Pasar al contenido principal

Resultados relacionados

Resultados totales (Incluyendo duplicados): 18
Encontrada(s) 1 página(s)

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

Towards an in-plane methodology to track breast lesions using mammograms and patient-specific finite-element simulations

  • Lapuebla-Ferri, Andrés|||0000-0001-9667-236X
  • Jimenez Mocholi, Antonio José
  • Cegoñino-Banzo, José
  • Perez del Palomar, Amaya
[EN] In breast cancer screening or diagnosis, it is usual to combine different images in order to locate a lesion as accurately as possible. These images are generated using a single or several imaging techniques. As x-ray-based mammography is widely used, a breast lesion is located in the same plane of the image (mammogram), but tracking it across mammograms corresponding to different views is a challenging task for medical physicians. Accordingly, simulation tools and methodologies that use patient-specific numerical models can facilitate the task of fusing information from different images.Additionally, these tools need to be as straightforward as possible to facilitate their translation to the clinical area. This paper presents a patient-specific, finite-element-based and semi-automated simulation methodology to track breast lesions across mammograms. A realistic three-dimensional computer model of a patient¿s breast was generated from magnetic resonance imaging to simulate mammographic compressions in cranio-caudal (CC, head-to-toe) and medio-lateral oblique (MLO, shoulder-to-opposite hip) directions. For each compression being simulated, a virtual mammogram was obtained and posteriorly superimposed to the corresponding real mammogram, by sharing the nipple as a common feature. Two-dimensional rigid-body transformations were applied, and the error distance measured between the centroids of the tumors previously located on each image was 3.84 mm and 2.41 mm for CC and MLO compression, respectively. Considering that the scope of this work is to conceive a methodology translatable to clinical practice, the results indicate that it could be helpful in supporting the tracking of breast lesions., This work was supported by the Spanish Ministry of Economy and Competitiveness through project DPI 2016-79302-R. The authors declare that there are no conflicts of interest.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publicación Artículo científico (JournalArticle)

Towards an in-plane methodology to track breast lesions using mammograms and patient-specific finite-element simulations

  • Lapuebla-Ferri, A.
  • Cegoñino-Banzo, J.
  • Jiménez-Mocholí, A.J.
  • Pérez Del Palomar, A.
In breast cancer screening or diagnosis, it is usual to combine different images in order to locate a lesion as accurately as possible. These images are generated using a single or several imaging techniques. As x-ray-based mammography is widely used, a breast lesion is located in the same plane of the image (mammogram), but tracking it across mammograms corresponding to different views is a challenging task for medical physicians. Accordingly, simulation tools and methodologies that use patient-specific numerical models can facilitate the task of fusing information from different images. Additionally, these tools need to be as straightforward as possible to facilitate their translation to the clinical area. This paper presents a patient-specific, finite-element-based and semi-automated simulation methodology to track breast lesions across mammograms. A realistic three-dimensional computer model of a patient''s breast was generated from magnetic resonance imaging to simulate mammographic compressions in cranio-caudal (CC, head-to-toe) and medio-lateral oblique (MLO, shoulder-to-opposite hip) directions. For each compression being simulated, a virtual mammogram was obtained and posteriorly superimposed to the corresponding real mammogram, by sharing the nipple as a common feature. Two-dimensional rigid-body transformations were applied, and the error distance measured between the centroids of the tumors previously located on each image was 3.84 mm and 2.41 mm for CC and MLO compression, respectively. Considering that the scope of this work is to conceive a methodology translatable to clinical practice, the results indicate that it could be helpful in supporting the tracking of breast lesions.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publicación Artículo científico (JournalArticle)

Stand-alone lumbar cage subsidence: A biomechanical sensitivity study of cage design and placement.

