración de datos del cuerpo completo en movimiento obtenidos con el escáner Move4D en sistemas de seguimiento basados en imágenes RGB para mejorar la estimación de pose 3D sin marcadores. Aunque los métodos tradicionales de captura de movimiento con marcadores son precisos, requieren mucho tiempo y solo funcionan en entornos controlados. Las técnicas sin marcadores, basadas en imágenes RGB y Deep Learning, ofrecen mayor flexibilidad, pero aún carecen de la precisión necesaria para aplicaciones en biomecánica e investigación clínica. El objetivo principal es evaluar si los datos del escáner Move4D pueden mejorar la estimación de pose basada en imágenes RGB, permitiendo así un sistema de captura de movimiento sin marcadores más preciso y accesible. La investigación incluyó tres estudios experimentales y la creación de una base de datos de a partir de los movimientos escaneados: Estudio 1: Se evaluó la precisión espacial del escáner Move4D en el análisis de la marcha, comparando sus resultados con los obtenidos mediante estereofotogrametría con marcadores. Los errores de posicionamiento estático y dinámico, así como las discrepancias en ángulos articulares (especialmente en cadera y rodilla), se mantuvieron dentro de rangos aceptables, confirmando la fiabilidad del escáner. Estudio 2: Se analizó la consistencia en la identificación de puntos anatómicos usando la topología de malla homóloga del escáner. Utilizando el conjunto de datos CAESAR, se asignaron 53 puntos anatómicos a vértices fijos en 3.165 escaneos corporales 3D. Los errores de posicionamiento fueron comparables a los obtenidos mediante técnicas de palpación manual, validando la topología de malla para la detección automática. No obstante, imponer la simetría en la identificación de puntos anatómicos aumentó los errores, evidenciando una controversia entre fidelidad anatómica y las restricciones de diseño. Estudio 3: Se investigó si los datos del Move4D podían entrenar redes neuronales para mejorar la estimación de pose sin marcadores basada en imagen RGB. Se creó una base de datos con 51.051 poses de 71 participantes, etiquetada con 53 puntos anatómicos y 33 puntos clave de Mediapipe. Se entrenaron tres modelos: un perceptrón multicapa, una red LSTM y un Transformer. Todos lograron precisiones dentro del rango de variabilidad intra-examinados para poses estáticas, destacando el modelo Transformer en la estimación de ángulos articulares y puntos anatómicos durante movimientos de traslación del cuerpo. El análisis de errores mostró que la precisión variaba según la región corporal: mayores errores en la pelvis y menores en los tobillos. Actividades como correr generaban más errores, mientras que las sentadillas y los saltos permitían estimaciones más precisas. Un análisis con modelos lineales mixtos indicó que el tipo de movimiento, el eje articular y la arquitectura de red influían más en la variabilidad del error que la forma corporal, lo que sugiere una buena capacidad de generalización de los modelos. En conclusión, los datos del escáner Move4D mejoran significativamente la precisión de la estimación de pose 3D sin marcadores basada en imágenes RGB. Gracias a su precisión y su malla topológicamente consistente, es posible generar bases de datos de alta calidad para entrenar modelos. Este trabajo ofrece un enfoque sólido y accesible para la estimación de pose, representando un avance clave para el análisis del movimiento fuera del entorno de laboratorio., [CA] Aquesta tesi explora la integració de dades del cos complet en moviment obtinguts amb l'escàner Move4D en sistemes de seguiment basats en imatges RGB per a millorar l'estimació de posa 3D sense marcadors. Encara que els mètodes tradicionals de captura de moviment amb marcadors són precisos, requereixen molt temps i només funcionen en entorns controlats. Les tècniques sense marcadors, basades en imatges RGB i Deep Learning, ofereixen major flexibilitat, però encara manquen de la precisió necessària per a aplicacions en biomecànica i recerca clínica. L'objectiu principal és avaluar si les dades de l'escàner Move4D poden millorar l'estimació de posa basada en imatges RGB, permetent així un sistema de captura de moviment sense marcadors més precís i accessible. La recerca va incloure tres estudis experimentals i la creació d'una base de dades d'a partir dels moviments escanejats: Estudi 1: Es va avaluar la precisió espacial de l'escàner Move4D en l'anàlisi de la marxa, comparant els seus resultats amb els obtinguts mitjançant estereofotogrametría amb marcadors. Els errors de posicionament estàtic i dinàmic, així com les discrepàncies en angles articulars (especialment en maluc i genoll), es van mantenir dins de rangs acceptables, confirmant la fiabilitat de l'escàner. Estudi 2: Es va analitzar la consistència