Computer Vision in Medical Imaging and its Impact on the Rare Musculoskeletal Diseases / 罕见病研究

There are over 6000 rare diseases in the world, affecting more than 300 million people. Early and precise diagnosis of rare diseases has always been the goal in clinical medicine. Emerging computer vision technology now greatly enhance medicine and healthcare and shows the potential in assisting the diagnosis and treatment for rare diseases. The technology can be a useful tool for extracting disease-relevant patterns from medical imaging. However, the effectiveness of its application depends on the complexity of the medical cases. In this paper, we summarize the challenges and emerging solution for the application of computer vision in diagnosis, rehabilitation as well as management of rare musculoskeletal diseases.

Clasificación no supervisada de imágenes médicas y minería de datos. Algoritmo S3 vs K-medias / Unsupervised classification of medical images and data mining: S3 algorithm vs. k-means

Uno de los desafíos que los programadores tienen que enfrentar es la alta dimensión de grupos de datos. El proceso de reconocimiento de patrones en imagen y la minería de datos para los volúmenes grandes de información son ejemplos de ellos, optimizar la cantidad de veces que se recorre el conjunto de datos, disminuye el tiempo de procesamiento. Éste documento tiene el objetivo de caracterizar el algoritmo de tres pasos (S3), paralelo a K-medias, como una alternativa para afrontar la alta dimensión del conjunto de datos, en la clasificación no supervisada de imagen. Para el análisis de la concurrencia, se escoge, flujo de datos y el esquema instrucción única con datos múltiples. El resultado obtenido confirma que la concurrencia en ambos es posible, S3 no depende de la selección inicial de los representantes y puede ser el proceso de escogimiento de los primeros vectores centrales en K-medias. S3 es una alternativa a ser tenida en cuenta en la clasificación no supervisada de imágenes médicas y procesos de minería de datos(AU)

One of the challenges to be faced by programmers is the large dimensions of data groups. The process of pattern recognition in images and data mining for great volumes of information is an example. Optimizing the number of times that the set of data is run saves processing time. The purpose of the study was to characterize the three-step (S3) algorithm, parallel to k-means, as an alternative to cope with the large dimension of the data set in unsupervised image classification. Concurrence analysis is based on data flow and the single instruction multiple data scheme. The result obtained confirms that concurrence of both is possible. S3 does not depend on initial selection of representatives, and may be the process for selection of the first central vectors in k-means. S3 is an alternative to be considered in the unsupervised classification of medical images and data mining(AU)

Role of artificial intelligence in the diagnosis and treatment of gastrointestinal diseases / 中华胃肠外科杂志

The rapid development of computer technologies brings us great changes in daily life and work. Artificial intelligence is a branch of computer science, which is to allow computers to exercise activities that are normally confined to intelligent life. The broad sense of artificial intelligence includes machine learning and robots. This article mainly focuses on machine learning and related medical fields, and deep learning is an artificial neural network in machine learning. Convolutional neural network (CNN) is a type of deep neural network, that is developed on the basis of deep neural network, further imitating the structure of the visual cortex of the brain and the principle of visual activity. The current machine learning method used in medical big data analysis is mainly CNN. In the next few years, it is the developing trend that artificial intelligence as a conventional tool will enter the relevant departments of medical image interpretation. In addition, this article also shares the progress of the integration of artificial intelligence and biomedicine combined with actual cases, and mainly introduces the current status of CNN application research in pathological diagnosis, imaging diagnosis and endoscopic diagnosis for gastrointestinal diseases.