  • Calvo-Echenique, A.
  • Cegoñino, J.
  • Chueca, R.
  • Pérez-del Palomar, A.
Background and objective: Spinal degeneration and instability are commonly treated with interbody fusion cages either alone or supplemented with posterior instrumentation with the aim to immobilise the segment and restore intervertebral height. The purpose of this work is to establish a tool which may help to understand the effects of intervertebral cage design and placement on the biomechanical response of a patient-specific model to help reducing post-surgical complications such as subsidence and segment instability. Methods: A 3D lumbar functional spinal unit (FSU) finite element model was created and a parametric model of an interbody cage was designed and introduced in the FSU. A Drucker–Prager Cap plasticity formulation was used to predict plastic strains and bone failure in the vertebrae. The effect of varying cage size, cross-sectional area, apparent stiffness and positioning was evaluated under 500 N preload followed by 7.5 Nm multidirectional rotation and the results were compared with the intact model. Results: The most influential cage parameters on the FSU were size, curvature congruence with the endplates and cage placement. Segmental stiffness was higher when increasing the cross-sectional cage area in all loading directions and when the cage was anteriorly placed in all directions but extension. In general, the facet joint forces were reduced by increasing segmental stiffness. However, these forces were higher than in the intact model in most of the cases due to the displacement of the instantaneous centre of rotation. The highest plastic deformations took place at the caudal vertebra under flexion and increased for cages with greater stiffness. Thus, wider cages and a more anteriorly placement would increase the volume of failed bone and, therefore, the risk of subsidence. Conclusions: Cage geometry plays a crucial role in the success of lumbar surgery. General considerations such as larger cages may be applied as a guideline, but parameters such as curvature or cage placement should be determined for each specific patient. This model provides a proof-of-concept of a tool for the preoperative evaluation of lumbar surgical outcomes.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publicación Artículo científico (JournalArticle)

Is there any advantage of using stand-alone cages? A numerical approach

  • Calvo-Echenique, A.
  • Cegonino, J.
  • Pérez del Palomar, A.
Background: Segment fusion using interbody cages supplemented with pedicle screw fixation is the most common surgery for the treatment of low back pain. However, there is still much controversy regarding the use of cages in a stand-alone fashion. The goal of this work is to numerically compare the influence that each surgery has on lumbar biomechanics. Methods: A non-linear FE model of the whole lumbar spine was developed to compare between two types of cages (OLYS and NEOLIF) with and without supplementary fixation. The motion of the whole spine was analysed and the biomechanical environment of the adjacent segments to the operated one was studied. Moreover, the risk of subsidence of the cages was qualitatively evaluated. Results: A great ROM reduction occurred when supplementary fixation was used. This stiffening increased the stresses at the adjacent levels. It might be hypothesised that the overloading of these segments could be related with the clinically observed adjacent disc degeneration. Meanwhile, the stand-alone cages allowed for a wider movement, and therefore, the influence of the surgery on adjacent discs was much lower. Regarding the risk of subsidence, the contact pressure magnitude was similar for both intervertebral cage designs and near the value of the maximum tolerable pressure of the endplates. Conclusions: A minimally invasive posterior insertion of an intervertebral cage (OLYS or NEOLIF) was compared using a stand-alone design or adding supplementary fixation. The outcomes of these two techniques were compared, and although stand-alone cage may diminish the risk of disease progression to the adjacent discs, the spinal movement in this case could compromise the vertebral fusion and might present a higher risk of cage subsidence.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publicación Artículo científico (JournalArticle)

In silico study of cuspid' periodontal ligament damage under parafunctional and traumatic conditions of whole-mouth occlusions. A patient-specific evaluation