en la identificació de punts anatòmics usant la topologia de malla homòloga de l'escàner. Utilitzant el conjunt de dades CAESAR, es van assignar 53 punts anatòmics a vèrtexs fixos en 3.165 escanejos corporals 3D. Els errors de posicionament van ser comparables als obtinguts mitjançant tècniques de palpació manual, validant la topologia de malla per a la detecció automàtica. No obstant això, imposar la simetria en la identificació de punts anatòmics va augmentar els errors, evidenciant una controvèrsia entre fidelitat anatòmica i les restriccions de disseny. Estudi 3: Es va investigar si les dades del Move4D podien entrenar xarxes neuronals per a millorar l'estimació de posa sense marcadors basada en imatge RGB. Es va crear una base de dades amb 51.051 poses de 71 participants, etiquetada amb 53 punts anatòmics i 33 punts clau de Mediapipe. Es van entrenar tres models: un perceptrón multicapa, una xarxa LSTM i un Transformer. Tots van aconseguir precisions dins del rang de variabilitat intra-examinadors per a poses estàtiques, destacant el model Transformer en l'estimació d'angles articulars i punts anatòmics durant moviments de translació del cos. L'anàlisi d'errors va mostrar que la precisió variava segons la regió corporal: majors errors en la pelvis i menors en els turmells. Activitats com córrer generaven més errors, mentre que els esquats i els salts permetien estimacions més precises. Una anàlisi amb models lineals mixtos va indicar que el tipus de moviment, l'eix articular i l'arquitectura de xarxa influïen més en la variabilitat de l'error que la forma corporal, la qual cosa suggereix una bona capacitat de generalització dels models. En conclusió, les dades de l'escàner Move4D milloren significativament la precisió de l'estimació de posa 3D sense marcadors basada en imatges RGB. Gràcies a la seva precisió i la seva malla topològicament consistent, és possible generar bases de dades d'alta qualitat per a entrenar models. Aquest treball ofereix un enfocament sòlid i accessible per a l'estimació de posa, representant un avanç clau per a l'anàlisi del moviment fora de l'entorn de laboratori., [EN] This thesis investigates the integration of real human motion data from the Move4D scanner with RGB-based motion tracking systems to improve the accuracy of markerless 3D pose estimation. Traditional motion capture methods, though precise, rely on marker-based systems that are time-consuming and limited to controlled environments. While deep learning-based, markerless techniques using RGB images offer more flexibility, they still lack the precision required for applications in biomechanics and clinical research. The core aim of this work is to determine whether Move4D scanner data can improve RGB-based pose estimation, ultimately enabling a more accessible and accurate markerless motion capture system. To this end, the research comprised three experimental studies and the creation of a large, labelled motion dataset. Each study contributed to validating different aspects of the proposed method: Study 1 evaluated the spatial accuracy of the Move4D scanner in dynamic gait analysis. It compared motion data from the scanner with results from traditional marker-based stereophotogrammetry. Findings showed that both static and dynamic positioning errors, as well as joint angle discrepancies (especially at the hip and knee), were within acceptable ranges. This affirmed the Move4D scanner's reliability for capturing motion data. Study 2 focused on the consistency and precision of anatomical landmark identification using the scanner's homologous mesh topology. Using the CAESAR dataset, 53 anatomical landmarks were assigned to fixed vertices across 3,165 3D body scans. The positional errors were found to be comparable to those from manual palpation techniques, validating the mesh structure's suitability for automated landmark detection. However, enforcing symmetry in the mesh topology introduced greater errors, indicating a trade-off between anatomical fidelity and design constraints. Study 3 explored whether Move4D data could train neural networks to improve RGB-based markerless pose estimation. A new dataset was developed, comprising 51,051 poses from 71 participants, each annotated with 53 anatomical landmarks and 33 Mediapipe keypoints. Three deep learning models were trained and tested: a multilayer perceptron, a long short-term memory network, and a Transformer model. All models achieved accuracy within the range of intra-examiner variability in static poses. Among them, the Transformer model delivered the best performance, particularly in estimating joint angles and anatomical landmarks during movements involving forward and backwards body translation. The error analysis showed that landmark prediction varied by body region, with