Methodology for registration of distended recutms in pelvic CT studies

Purpose: Accurate delineation of the rectum is of high importance in off-line adaptive radiation therapy since it is a major dose-limiting organ in prostate cancer radiotherapy. The intensity-based deformable image registration (DIR) methods cannot create a correct spatial transformation if there is no correspondence between the template and the target images. The variation of rectal filling, gas, or feces, creates a noncorrespondence in image intensities that becomes a great obstacle for intensity-based DIR. Methods: In this study the authors have designed and implemented a semiautomatic method to create a rectum mask in pelvic computed tomography (CT) images. The method, that includes a DIR based on the demons algorithm, has been tested in 13 prostate cancer cases, each comprising of two CT scans, for a total of 26 CT scans. Results: The use of the manual segmentation in the planning image and the proposed rectum mask method (RMM) method in the daily image leads to an improvement in the DIR performance in pelvic CT images, obtaining a mean value of overlap volume index = 0.89, close to the values obtained using the manual segmentations in both images. Conclusions: The application of the RMM method in the daily image and the manual segmentations in the planning image during prostate cancer treatments increases the performance of the registration in presence of rectal fillings, obtaining very good agreement with a physician's manual contours.

TELMA: Technology enhanced learning environment for minimally invasive surgery

Background: Cognitive skills training for minimally invasive surgery has traditionally relied upon diverse tools, such as seminars or lectures. Web technologies for e-learning have been adopted to provide ubiquitous training and serve as structured repositories for the vast amount of laparoscopic video sources available. However, these technologies fail to offer such features as formative and summative evaluation, guided learning, or collaborative interaction between users. Methodology: The "TELMA" environment is presented as a new technology-enhanced learning platform that increases the user's experience using a four-pillared architecture: (1) an authoring tool for the creation of didactic contents; (2) a learning content and knowledge management system that incorporates a modular and scalable system to capture, catalogue, search, and retrieve multimedia content; (3) an evaluation module that provides learning feedback to users; and (4) a professional network for collaborative learning between users. Face validation of the environment and the authoring tool are presented. Results: Face validation of TELMA reveals the positive perception of surgeons regarding the implementation of TELMA and their willingness to use it as a cognitive skills training tool. Preliminary validation data also reflect the importance of providing an easy-to-use, functional authoring tool to create didactic content. Conclusion: The TELMA environment is currently installed and used at the Jesús Usón Minimally Invasive Surgery Centre and several other Spanish hospitals. Face validation results ascertain the acceptance and usefulness of this new minimally invasive surgery training environment.

Learning curves of basic laparoscopic psychomotor skills in SINERGIA VR simulator

Purpose: Surgical simulators are currently essential within any laparoscopic training program because they provide a low-stakes, reproducible and reliable environment to acquire basic skills. The purpose of this study is to determine the training learning curve based on different metrics corresponding to five tasks included in SINERGIA laparoscopic virtual reality simulator. Methods: Thirty medical students without surgical experience participated in the study. Five tasks of SINERGIA were included: Coordination, Navigation, Navigation and touch, Accurate grasping and Coordinated pulling. Each participant was trained in SINERGIA. This training consisted of eight sessions (R1–R8) of the five mentioned tasks and was carried out in two consecutive days with four sessions per day. A statistical analysis was made, and the results of R1, R4 and R8 were pair-wise compared with Wilcoxon signed-rank test. Significance is considered at P value <0.005. Results: In total, 84.38% of the metrics provided by SINERGIA and included in this study show significant differences when comparing R1 and R8. Metrics are mostly improved in the first session of training (75.00% when R1 and R4 are compared vs. 37.50% when R4 and R8 are compared). In tasks Coordination and Navigation and touch, all metrics are improved. On the other hand, Navigation just improves 60% of the analyzed metrics. Most learning curves show an improvement with better results in the fulfillment of the different tasks. Conclusions: Learning curves of metrics that assess the basic psychomotor laparoscopic skills acquired in SINERGIA virtual reality simulator show a faster learning rate during the first part of the training. Nevertheless, eight repetitions of the tasks are not enough to acquire all psychomotor skills that can be trained in SINERGIA. Therefore, and based on these results together with previous works, SINERGIA could be used as training tool with a properly designed training program.

Data mining applied to the cognitive rehabilitation of patients with acquired brain injury

Acquired brain injury (ABI) is one of the leading causes of death and disability in the world and is associated with high health care costs as a result of the acute treatment and long term rehabilitation involved. Different algorithms and methods have been proposed to predict the effectiveness of rehabilitation programs. In general, research has focused on predicting the overall improvement of patients with ABI. The purpose of this study is the novel application of data mining (DM) techniques to predict the outcomes of cognitive rehabilitation in patients with ABI. We generate three predictive models that allow us to obtain new knowledge to evaluate and improve the effectiveness of the cognitive rehabilitation process. Decision tree (DT), multilayer perceptron (MLP) and general regression neural network (GRNN) have been used to construct the prediction models. 10-fold cross validation was carried out in order to test the algorithms, using the Institut Guttmann Neurorehabilitation Hospital (IG) patients database. Performance of the models was tested through specificity, sensitivity and accuracy analysis and confusion matrix analysis. The experimental results obtained by DT are clearly superior with a prediction average accuracy of 90.38%, while MLP and GRRN obtained a 78.7% and 75.96%, respectively. This study allows to increase the knowledge about the contributing factors of an ABI patient recovery and to estimate treatment efficacy in individual patients.