  • Ortún-Terrazas, Javier
  • Cegoñino, José
  • Pérez del Palomar, Amaya
Background and objective: Although traumatic loading has been associated with periodontal ligament (PDL) damage and therefore with several oral disorders, the damage phenomena and the traumatic loads involved are still unclear. The complex composition and extremely thin size of the PDL make experimentation difficult, requiring computational studies that consider the macroscopic loading conditions, the microscopic composition and fine detailed geometry of the tissue. In this study, a new methodology to analyse the damage phenomena in the collagen network and the extracellular matrix of the PDL caused by parafunctional and traumatic occlusal forces was proposed. Methods: The entire human mandible and a portion thereof containing a full cuspid tooth were separately modelled using finite element analysis based on computed tomography and micro-computed tomography images, respectively. The first model was experimentally validated by occlusion analysis and subjected to the muscle loads produced during hard and soft chewing, traumatic cuspid occlusion, grinding, clenching, and simultaneous grinding and clenching. The occlusal forces computed by the first model were subsequently applied to the single tooth model to evaluate damage to the collagen network and the extracellular matrix of the PDL. Results: Early occlusal contact on the left cuspid tooth guided the mandible to the more occluded side (16.5% greater in the right side) and absorbed most of the lateral load. The intrusive occlusal loads on the posterior teeth were 0.77–13.3% greater than those on the cuspid. According to our findings, damage to the collagen network and the extracellular matrix of the PDL could occur in traumatic and grinding conditions, mainly due to fibre overstretching (>60%) and interstitial fluid overpressure (>4.7 kPa), respectively. Conclusions: Our findings provide important biomechanical insights into the determination of damage mechanisms which are caused by mechanical loading and the key role of the porous-fibrous behaviour of the PDL in parafunctional and traumatic loading scenarios. Besides, the 3D loading conditions computed from occlusal contacts will help future studies in the design of new orthodontics appliances and encourage the application of computing methods in medical practice.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publicación Artículo científico (JournalArticle)

A porous fibrous hyperelastic damage model for human periodontal ligament: Application of a microcomputerized tomography finite element model

  • Ortún-Terrazas, J.
  • Cegoñino, J.
  • Santana-Penín, U.
  • Santana-Mora, U.
  • Pérez del Palomar, A.
The periodontal ligament (PDL) is a soft biological tissue that connects the tooth with the trabecular bone of the mandible. It plays a key role in load transmission and is primarily responsible for bone resorption and most common periodontal diseases. Although several numerical studies have analysed the biomechanical response of the PDL, most did not consider its porous fibrous structure, and only a few analysed damage to the PDL. This study presents an innovative numerical formulation of a porous fibrous hyperelastic damage material model for the PDL. The model considers two separate softening phenomena: fibre alignment during loading and fibre rupture. The parameters for the material model characterization were fitted using experimental data from the literature. Furthermore, the experimental tests used for characterization were computationally modelled to verify the material parameters. A finite element model of a portion of a human mandible, obtained by microcomputerized tomography, was developed, and the proposed constitutive model was implemented for the PDL. Our results confirm that damage to the PDL may occur mainly because of overpressure of the interstitial fluid, while large forces must be applied to damage the PDL fibrous network. Moreover, this study clarifies some aspects of the relationship between PDL damage and the bone remodelling process.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publicación Artículo científico (JournalArticle)

Towards an early 3D-diagnosis of craniofacial asymmetry by computing the accurate midplane: A PCA-based method

  • Ortún-Terrazas, Javier
  • Fagan, Michael J.
  • Cegoñino, José
  • Illipronti-Filho, Edson
  • Pérez del Palomar, Amaya
Background and objective: Craniofacial asymmetry is a common growth disorder often caused by unilateral chewing. Although an early orthodontic treatment would avoid surgical procedures later in life, the uncertainty of defining the accurate sagittal midplane potentially leads to misdiagnosis and therefore inaccurate orthodontic treatment plans. This novel study aims to 3D-diagnose craniofacial complex malformations in children with unilateral crossbite (UXB) considering a midplane which compensates the asymmetric morphology. Methods: The sagittal midplane of 20 children, fifteen of whom exhibited UXB, was computed by a PCA-based method which compensates the asymmetry mirroring the 3D models obtained from cone-beam computed tomography data. Once determined, one side of the data was mirrored using the computed midplane to visualize the malformations on the hard and soft tissues by 3D-computing the distances between both halves. Additionally, 31 skull's landmarks were manually placed in each model to study the principal variation modes and the significant differences in the group of subjects with and without UXB through PCA and Mann-Whitney U test analyses respectively. Results: Morphological 3D-analysis showed pronounced deformities and aesthetic implications for patients with severe asymmetry (jaw deviation > 0.8¿mm) in whole craniofacial system, while initial signs of asymmetry were found indistinctly in the mandible or maxilla. We detected significant (p < 0.05) malformations for example in mandibular ramus length (0.0086), maxillary palate width (0.0481) and condylar head width (0.0408). Craniofacial malformations increased the landmarks’ variability in the group of patients with UXB over the control group requiring 8 variation modes more to define 99% of the sample’ variability. Conclusions: Our findings demonstrated the viability of early diagnosis of craniofacial asymmetry through computing the accurate sagittal midplane which compensates the individual's asymmetrical morphology. Furthermore, this study provides important computational insights into the determination of craniofacial deformities which are caused by UXB, following some empirical findings of previous clinical studies. Hence, this computational approach can be useful for the development of new software in craniofacial surgery or for its use in biomedical research and clinical practice.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publicación Artículo científico (JournalArticle)