Progress on Biomechanical Research of Image-Based Subject-Specific OpenSim Lower Extremity Musculoskeletal Model / 医用生物力学

OpenSim musculoskeletal modelling has developed rapidly and been widely utilized due to its open-source. Apart from calculation of the basic kinematic and kinetic data, subject-specific OpenSim model could reveal information of neuromuscular control, muscle forces and geometry, and contact forces. Image-based model-ling of the neuromuscular control in pathological gait and ergonomic evaluation of the prostheses confirmed the reliability and feasibility, but limitations in time-consumption and foot-ankle modelling also existed. The subject-specific modelling of pathological gait could improve the accuracy and diversity of clinical biomechanics and medical engineering research. It could also reveal the pathological features, and provide scientific evidence to design specific and accurate protocols of motor function diagnosis and rehabilitation, health monitoring and evaluation, and ergonomic customization and assessment of devices, as well as future directions and implications in the research field.

Accuracy evaluation for 3D bone reconstruction based on medical 3D reconstruction software / 中华创伤骨科杂志

Objective@#To investigate the feasibility of an accuracy evaluation method for 3D reconstructed bone model based on 3D reconstruction software Arigin3D Pro.@*Methods@#Pig femurs were used as solid models which were scanned by CT and MRI respectively. The scan data were imported into software Arigin3D Pro for 3D model reconstruction by 3 operators with different reconstruction experience (≤1 year, 2 to 3 years, and ≥4 years, respectively). Each operator reconstructed the femurs 3 times and in each reconstruction measured the diameter of the femoral head, the length of the femur and the width of the knee joint at the distal end of the femur 3 times respectively using software Geomagic Wrap. The above parameters of the solid models were measured using a vernier caliper. The parameter values of reconstructed models and solid models were compared and the differences were analyzed.@*Results@#The measurements by Geomagic Wrap showed deviations between the CT and MRI reconstruction models and the solid models, and the maximum deviation percentages were 1.47% and 1.08%, respectively. The percentages of intra-operater difference ranged from 0.29% to 1.53%; the 3D models reconstructed by operators with different reconstruction experience were not identical.@*Conclusions@#It is a feasible accuracy evaluation method to compare key parameters between the 3D bone model reconstructed by software Arigin3D Pro and the real animal bone. The deviations of 3D reconstructed bone model based on CT and MRI images are acceptable. The accuracy of 3D bone construction is related to the difference in operators.

Accuracy evaluation for 3D bone reconstruction based on medical 3D reconstruction software / 中华创伤骨科杂志

Objective To investigate the feasibility of an accuracy evaluation method for 3D reconstructed bone model based on 3D reconstruction software Arigin3D Pro.Methods Pig femurs were used as solid models which were scanned by CT and MRI respectively.The scan data were imported into software Arigin3D Pro for 3D model reconstruction by 3 operators with different reconstruction experience (≤1 year,2 to 3 years,and ≥4 years,respectively).Each operator reconstructed the femurs 3 times and in each reconstruction measured the diameter of the femoral head,the length of the femur and the width of the knee joint at the distal end of the femur 3 times respectively using software Geomagic Wrap.The above parameters of the solid models were measured using a vernier caliper.The parameter values of reconstructed models and solid models were compared and the differences were analyzed.Results The measurements by Geomagic Wrap showed deviations between the CT and MRI reconstruction models and the solid models,and the maximum deviation percentages were 1.47％ and 1.08％,respectively.The percentages of intra-operater difference ranged from 0.29％ to 1.53％;the 3D models reconstructed by operators with different reconstruction experience were not identical.Conclusions It is a feasible accuracy evaluation method to compare key parameters between the 3D bone model reconstructed by software Arigin3D Pro and the real animal bone.The deviations of 3D reconstructed bone model based on CT and MR1 images are acceptable.The accuracy of 3D bone construction is related to the difference in operators.

Applications of generative adversarial networks in medical image processing / 生物医学工程学杂志

In recent years, researchers have introduced various methods in many domains into medical image processing so that its effectiveness and efficiency can be improved to some extent. The applications of generative adversarial networks (GAN) in medical image processing are evolving very fast. In this paper, the state of the art in this area has been reviewed. Firstly, the basic concepts of the GAN were introduced. And then, from the perspectives of the medical image denoising, detection, segmentation, synthesis, reconstruction and classification, the applications of the GAN were summarized. Finally, prospects for further research in this area were presented.