Methodology for registration of distended rectums in pelvic CT studies

Purpose: Accurate delineation of the rectum is of high importance in off-line adaptive radiation therapy since it is a major dose-limiting organ in prostate cancer radiotherapy. The intensity-based deformable image registration (DIR) methods cannot create a correct spatial transformation if there is no correspondence between the template and the target images. The variation of rectal filling, gas, or feces, creates a noncorrespondence in image intensities that becomes a great obstacle for intensity-based DIR. Methods: In this study the authors have designed and implemented a semiautomatic method to create a rectum mask in pelvic computed tomography (CT) images. The method, that includes a DIR based on the demons algorithm, has been tested in 13 prostate cancer cases, each comprising of two CT scans, for a total of 26 CT scans. Results: The use of the manual segmentation in the planning image and the proposed rectum mask method (RMM) method in the daily image leads to an improvement in the DIR performance in pelvic CT images, obtaining a mean value of overlap volume index = 0.89, close to the values obtained using the manual segmentations in both images. Conclusions: The application of the RMM method in the daily image and the manual segmentations in the planning image during prostate cancer treatments increases the performance of the registration in presence of rectal fillings, obtaining very good agreement with a physician's manual contours.

Automatic generation of models for abdominal aortic aneurysms and intraluminal thrombus based on hexahedral meshes

To propose an automated patient-specific algorithm for the creation of accurate and smooth meshes of the aortic anatomy, to be used for evaluating rupture risk factors of abdominal aortic aneurysms (AAA). Finite element (FE) analyses and simulations require meshes to be smooth and anatomically accurate, capturing both the artery wall and the intraluminal thrombus (ILT). The two main difficulties are the modeling of the arterial bifurcations, and of the ILT, which has an arbitrary shape that is conforming to the aortic wall.

A method for incorporating residual stresses into patient-specific finite element simulations of arteries with an example on AAAs

Through progress in medical imaging, image analysis and finite element (FE) meshing tools it is now possible to extract patient-specific geometries from medical images of abdominal aortic aneurysms(AAAs), and thus to study clinically-relevant problems via FE simulations. Such simulations allow additional insight into human physiology in both healthy and diseased states. Medical imaging is most often performed in vivo, and hence the reconstructed model geometry in the problem of interest will represent the in vivo state, e.g., the AAA at physiological blood pressure. However, classical continuum mechanics and FE methods assume that constitutive models and the corresponding simulations begin from an unloaded, stress-free reference condition.

Extracción Automática de la Línea Central de Estructuras Tubulares: Implementación Matricial

Se propone una nueva implementación matricial de un algoritmo para la extracción automática de la línea central de estructuras tubulares. El algoritmo seleccionado calcula la línea central de estructuras complejas sin la necesidad de interacción con el usuario. En el trabajo se explica detalladamente cómo llevar a cabo la implementación matricial utilizando el lenguaje de computación de Matlab. La implementación matricial permite el cálculo de la línea central en pocos segundos, mejorando en varios grados de magnitud la implementación disponible en ITK.

Validación de dos algoritmos para la localización 3D del instrumental laparoscópico a partir de la imagen de vídeo endoscópico

Uno de los aspectos fundamentales en un sistema de cirugía guiada por imagen (CGI) es la localización del instrumental quirúrgico con respecto a la anatomía del paciente. Los sistemas basados en sensores ofrecen buenos niveles de precisión, pero son sensibles a distintas fuentes de ruido en el quirófano y contribuyen a la sobrecarga tecnológica del mismo. Una alternativa novedosa es analizar la imagen del vídeo endoscópico para llevar a cabo la detección y localización espacial del instrumental. Se presenta en este trabajo la validación de dos métodos, basados en el diámetro aparente y en la sección transversal del instrumental, para la localización espacial del instrumental a partir de los bordes y la posición 2D de la punta en la imagen. La validación, llevada a cabo en un simulador físico, se realiza comparando los resultados con el sistema Kinescan/IBV. Los resultados muestran para cada método un error medio de 12,7 y 12,8 mm respectivamente. La incorporación de estos algoritmos dentro del paradigma de navegación propuesto en el proyecto THEMIS permitirá al cirujano conocer la posición del instrumental de forma no intrusiva y transparente, sin necesidad de equipamiento adicional en el quirófano.