Numerical simulations of bone remodelling and formation following nucleotomy

  • Calvo-Echenique, A.
  • Bashkuev, M.
  • Reitmaier, S.
  • Pérez-del Palomar, A.
  • Schmidt, H.
Nucleotomy is the gold standard treatment for disc herniation and has proven ability to restore stability by creating a bony bridge without any additional fixation. However, the evolution of mineral density in the extant and new bone after nucleotomy and fixation techniques has to date not been investigated in detail. The main goal of this study is to determine possible mechanisms that may trigger the bone remodelling and formation processes. With that purpose, a finite element model of the L4–L5 spinal segment was used. Bone mineral density (BMD), new tissue composition, and endplate deflection were determined as indicators of lumbar fusion. A bone-remodelling algorithm and a tissue-healing algorithm, both mechanically driven, were implemented to predict vertebral bone alterations and fusion patterns after nucleotomy, internal fixation, and anterior plate placement. When considering an intact disc height, neither nucleotomy nor internal fixation were able to provide the necessary stability to promote bony fusion. However, when 75% of the disc height was considered, bone fusion was predicted for both techniques. By contrast, an anterior plate allowed bone fusion at all disc heights. A 50% disc-height reduction led to osteophyte formation in all cases. Changes in the intervertebral disc tissue caused BMD alterations in the endplates. From this observations it can be drawn that fusion may be self-induced by controlling the mechanical stabilisation without the need of additional fixation. The amount of tissue to be removed to achieve this stabilisation remains to be determined.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publicación Artículo científico (JournalArticle)

Analysis of temporomandibular joint dysfunction in paediatric patients with unilateral crossbite using automatically generated finite element models

  • Ortún-Terrazas, J.
  • Cegoñino, J.
  • Illipronti-Filho, E.
  • Pérez del Palomar, A.
The evaluation of temporomandibular joint (TMJ) dysfunction using finite element models is a time consuming process that requires extensive technical knowledge. We combined a statistical active appearance model with automated modelling algorithms to biomechanically study the relationship between TMJ malformations and dysfunction in radiographs from 20 paediatric patients with unilateral crossbite. A fitting algorithm (fitting error < 4%) recognised the TMJ shape and adjusted the dimensions of each patient-specific 2D FE model, which was then used to compute 2 different joint movements. Significant functional differences were observed between the crossbite and non-cross bite sides, and the shape-function relation was verified.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publicación Artículo científico (JournalArticle)

The Innate Immune Cell Profile of the Cornea Predicts the Onset of Ocular Surface Inflammatory Disorders