Medical image fusion of multimoding based on wavelet transformation / 中国医学装备

Objective: To fuse many non-homologous medical images based on wavelet transformation, and integrate and stand out complementation information, and strengthen image quality, and reduce redundancy so as to enhance the precision of clinically auxiliary diagnosis and treatment for medical image. Methods: Through studied and researched the relevant knowledge of wavelet theory in the application of image fusion to proposed a improved fusion algorithm by mutual combination for maximum low frequency energy and maximum high frequency variance. The fusion experiment of non-homologous standard medical image, including CT, MRI and multi-focus images, were implemented, and their data were compared. And then, the performance of image fusion was compared and analyzed under different fusion rule and different fusion method. Results: In the two kinds of contrastive analysis experiments, the fusion image which depended on algorithm included more abundantly effective information amount of source image, and the luminance of image was reasonable enhanced. Besides, the mean value, mutual information and information entropy of fusion image were optimal. Conclusion: The fusion image which comes from multimoding medical image algorithm has better visual effects and quantization indicator, and it has strengthener fused performance. Therefore, it can reflect the effectiveness of the method.

Una solución pacs cubana bajo software libre que sirve de plataforma a especializaciones médicas / A Cuban pacs solutions under free software as a framework for medical specializations

En nuestro país se lucha por la soberanía tecnológica y se impulsa la utilización del software libre. Por esto y con un mejor diseño de las soluciones implementadas por el Centro de Biofísica Médica, el imagis 2.0, como paquete de soluciones de un sistema de almacenamiento y transmisión de imágenes médicas, fue desarrollado bajo la plataforma Linux, haciendo uso de herramientas libres, convirtiéndolo en un sistema más eficiente, estable y robusto. Con este trabajo pretendemos primero realizar un breve recorrido por la historia de los sistemas de almacenamiento y transmisión de imágenes médicas en nuestro país, para luego adentrarnos en una solución implementada bajo plataforma libre que sirve de base a la implementación para especializaciones médicas, exponiendo su uso en los procesos radioterapéuticos(AU)

In our country we fight for technological sovereignty and the use of free software is promoted. For this and with better design solutions implemented by the Center of Medical Biophysics, the imagis 2.0, as a package of solutions of storage and transmission of medical images systems, was developed under the Linux platform, using free tools, making it more efficient, stable and robust. In this paper we first present a brief review of storage and transmission of medical images history in our country, for then get into a deployed under open platform solution that can be using as framework for the implementation for medical specializations, exposing it use in radiotherapeutic procedures(AU)

Modelación y simulación de la arteria aorta a partir de datos clínicos utilizando un modelo fraccional viscoelástico y el método del elemento finito / Modeling and simulation of the aorta from clinical data using a fractional viscoelastic model and finite element method

Los modelos y simulaciones de los efectos biomecánicos presentes en la arteria aorta, le proporcionan al especialista de la salud una herramienta computacional, que puede ser empleada en la prevención y el tratamiento de las enfermedades cardiovasculares. Es por esto que en la presente investigación se desarrolla un modelo matemático con la finalidad de implementarlo en simulaciones tridimensionales digitales que permitan analizar el comportamiento mecánico de arterias. Primero se describe la metodología utilizada en la construcción de la geometría de la arteria basada en imágenes provenientes de una tomografía axial computarizada, los ensayos experimentales necesarios para la obtención de los parámetros mecánicos requeridos por el modelo y por último su orden fraccional. Con lo que se obtiene una simulación mediante elementos finitos donde se identifican las zonas de mayor concentración de esfuerzos y el campo de desplazamientos. Para poder obtener estos resultados se empleó una formulación novedosa basada en modelos viscoelásticos de orden fraccional donde además se obtuvieron, a través del módulo complejo, los valores requeridos para la simulación.