the pelvis yielding the highest and ankles the lowest errors. Activities like running produced higher errors, while squats and jumps led to more accurate estimations. A linear mixed model analysis further revealed that error variance was primarily influenced by movement type, joint axis, and neural network architecture, while individual body shape had minimal effect, suggesting effective generalisation of the models across different body types. In summary, this thesis demonstrates that Move4D scan data significantly enhances the accuracy of deep learning-based markerless pose estimation from RGB images. The scanner's precision and its topologically consistent mesh structure enable the creation of high-quality datasets that support the training of effective pose estimation models. The research successfully fulfils its goal of developing and validating a straightforward yet robust markerless 3D pose estimation approach, marking a meaningful step forward for practical applications in motion analysis outside of laboratory settings., Activity developed within the framework of the IBERUS project. Technological Network of Biomedical Engineering applied to degenerative pathologies of the neuromusculoskeletal system in clinical and outpatient settings (CER-20211003), CER-VERA Network financed by the Ministry of Science and Innovation through the Center for Industrial Technological Development (CDTI), charged to the General State Budgets 2021 and the Recovery, Transformation and Resilience Plan. Activity developed within the framework of the "IBV 2023 Non-economic Activities Plan" (IMAMCA/2023/7), financed by the Nominative Line S8021000 distributed in favour of the technology centres of the Valencian Community, approved by the Budget Law of the Generalitat for 2023. This work was supported by the project IMDEEA/2023/77 and by the Instituto Valenciano de Competitividad Empresarial (IVACE), 2023 Call for Proposals for Technology Centers of the Valencian Region, funded by European Union and the project CONV23/DGINN/14 funded by Conselleria de Innovación, Industria, Comercio y Turismo, Grants for Technology Centers for Innovation Projects in Collaboration with Companies, within the Smart Specialization Approach (S3)
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Variabilidad interobservador del baremo de tráfico (Ley 35/2015) y utilidad de las pruebas biomecánicas en la valoración de las secuelas

  • Herrera-Ligero, Cristina
  • Vivas Broseta, María José
  • Ferrer Gómez, Carmen
  • De Rosario Martínez, Helios
  • Garrido-Lestache López-Belmonte, Elvira
  • Bermejo Bosch, Ignacio
[ES] Introducción La Ley 35/2015 valora daños y perjuicios en los accidentes de circulación y trata de garantizar una respuesta igualitaria ante situaciones idénticas. La valoración funcional biomecánica (VFB) es una prueba médica complementaria que aporta objetividad en la evaluación de la cervicalgia postraumática. Este estudio analiza la variabilidad interobservador del sistema de valoración de la Ley 35/2015 y estudia el efecto de la VFB y del perfil del evaluador en la determinación de secuelas. Material y métodos para ello, 49 profesionales de la valoración del daño corporal evaluaron 5 casos clínicos de accidentados con cervicalgia postraumática; un mes más tarde, 35 de los anteriores valoraron los mismos casos incluyendo informe de VFB. Resultados hay una elevada variabilidad en la determinación de los días de perjuicio personal (coeficientes kappa entre 0,04 y 0,073) con o sin VFB; una elevada variabilidad interobservador en la valoración de las secuelas en los casos sin VFB (coeficientes kappa entre 0,022 y 0,044), que mejora discretamente con VFB (coeficientes kappa entre 0,128 y 0,26), aún mostrando concordancia débil. El resultado de la VFB tiene influencia en la determinación de las secuelas (p¿<¿0,01), no así el perfil del evaluador. Más del 79,4% de los evaluadores encontraron útiles las pruebas para poner de manifiesto o confirmar sintomatología, recuperación o simulación. Conclusiones existe variabilidad en la aplicación de la Ley 35/2015 para la valoración de los accidentados de tráfico con cervicalgia postraumática entre los profesionales del daño corporal. La VFB resulta de utilidad para los evaluadores y tiene influencia en la determinación de las secuelas., [EN] Introduction Law 35/2015 assesses damages in traffic accidents and tries to guarantee an equal response to identical situations. Biomechanical functional assessment (BFA) is a complementary medical test that provides objectivity in the evaluation of post-traumatic neck pain. This study analyzes the interobserver variability of the assessment system defined by Law 35/2015 and studies the effect of having BFA tests and the profile of the evaluator in determining sequelae. Materials