  • Perez del Palomar, Amaya
  • Montolio, Alberto
  • Cegonino, Jose
  • Dhanda, Sandeep Kumar
  • Lio, Chit Tong
  • Bose, Tanima
Ocular surface inflammatory disorder (OSID) is a spectrum of disorders that have features of several etiologies whilst displaying similar phenotypic signs of ocular inflammation. They are complicated disorders with underlying mechanisms related to several autoimmune disorders, such as rheumatoid arthritis (RA), Sjogren's syndrome, and systemic lupus erythematosus (SLE). Current literature shows the involvement of both innate and adaptive arms of the immune system in ocular surface inflammation. The ocular surface contains distinct components of the immune system in the conjunctiva and the cornea. The normal conjunctiva epithelium and sub-epithelial stroma contains resident immune cells, such as T cells, B cells (adaptive), dendritic cells, and macrophages (innate). The relative sterile environment of the cornea is achieved by the tolerogenic properties of dendritic cells in the conjunctiva, the presence of regulatory lymphocytes, and the existence of soluble immunosuppressive factors, such as the transforming growth factor (TGF)-beta and macrophage migration inhibitory factors. With the presence of both innate and adaptive immune system components, it is intriguing to investigate the most important leukocyte population in the ocular surface, which is involved in immune surveillance. Our meta-analysis investigates into this with a focus on both infectious (contact lens wear, corneal graft rejection, Cytomegalovirus, keratitis, scleritis, ocular surgery) and non-infectious (dry eye disease, glaucoma, graft-vs-host disease, Sjogren's syndrome) situations. We have found the predominance of dendritic cells in ocular surface diseases, along with the Th-related cytokines. Our goal is to improve the knowledge of immune cells in OSID and to open new dimensions in the field. The purpose of this study is not to limit ourselves in the ocular system, but to investigate the importance of dendritic cells in the disorders of other mucosal organs (e.g., lungs, gut, uterus). Holistically, we want to investigate if this is a common trend in the initiation of any disease related to the mucosal organs and find a unified therapeutic approach. In addition, we want to show the power of computational approaches to foster a collaboration between computational and biological science.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publicación Artículo científico (JournalArticle)

Swept source optical coherence tomography to early detect multiple sclerosis disease. The use of machine learning techniques

  • Pérez del Palomar, Amaya
  • Cegoñino, José
  • Montolío, Alberto
  • Orduna, Elvira
  • Vilades, Elisa
  • Sebastián, Berta
  • Pablo, Luis E.
  • Garcia-Martin, Elena
Objective To compare axonal loss in ganglion cells detected with swept-source optical coherence tomography (SS-OCT) in eyes of patients with multiple sclerosis (MS) versus healthy controls using different machine learning techniques. To analyze the capability of machine learning techniques to improve the detection of retinal nerve fiber layer (RNFL) and the complex Ganglion Cell Layer–Inner plexiform layer (GCL+) damage in patients with multiple sclerosis and to use the SS-OCT as a biomarker to early predict this disease. Methods Patients with relapsing-remitting MS (n = 80) and age-matched healthy controls (n = 180) were enrolled. Different protocols from the DRI SS-OCT Triton system were used to obtain the RNFL and GCL+ thicknesses in both eyes. Macular and peripapilar areas were analyzed to detect the zones with higher thickness decrease. The performance of different machine learning techniques (decision trees, multilayer perceptron and support vector machine) for identifying RNFL and GCL+ thickness loss in patients with MS were evaluated. Receiver-operating characteristic (ROC) curves were used to display the ability of the different tests to discriminate between MS and healthy eyes in our population. Results Machine learning techniques provided an excellent tool to predict MS disease using SS-OCT data. In particular, the decision trees obtained the best prediction (97.24%) using RNFL data in macular area and the area under the ROC curve was 0.995, while the wide protocol which covers an extended area between macula and papilla gave an accuracy of 95.3% with a ROC of 0.998. Moreover, it was obtained that the most significant area of the RNFL to predict MS is the macula just surrounding the fovea. On the other hand, in our study, GCL+ did not contribute to predict MS and the different machine learning techniques performed worse in this layer than in RNFL. Conclusions Measurements of RNFL thickness obtained with SS-OCT have an excellent ability to differentiate between healthy controls and patients with MS. Thus, the use of machine learning techniques based on these measures can be a reliable tool to help in MS diagnosis.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publicación Artículo científico (JournalArticle)

Physiological changes in retinal layers thicknesses measured with swept source optical coherence tomography