The modeling and simulation of the biomechanical effects present in the aorta, give the health specialist a computational tool that can be used in the prevention and treatment of cardiovascular diseases. For that reason on this research a mathematical model was developed in order to implement digital dimensional simulations to analyze the mechanical behavior of arteries. First, its described the methodology used in the construction of the geometry of the artery based on images from a CT scan, next the necessary experimental tests to obtain mechanical parameters required by the model and finally his fractional order. Obtaining a finite element simulation where the areas of greatest stress concentration and the displacement field are identified. To obtain these results a novel formulation based on fractional order viscoelastic models was used and the values required for simulation were obtained through the complex modulus.

Análise dos softwares gratuitos para tomografia computadorizada de feixe cônico de interesse aos cirurgiões-dentistas / Analysis of free software for computed tomograp hy cone beam of interest to dentists

O objetivo deste artigo foi avaliar programas de imagens odontológicas disponíveis gratuitamente na internet como ferramentas para prática do cirurgião-dentista e que funcionam em computadores pessoais. Foi realizada uma pesquisa no Medline entre os anos de 2005 a 2014, com enfoque em software utilizado para visualizar imagens tomográficas na Odontologia. Para a pesquisa dos softwares foi realizado busca no site Google e em sites especializados por programas gratuitos disponíveis para windows. Foram utilizados as palavras chaves "free dicom viewer" e "dental software". De acordo com os critérios de inclusão e exclusão três softwares foram analisados. Conhecer a existência de softwares gratuitos e que dão um grande suporte para os profissionais é de extrema importância para a Odontologia atual.

The aim of this study was to review the dental imaging programs available for free on the Internet as tools to practice DDS and running on personal computers. A survey was conducted in Medline from 2005 to 2014, with focus on the software used to view tomographic images in dentistry. A research about the software was done at Google search site and specialized sites for free programs available for Windows. Key words were used "free dicom viewer" and "dental software." According to the criteria of inclusion and exclusion 3 softwares were analyzed. Knowing the existence of free softwares which are able to support professionals improve their jobs is a matter of great importance for the current dentistry.

Modelo mecanobiológico de una tibia humana para determinar su respuesta ante estímulos mecánicos externos / Mechanobiological model of a human tibia to determine its response to external mechanical stimuli

INTRODUCCIÓN: las nuevas herramientas informáticas cada vez más influyen en los avances científicos tecnológicos en la esfera de la salud. Una de las áreas de mayor implementación de estas nuevas herramientas de cómputo es sin dudas, la ortopedia, debido en lo fundamental, a la similitud que tiene el cuerpo humano y su estructura ósea con las propiedades de los materiales de ingeniería que permite modelar tejidos y órganos en correspondencia con teorías remodeladoras que predicen su posible comportamiento mecanobiológico. OBJETIVO: describir el comportamiento de una tibia humana ante estímulos externos de torsión. MÉTODOS: se utilizó el Método de los Elementos Finitos implementado en el software profesional ABAQUS/CAE para el cálculo de tensiones y deformaciones presentes en la tibia humana. RESULTADOS: se obtuvo el modelo específico del paciente con síndrome de torsión tibial, a partir de las imágenes médicas de un tomógrafo; las cuales fueron procesadas por ordenador, se utiliza el software de tratamiento de imágenes médicas MIMICS 10.01 para establecer una relación entre la escala de grises (unidades Hounsfield), la densidad ósea y el módulo de Young. CONCLUSIONES: se determinó que el valor correcto de momento torsor estático que debe ser aplicado a este paciente para iniciar el proceso remodelador es de 30 Nm.

INTRODUCTION: new information tools have a growing influence on technological and scientific advances in health care. Orthopedics is one of the areas where new computational tools have been largely implemented, mainly owing to the similarity of the human body and its bone structure to the properties of engineering materials, enabling the modeling of tissues and organs in accordance with remodeling theories predicting their potential mechanobiological behavior. OBJECTIVE: describe the response of a human tibia to external torsion stimuli. METHODS: estimation of tensions and deformities in the human tibia was based on the Finite Element Method incorporated into the professional software Abaqus / CAE. RESULTS: starting from medical tomographic images, a specific model was obtained for a patient with tibial torsion syndrome. The images were computer processed with the medical imaging treatment software MIMICS 10.01 to establish a relationship between the grayscale (Hounsfield units), bone density and Young's modulus. CONCLUSIONS: it was determined that 30 Nm is the correct static torsion moment value to be applied to this patient to start the remodeling process.