and methods To do this, 49 professionals in the assessment of bodily injury evaluated 5 clinical cases of accident victims with post-traumatic neck pain; a month later, 35 of them assessed the same cases including a BFA report. Results The results show high variability in determining the days of personal injury (Kappa coefficients between 0.04 and 0.073) with or without BFA; high interobserver variability in the assessment of sequelae in cases without BFA (Kappa coefficients between 0.022 and 0.044), which slightly improves with BFA (Kappa coefficients between 0.128 and 0.26), even showing weak concordance. The BFA has an influence on the determination of sequelae (p¿<¿0.01), but the profile of the evaluator does not. More than 79.4% of the evaluators found the BFA tests useful to reveal or to confirm symptoms, recovery, or simulation. Conclusions There is variability in the application of Law 35/2015 for the assessment of traffic accidents among professionals of bodily injury. The BFA is useful for evaluators and influences the determination of sequelae., Actividad desarrollada en el marco del proyecto IBERUS. Red Tecnológica de Ingeniería Biomédica aplicada a patologías degenerativas del sistema neuromusculoesquelético en entornos clínicos y extrahospitalarios (CER-20211003), Red CERVERA financiada por el Ministerio de Ciencia e Innovación a través del Centro para el Desarrollo Tecnológico Industrial (CDTI), con cargo a los Presupuestos Generales del Estado 2021 y el Plan de Recuperación, Transformación y Resiliencia. Parte de esta investigación se ha llevado a cabo en el marco del Proyecto Mov-Hum (Generación de criterios para la valoración y diseño de productos basados en el estudio de modelos biomecánicos de miembro superior y raquis), que cuenta con el apoyo del Instituto Valenciano de Competitividad Empresarial (IVACE) y la cofinanciación en un 50% por la Unión Europea a través del Fondo Europeo de Desarrollo Regional (FEDER), dentro del Programa Operativo FEDER de la Comunitat Valenciana 2014-2020, con el número de expediente IMDEEA/2018/89.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Accuracy of a 3D temporal scanning system for gait analysis: Comparative with a marker-based photogrammetry system

  • RUESCAS NICOLAU, ANA VIRGINIA
  • De Rosario Martínez, Helios
  • Basso Della-Vedova, Fermín
  • Parrilla Bernabé, Eduardo
  • Lopez Pascual, Juan
  • Juan, M. Carmen|||0000-0002-8764-1470
[EN] Background Combining the accuracy of marker¿based stereophotogrammetry and the usability and comfort of markerless human movement analysis is a difficult challenge. 3D temporal scanners are a promising solution, since they provide moving meshes with thousands of vertices that can be used to analyze human movements. Research question Can a 3D temporal scanner be used as a markerless system for gait analysis with the same accuracy as traditional, marker¿based stereophotogrammetry systems? Methods A comparative study was carried out using a 3D temporal scanner synchronized with a marker¿based stereophotogrammetry system. Two gait cycles of twelve healthy adults were measured simultaneously, extracting the positions of key anatomical points from both systems, and using them to analyze the 3D kinematics of the pelvis, right hip and knee joints. Measurement differences of marker positions and joint angles were described by their root mean square. A t¿test was performed to rule out instrumental errors, and an F¿test to evaluate the amplifications of marker position errors in dynamic conditions. Results The differences in 3D landmark positions were between 1.9 and 2.4 mm in the reference pose. Marker position errors were significantly increased during motion in the medial¿lateral and vertical directions. The angle relative errors were between 3% and 43% of the range of motion, with the greatest difference being observed in hip axial rotation. Significance The differences in the results obtained between the 3D temporal scanner and the marker¿based system were smaller than the usual errors due to lack of accuracy in the manual positioning of markers on anatomical landmarks and to soft¿tissue artefacts. That level of accuracy is greater than other markerless systems, and proves that such technology is a good alternative to traditional, marker¿based motion capture., Activity developed within the framework of the IBERUS project. Technological Network of Biomedical Engineering applied to degenerative pathologies of the neuromusculoskeletal system in clinical and outpatient settings (CER-20211003), CERVERA Network financed by the Ministry of Science and Innovation through the Center for Industrial Technological Development (CDTI), charged to the General State Budgets 2021 and the Recovery, Transformation and Resilience Plan.