  • Viladés, E.
  • Perez del Palomar, A.
  • Cegoñino, J.
  • Obis, J.
  • Satue, M.
  • Orduna, E.
  • Pablo, L.E.
  • Ciprés, M.
  • Garcia Martin, E.
Purpose To evaluate the physiological changes related with age of all retinal layers thickness measurements in macular and peripapillary areas in healthy eyes. Methods Wide protocol scan (with a field of view of 12x9 cm) from Triton SS-OCT instrument (Topcon Corporation, Japan) was performed 463 heathy eyes from 463 healthy controls. This protocol allows to measure the thickness of the following layers: Retina, Retinal nerve fiber layer (RNFL), Ganglion cell layer (GCL +), GCL++ and choroid. In those layers, mean thickness was compared in four groups of ages: Group 1 (71 healthy subjects aged between 20 and 34 years); Group 2 (65 individuals aged 35–49 years), Group 3 (230 healthy controls aged 50–64 years) and Group 4 (97 healthy subjects aged 65–79 years). Results The most significant thinning of all retinal layers occurs particularly in the transition from group 2 to group 3, especially in temporal superior quadrant at RNFL, GCL++ and retinal layers (p=0.001), and temporal superior, temporal inferior, and temporal half in choroid layer (p<0.001). Curiously group 2 when compared with group 1 presents a significant thickening of RNFL in temporal superior quadrant (p = 0.001), inferior (p<0.001) and temporal (p = 0.001) halves, and also in nasal half in choroid layer (p = 0.001). Conclusions Excepting the RNFL, which shows a thickening until the third decade of life, the rest of the layers seem to have a physiological progressive thinning. Copyright:
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publicación Artículo científico (JournalArticle)

Comparison of Machine Learning Methods Using Spectralis OCT for Diagnosis and Disability Progression Prognosis in Multiple Sclerosis

  • Montolío, A.
  • Cegoñino, J.
  • Garcia-Martin, E.
  • Pérez del Palomar, A.
Machine learning approaches in diagnosis and prognosis of multiple sclerosis (MS) were analysed using retinal nerve fiber layer (RNFL) thickness, measured by optical coherence tomography (OCT). A cross-sectional study (72 MS patients and 30 healthy controls) was used for diagnosis. These 72 MS patients were involved in a 10-year longitudinal follow-up study for prognostic purposes. Structural measurements of RNFL thickness were performed using different Spectralis OCT protocols: fast macular thickness protocol to measure macular RNFL, and fast RNFL thickness protocol and fast RNFL-N thickness protocol to measure peripapillary RNFL. Binary classifiers such as multiple linear regression (MLR), support vector machines (SVM), decision tree (DT), k-nearest neighbours (k-NN), Naïve Bayes (NB), ensemble classifier (EC) and long short-term memory (LSTM) recurrent neural network were tested. For MS diagnosis, the best acquisition protocol was fast macular thickness protocol using k-NN (accuracy: 95.8%; sensitivity: 94.4%; specificity: 97.2%; precision: 97.1%; AUC: 0.958). For MS prognosis, our model with a 3-year follow up to predict disability progression 8 years later was the best predictive model. DT performed best for fast macular thickness protocol (accuracy: 91.3%; sensitivity: 90.0%; specificity: 92.5%; precision: 92.3%; AUC: 0.913) and SVM for fast RNFL-N thickness protocol (accuracy: 91.3%; sensitivity: 87.5%; specificity: 95.0%; precision: 94.6%; AUC: 0.913). This work concludes that measurements of RNFL thickness obtained with Spectralis OCT have a good ability to diagnose MS and to predict disability progression in MS patients. This machine learning approach would help clinicians to have valuable information. © 2022, The Author(s).
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publicación Artículo científico (JournalArticle)

The macular retinal ganglion cell layer as a biomarker for diagnosis and prognosis in multiple sclerosis: A deep learning approach