Visor de imágenes médicas digitales web / Web viewer for digital medical images

El desarrollo de las tecnologías ha facilitado la creación de soluciones médicas que agilizan las tareas de los especialistas. A través del uso de este tipo de sistemas, consultorios médicos, clínicas, centros de imagen y departamentos de diagnóstico por imágenes pueden tener a gran velocidad servicios de Internet a su alcance y capacidades de archivo que alguna vez fueron solo para los grandes hospitales. La Universidad de las Ciencias Informáticas ha desarrollado la solución alas PACS-RIS, pero este sistema debido a sus características específicas no causa el impacto deseado debido a los grandes costos de tiempo de instalación y configuración, su uso limitado a plataformas propietarias, la necesidad de equipo de altas prestaciones, entre otras causas. Se presenta la propuesta de solución de software desarrollada para la visualización de imágenes médicas. El sistema está diseñado para ofrecer al personal médico, servicios en línea que faciliten la ejecución de sus tareas. Facilita a los especialistas una gama de herramientas básicas para la visualización y procesamiento de imágenes médicas y creación de los reportes de estudios realizados a pacientes. La aplicación fue desarrollada sobre la plataforma .NET 4.0 con lenguajes de programación C#, JavaScript, HTML5, CSS3. Se utilizaron además las librerías jQuery, Knockout y además de MVVM como patrón arquitectónico para la capa de presentación. El sistema se integra a la solución alas PACS-RIS y está diseñada para que finalmente sea utilizada en dispositivos móviles como celulares y Tablet PC(AU)

The actual development of technologies increases the creation of medical solutions that improve the specialist's tasks. Through the usage of this kind of systems, medical institutions, clinics, image centers and image diagnostic departments can use huge speed internet services and storage capacities that once could only have big hospitals. The Informatics Science University has develop the alas PACS-RIS solution, but this system, due to its specific characteristics, didn't cause the wished impact, due to the big time costs of installation and configuration, the limited use under proprietary platforms, the use in high end computers, among others. This paper has the fundamental objective of develop a web application that allows the medical images visualization and management of image studies reports. Presents a software solution developed in .NET 4.0 platform, with programming languages such as: C#, JavaScript, HTML5, CSS3. Also were used the libraries, jQuery, Knockout, and MVVM as architectural pattern for presentation layer. This solution is integrated with alas PACS-RIS system and is designed for the usage in mobile platforms such as: Mobiles and Tablets PC(AU)

Texture analysis of masses in digitized mammograms using Gleason and Menhinick diversity indexes

INTRODUCTION: Breast cancer is the second most common type of cancer in the world, being more common among women and representing 22% of all new cancer cases every year. The sooner it is diagnosed, the better the chances of a successful treatment are. Mammography is one way to detect non-palpable tumors that cause breast cancer. However, it is known that the sensitivity of this exam can vary considerably due to factors such as the specialist's experience, the patient's age and the quality of the images obtained in the exam. The use of computational techniques involving artificial intelligence and image processing has contributed more and more to support the specialists in obtaining a more precise diagnosis. METHODS: This paper proposes a methodology that exclusively uses texture analysis to describe features of masses in digitized mammograms. To increase the efficiency of texture feature extraction, the diversity index's capability to detect patterns of species co-occurrence is used. For this purpose, the Gleason and Menhinick indexes are used. Finally, the extracted texture is classified using the Support Vector Machine, looking to differentiate the malignant masses from the benign. RESULTS: The best result was obtained using the Gleason index, with 86.66% accuracy, 90% sensitivity, 83.33% specificity and an area under the ROC Curve (Az) of 0.86. CONCLUSION: Both indexes showed statistically similar performance; however, the Gleason index was slightly superior.