Integración de tecnologías eye-tracking con vídeos interactivos para neurorrehabilitación cognitiva

La integración de las nuevas tecnologías en el proceso de rehabilitación permite la generación de terapias personalizadas, ubicuas y basadas en la evidencia. Tecnologías como el vídeo interactivo son propicias para el desarrollo de entornos virtuales en los que el paciente se ve inmerso dentro de actividades de la vida diaria en los que tiene que lograr un objetivo ecológico en un contexto seguro, controlado y adaptado a su perfil disfuncional. Dentro de este marco de rehabilitación la interacción visual paciente-entorno virtual se entiende como el mecanismo de comunicación principal, siendo además la atención visual un reflejo del estado cognitivo del paciente. El trabajo presentado en este artículo permite la integración de un sistema de eye-tracking con un entorno de neurorrehabilitación basado en vídeo interactivo. El objetivo último del sistema es la monitorización en tiempo real de la atención visual del usuario durante el proceso de neurorrehabilitación. Esta monitorización permite no sólo reproducir la ejecución de la actividad junto con el foco de la mirada, sino también detectar faltas de atención por parte del usuario, que permiten al vídeo interactivo reaccionar y adaptar la presentación de estímulos para ayudar a centrar su atención y así completar el objetivo de la actividad.

Herramienta de modelado disfuncional tridimensional basado en estudios de neuroimagen

El modelado disfuncional basado en estudios de neuroimagen mejora la comprensión de los cambios estructurales provocados ante la presencia de lesiones cerebrales. Actualmente, existen numerosas herramientas para el análisis y procesado de estudios de neuroimagen. Algunas de ellas, como el 3D Slicer, BrainVoyager y el FreeSurfer permiten la creación y navegación sobre modelos tridimensionales cerebrales sin alteraciones estructurales. Sin embargo, no se han detectado herramientas que permitan modelar tridimensionalmente lesiones a partir de estudios de neuroimagen, concretamente de estudios de resonancia magnética. El objetivo de este trabajo es el diseño de una metodología que permite la creación de este tipo de modelos y su visualización y navegación.

Identificación de Oportunidades de Mejora en Procesos de Neurorrehabilitación

El modelado de procesos es una técnica de gestión empresarial destinada a la mejora continua de los procesos de una organización, como base operativa y estructural de la misma. En el ámbito de la Neurorrehabilitación, crece el interés por los mapas de procesos como herramienta de comprensión, representación y análisis de los procesos clínicos. El presente trabajo se centra en la identificación de oportunidades de mejora de las actividades de rehabilitación, con el objetivo de definir nuevas estrategias de monitorización y automatización que permitan su evolución hacia el nuevo modelo de rehabilitación ubicua, personalizada y basada en la evidencia.

Segmentación automática de aneurismas aórticos abdominales a partir de imágenes de resonancia magnética basada en modelos activos de forma y modelos de textura

Se presenta un algoritmo semiautomático de segmentación de aneurismas aórticos abdominales (AAA) basado en modelos activos de forma (ASM) y modelos de textura. La información de textura viene dada por un conjunto de cuatro imágenes 3D de resonancia magnética (RM) compuestas por cortes axiales de la zona abdominal. En estas imágenes son visibles la luz aórtica, la pared aórtica y el trombo intraluminal (ILT). Dado el tamaño limitado del conjunto de imágenes de RM, se han implementado un ASM que capture las características específicas del conjunto de entrenamiento compuesto por 35 imágenes de tomografía axial computarizada (CTA), de modo que la variación de forma pueda ser adecuadamente caracterizada. La textura se caracteriza a partir de las imágenes de RM. Para la evaluación del algoritmo se ha llevado a cabo una validación cruzada dejando uno fuera sobre el conjunto de imágenes de RM.

Using a causal smoothing to improve the performance of an on-line neural network glucose prediction algorithm

This work evaluates a spline-based smoothing method applied to the output of a glucose predictor. Methods:Our on-line prediction algorithm is based on a neural network model (NNM). We trained/validated the NNM with a prediction horizon of 30 minutes using 39/54 profiles of patients monitored with the Guardian® Real-Time continuous glucose monitoring system The NNM output is smoothed by fitting a causal cubic spline. The assessment parameters are the error (RMSE), mean delay (MD) and the high-frequency noise (HFCrms). The HFCrms is the root-mean-square values of the high-frequency components isolated with a zero-delay non-causal filter. HFCrms is 2.90±1.37 (mg/dl) for the original profiles.

Automatic generation of real-time deformable parametric model of the aorta for a VR-based catheterism guidance system

The purpose of this work is twofold: first, to develop a process to automatically create parametric models of the aorta that can adapt to any possible intraoperative deformation of the vessel. Second, it intends to provide the tools needed to perform this deformation in real time, by means of a non-rigid registration method. This dynamically deformable model will later be used in a VR-based surgery guidance system for aortic catheterism procedures, showing the vessel changes in real time.