  • Montolío, Alberto
  • Cegoñino, José
  • Garcia-Martin, Elena
  • Pérez del Palomar, Amaya
AbstractPurposeThe macular ganglion cell layer (mGCL) is a strong potential biomarker of axonal degeneration in multiple sclerosis (MS). For this reason, this study aims to develop a computer‐aided method to facilitate diagnosis and prognosis in MS.MethodsThis paper combines a cross‐sectional study of 72 MS patients and 30 healthy control subjects for diagnosis and a 10‐year longitudinal study of the same MS patients for the prediction of disability progression, during which the mGCL was measured using optical coherence tomography (OCT). Deep neural networks were used as an automatic classifier.ResultsFor MS diagnosis, greatest accuracy (90.3%) was achieved using 17 features as inputs. The neural network architecture comprised the input layer, two hidden layers and the output layer with softmax activation. For the prediction of disability progression 8 years later, accuracy of 81.9% was achieved with a neural network comprising two hidden layers and 400 epochs.ConclusionWe present evidence that by applying deep learning techniques to clinical and mGCL thickness data it is possible to identify MS and predict the course of the disease. This approach potentially constitutes a non‐invasive, low‐cost, easy‐to‐implement and effective method.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publicación Artículo científico (JournalArticle)

A mathematical model to predict the evolution of retinal nerve fiber layer thinning in multiple sclerosis patients

  • Montolío, Alberto
  • Cegoñino, José
  • Orduna, Elvira
  • Sebastian, Berta
  • Garcia-Martin, Elena
  • Pérez del Palomar, Amaya
Multiple sclerosis (MS) is a neurodegenerative disease of the central nervous system (CNS). Many studies of MS patients have described axonal loss in the optic nerve of the retina, and specifically progressive thinning of the retinal nerve fiber layer (RNFL). We hypothesize that RNFL thinning involves the participation of 2 processes that cause CNS damage: autoimmune inflammation and axonal degeneration. To test this hypothesis, we developed a mathematical model based on ordinary differential equations to relate the evolution of RNFL thickness (measured by optical coherence tomography [OCT]) with that of the Expanded Disability Status Scale (EDSS) score in MS patients. Data were obtained from a longitudinal study of 114 MS patients who were followed-up for 10 years. After adjusting the parameters using a genetic algorithm, the model''s prediction of the evolution of RNFL thickness accurately reflected the progression revealed by the 10-year clinical data. Our findings suggest that differences in the relative contributions of autoimmune inflammation and axonal degeneration can account for the complex dynamics of MS, which vary from one patient to the next. Moreover, our results show that CNS damage occurs cumulatively from the onset of MS and that most RNFL thinning occurs before the appearance of significant disability. RNFL thickness could therefore serve as a reliable biomarker of MS disease course. Our proposed methodology would enable the use of OCT data from new MS patients to predict the evolution of RNFL thinning and hence the progression of MS in individual patients, and to facilitate the selection of patient-specific therapies.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publicación Artículo científico (JournalArticle)

Machine learning in diagnosis and disability prediction of multiple sclerosis using optical coherence tomography

  • Montolío A.
  • Martín-Gallego A.
  • Cegoñino J.
  • Orduna E.
  • Vilades E.
  • Garcia-Martin E.
  • Pérez del Palomar A.
Background: Multiple sclerosis (MS) is a neurodegenerative disease that affects the central nervous system, especially the brain, spinal cord, and optic nerve. Diagnosis of this disease is a very complex process and generally requires a lot of time. In addition, treatments are applied without any information on the disability course in each MS patient. For these two reasons, the objective of this study was to improve the MS diagnosis and predict the long-term course of disability in MS patients based on clinical data and retinal nerve fiber layer (RNFL) thickness, measured by optical coherence tomography (OCT). Material and methods: A total of 104 healthy controls and 108 MS patients, 82 of whom had a 10-year follow-up, were enrolled. Classification algorithms such as multiple linear regression (MLR), support vector machines (SVM), decision tree (DT), k-nearest neighbours (k-NN), Naïve Bayes (NB), ensemble classifier (EC) and long short-term memory (LSTM) recurrent neural network were tested to develop two predictive models: MS diagnosis model and MS disability course prediction model. Results: For MS diagnosis, the best result was obtained using EC (accuracy: 87.7%; sensitivity: 87.0%; specificity: 88.5%; precision: 88.7%; AUC: 0.8775). In line with this good performance, the accuracy was 85.4% using k-NN and 84.4% using SVM. And, for long-term prediction of MS disability course, LSTM recurrent neural network was the most appropriate classifier (accuracy: 81.7%; sensitivity: 81.1%; specificity: 82.2%; precision: 78.9%; AUC: 0.8165). The use of MLR, SVM and k-NN also showed a good performance (AUC = 0.8). Conclusions: This study demonstrated that machine learning techniques, using clinical and OCT data, can help establish an early diagnosis and predict the course of MS. This advance could help clinicians select more specific treatments for each MS patient. Therefore, our findings underscore the potential of RNFL thickness as a reliable MS biomarker. © 2021 The Author(s)
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publicación Artículo científico (JournalArticle)