Visualizadores de imagens médicas gratuitos: é possível trabalhar apenas com eles? / Freeware medical image viewers: can we rely only on them?

OBJETIVO: Pesquisar visualizadores de imagens médicas gratuitos disponíveis na internet capazes de funcionar como cliente PACS (picture archiving and communication system) e avaliar suas principais funções e a viabilidade do uso em computadores pessoais. MATERIAIS E MÉTODOS: Foi feita pesquisa, no Google e em sites especializados, por programas gratuitos disponíveis para o Windows. Foram encontrados cerca de 70, sendo 11 capazes de funcionar como cliente PACS, e selecionados seis destes para análise: ClearCanvas Workstation, KPACS, Onis, Synedra View Personal, Mito e Tudor DicomViewer. Com base nas necessidades dos autores, 16 funções foram avaliadas. RESULTADOS: Dos seis programas avaliados, dois possuem 10 das 16 funções avaliadas e um possui apenas duas. Três realizam MPR (reconstrução multiplanar), um realiza MIP (reconstrução por projeção de intensidade máxima), dois realizam VR (renderizações volumétricas), dois funcionam como servidor PACS, dois geram CDs, um realiza fusão de imagens, três permitem utilizar múltiplos monitores e apenas um não é compatível com Windows 7. CONCLUSÃO: Diversos programas gratuitos estão disponíveis e não existe nenhum completo. Cabe ao usuário analisar e selecionar o programa que melhor se enquadra nas suas necessidades, porém, os programas Onis, Synedra e ClearCanvas se destacam, cada um com suas peculiaridades. É totalmente viável o uso de programas gratuitos para o dia-a-dia do radiologista.

OBJECTIVE: To search in the internet for freeware medical image viewers capable of running as a PACS (picture archiving and communication system) client, and to evaluate their main functions as well as the feasibility of their use in personal computers. MATERIALS AND METHODS: The Google search engine and specialized sites were utilized in the search for freeware softwares for Windows. The authors have found about 70 and among them 11 were able to run as PACS clients. Six were selected for analysis: ClearCanvas Workstation, KPACS, Onis, Synedra View Personal, Mito and Tudor DicomViewer. Sixteen functions selected according to the authors' needs were evaluated. RESULTS: Among the six applications, two presented 10 of the 16 functions, and one of them presented only two. Three perform MPR (multiplanar reconstruction), one performs MIP (maximum intensity projection), two perform VR (volume rendering), two can run as a PACS server, two can create CDs, one performs images fusion, three allow the use of multiple monitors and only one is not compatible with Windows 7. CONCLUSION: Although several freeware applications are available, no one of them is complete. It is up to the users to analyze and select the software that best suits their needs. However, Onis, Synedra and ClearCanvas stand out because of their own peculiarities. The use of freeware image viewers is entirely feasible in the radiologists' daily routine.

Technique of 3D virtual surgery based on medical images / 中国医学影像技术

Objective To research the key technique in the virtual surgery, and to build a 3D virtual surgical system for medical imaging. Methods In VC++ platform, the volume data were segmented by using a 3D segmentation algorithm. The 3D reconstruction and virtual cutting were achieved with the help of VTK, and the performance of the varied cutting methods was analyzed. Results The interactive 3D reconstruction, virtual scalpel cutting and virtual surgical planning of the medical CT/MR images were realized. Conclusion This system offers great help for doctor to simulate the operative process, to observe the 3D models of human organs and to carry out the computer-aided surgery.

Implementação e avaliação de um sistema de gerenciamento de imagens médicas com suporte à recuperação baseada em conteúdo / Implementation and evaluation of a medical image management system with content-based retrieval support

OBJETIVO: Neste artigo são descritas a implementação e avaliação de um sistema de gerenciamento de imagens médicas com suporte à recuperação baseada em conteúdo (PACS-CBIR), integrando módulos voltados para a aquisição, armazenamento e distribuição de imagens, e a recuperação de informação textual por palavras-chave e de imagens por similaridade. MATERIAIS E MÉTODOS: O sistema foi implementado com tecnologias para Internet, utilizando-se programas livres, plataforma Linux e linguagem de programação C++, PHP e Java. Há um módulo de gerenciamento de imagens compatível com o padrão DICOM e outros dois módulos de busca, um baseado em informações textuais e outro na similaridade de atributos de textura de imagens. RESULTADOS: Os resultados obtidos indicaram que as imagens são gerenciadas e armazenadas corretamente e que o tempo de retorno das imagens, sempre menor do que 15 segundos, foi considerado bom pelos usuários. As avaliações da recuperação por similaridade demonstraram que o extrator escolhido possibilitou a separação das imagens por região anatômica. CONCLUSÃO: Com os resultados obtidos pode-se concluir que é viável a implementação de um PACS-CBIR. O sistema apresentou-se compatível com as funcionalidades do DICOM e integrável ao sistema de informação local. A funcionalidade de recuperação de imagens similares pode ser melhorada com a inclusão de outros descritores.

OBJECTIVE: The present paper describes the implementation and evaluation of a medical images management system with content-based retrieval support (PACS-CBIR) integrating modules focused on images acquisition, storage and distribution, and text retrieval by keyword and images retrieval by similarity. MATERIALS AND METHODS: Internet-compatible technologies were utilized for the system implementation with freeware, and C++, PHP and Java languages on a Linux platform. There is a DICOM-compatible image management module and two query modules, one of them based on text and the other on similarity of image texture attributes. RESULTS: Results demonstrate an appropriate images management and storage, and that the images retrieval time, always < 15 sec, was found to be good by users. The evaluation of retrieval by similarity has demonstrated that the selected images extractor allowed the sorting of images according to anatomical areas. CONCLUSION: Based on these results, one can conclude that the PACS-CBIR implementation is feasible. The system has demonstrated to be DICOM-compatible, and that it can be integrated with the local information system. The similar images retrieval functionality can be enhanced by the introduction of further descriptors.

Multimodal Medical Image Registration Based on Gradient Vector Flow and Particle Swarm Optimization / 航天医学与医学工程

Objective To study the method based on gradient vector flow (GVF) and particle swarm optimization (PSO) for realizing multimodal medical image registration and improving its accuracy. Methods In view of three major components of image registration, i.e. the feature space, the similarity metric and the search strategy, a novel method was proposed with three improvements. Firstly, the GVF field was employed as the feature space. Then three similarity metrics were proposed based on GVF field. Finally, an improved PSO combined with crossover mechanism of genetic algorithm was utilized to search for the optimal transformation of two images. Results With 54 times of experiments on both simulated and real medical images, it was demonstrated that this method accurately registered the multimodal medical images to be superior to the method based on PSO of pixels, and the Walsh transform method. Conclusion The method based on GVF and PSO is effective for multimodal medical image registration.

A Non-rigid Medical Image Registration Algorithm Based on Modally Controlled Free Form Deformation / 航天医学与医学工程

Objective To register pre-operative MRI/CT images with intra-operative ultrasound images based on vessels visible in both of the modalities. Method A non-rigid registration method of multimodal medical images based on Free Form Deformation(FFD) was proposed. When the images were aligned, the centerline points of the vessels in one image aligned with the intensity ridge points in the other image. Rigid transformation was adopted in global registration while local deformation was described by a Free Form Deformation based on a modally controlled B-spline. The method applied an optimization strategy combining the genetic algorithm with the conjugated gradients algorithm to minimize the objective function. Result Two experiments were designed on phantom and clinical data to evaluate the method. The results demonstrated that the registration method was consistent accurate. The average standard deviation of the final transformation parameters was sub-voxel, sub-millimeter, and within 0.010 radians. Conclusion The results show that the method has good registration accuracy and convergence rate. It can be applied efficiently in the ultrasound-image-guided surgery system.