Computational characterization of the porous-fibrous behavior of the soft tissues in the temporomandibular joint

  • Ortún-Terrazas, J.
  • Cegoñino, J.
  • Pérez del Palomar, A.
The prevalence and severity of temporomandibular joint (TMJ) disorders have led to growing research interest in the development of new biomaterials and medical devices for TMJ implant designs. In computational designs, however, the time and stretch direction dependences of the TMJ soft tissues behavior are not considered and they are frequently based on measurements taken from non-human species or from joints that differ markedly from the human TMJ. The aim of this study was to accurately characterize the porous-fibrous properties of the TMJ soft tissues by simulating previously published experimental tests, to assist professionals in the design of new TMJ implants. To that end, material parameters were determined assuming a uniform fiber orientation throughout the entire sample. This assumption was then tested by comparing these results with those of considering multiple regions and distinct fiber orientations in each sample. Our findings validated the use of a transversely isotropic hyperelastic material model to characterize the direction dependent behavior of TMJ soft tissues and its combination with porous hyperfoam material models to mimic the compressive response of the TMJ disc. In conclusion, constitutive model proposed accurately reproduce the mechanical response of the TMJ soft tissues at different strain rates and stretch directions.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publicación Artículo científico (JournalArticle)

In silico approach towards Neuro-Occlusal Rehabilitation for the early correction of asymmetrical development in a unilateral crossbite patient.

  • Ortún-Terrazas, Javier
  • Cegoñino Banzo, José
  • Pérez del Palomar Aldea, Amaya
Neuro-occlusal rehabilitation (N.O.R.) is a discipline of the stomatognathic medicine that defends early treatments of functional malocclusions, such as unilateral crossbite, for the correction of craniofacial development, avoiding surgical procedures later in life. Nevertheless, N.O.R.'s advances have not been proved analytically yet due to the difficulties of evaluate the mechanical response after the treatment. This study aims to evaluate computationally the effect of N.O.R.'s treatments during childhood. Therefore, bilateral chewing and maximum intercuspation occlusion were modelled through a detailed finite element model of a paediatric craniofacial complex, before and after different selective grinding-alternatives. This model was subjected to the muscular forces derived from a musculoskeletal model and was validated by the occlusal contacts recorded experimentally. This approach yielded errors below 2% and reproduced successfully the occlusal, muscular, functional and mechanical imbalance before the therapies. Treatment strategies balanced the occlusal plane and reduced the periodontal overpressure (>4.7 kPa) and the mandibular over deformation (>0.002 ε) on the crossed side. Based on the principles of the mechanostat theory of bone remodelling and the pressure-tension theory of tooth movement, these findings could also demonstrate how N.O.R.'s treatments correct the malocclusion and the asymmetrical development of the craniofacial complex. Besides, N.O.R.'s treatments slightly modified the stress state and functions of the temporomandibular joints, facilitating the chewing by the unaccustomed side. These findings provide important biomechanical insights into the use of N.O.R.'s treatments for the correction of unilateral crossbite, but also encourage the application of computing methods in biomedical research and clinical practise.