Fernando Arámbula Cosío

Automatic initialization of an active shape model of the prostate

In this work is reported a new method for automatic segmentation of the boundary of the prostate, in transurethral ultrasound images. The scheme is based on a robust automatic initialization of an active shape model (ASM) of the prostate, which is subsequently fitted to the boundary of the gland. The initialization of the ASM is based on pixel classification to estimate the prostate region in an ultrasound image, followed by automatic adjustment - using a multipopulation genetic algorithm (MPGA) - of the initial pose of the ASM to the binary image produced by the classifier. The initial pose is next adjusted to the gray level ultrasound image, using the MPGA. After automatic initialization, the ASM is adjusted to the gray level ultrasound image to produce the final prostate contour. The method provides fast and robust segmentation of the prostate boundary. Validation results on 22 ultrasound images are reported with 1.74 mm of mean boundary error and an estimated processing time of 66 per image. Our automatic initialization method can be applied with the ASMs of different organs in various imaging modalities.

Local Autonomous Robot Navigation Using Potential Fields

The potential fields method for autonomous robot navigation consists essentially in the assignment of an attractive potential to the goal point and a repulsive potential to each of the obstacles in the environment. Several implementations of potential fields for autonomous robot navigation have been reported. The most simple implementation considers a known environment where fixed potentials can be assigned to the goal and the obstacles. When the obstacles are unknown the potential fields have to be adapted as the robot advances, and detects new obstacles. The implementation of the potential fields method with one attraction potential assigned to the goal point and fixed repulsion points assigned to the obstacles, has the important limitation that for some obstacle configurations it may not be possible to produce appropriate resultant forces to avoid the obstacles. Recently the use of several adjustable attraction points, and the progressive insertion of repulsion points as obstacles are detected online, have proved to be a viable method to avoid large obstacles using potential fields in environments with unknown obstacles. In this chapter we present the main characteristics of the different approaches to implement local robot navigation algorithms using potential fields for known and partially known environments. Different strategies to escape from local minima, that occur when the attraction and repulsion forces cancel each other, are also considered.

Automatic segmentation of the fetal cerebellum on ultrasound volumes, using a 3D statistical shape model

Previous work has shown that the segmentation of anatomical structures on 3D ultrasound data sets provides an important tool for the assessment of the fetal health. In this work, we present an algorithm based on a 3D statistical shape model to segment the fetal cerebellum on 3D ultrasound volumes. This model is adjusted using an ad hoc objective function which is in turn optimized using the Nelder–Mead simplex algorithm. Our algorithm was tested on ultrasound volumes of the fetal brain taken from 20 pregnant women, between 18 and 24 gestational weeks. An intraclass correlation coefficient of 0.8528 and a mean Dice coefficient of 0.8 between cerebellar volumes measured using manual techniques and the volumes calculated using our algorithm were obtained. As far as we know, this is the first effort to automatically segment fetal intracranial structures on 3D ultrasound data.

Deformable model of the prostate for TURP surgery simulation

Abstract During a prostatectomy, a surgeon removes the inner prostate tissue that obstructs the urinary flow. The standard procedure, called transurethral resection of the prostate (TURP), is a minimally invasive surgery in which a resectoscope is introduced through the urethra of the patient to remove the obstructing tissue. In this paper, we present a three-dimensional (3D) computer model of the prostate for TURP simulation. The prostate model is designed to be the basis of a computer simulator for TURP training. The model was built from a set of ultrasound images with a technique that constructs a 3D volumetric mesh of the prostate shape, which is able to closely reproduce tissue resections as they are performed during real TURP procedures. A mass-spring method is used to model tissue deformation due to surgical tool interaction. The model simulates, in real-time: resections; tissue deformations; the cavity produced by the user as the surgical procedure progresses; and the corresponding reduction of the prostate volume.

Automatic segmentation of the cerebellum of fetuses on 3D ultrasound images, using a 3D Point Distribution Model

Analysis of fetal biometric parameters on ultrasound images is widely performed and it is essential to estimate the gestational age, as well as the fetal growth pattern. The use of three dimensional ultrasound (3D US) is preferred over other tomographic modalities such as CT or MRI, due to its inherent safety and availability. However, the image quality of 3D US is not as good as MRI and therefore there is little work on the automatic segmentation of anatomic structures in 3D US of fetal brains. In this work we present preliminary results of the development of a 3D Point Distribution Model (PDM), for automatic segmentation, of the cerebellum in 3D US of the fetal brain. The model is adjusted to a fetal 3D ultrasound, using a genetic algorithm which optimizes a model fitting function. Preliminary results show that the approach reported is able to automatically segment the cerebellum in 3D ultrasounds of fetal brains.

Autonomous robot navigation based on the evolutionary multi-objective optimization of potential fields

This article presents the application of a new multi-objective evolutionary algorithm called RankMOEA to determine the optimal parameters of an artificial potential field for autonomous navigation of a mobile robot. Autonomous robot navigation is posed as a multi-objective optimization problem with three objectives: minimization of the distance to the goal, maximization of the distance between the robot and the nearest obstacle, and maximization of the distance travelled on each field configuration. Two decision makers were implemented using objective reduction and discrimination in performance trade-off. The performance of RankMOEA is compared with NSGA-II and SPEA2, including both decision makers. Simulation experiments using three different obstacle configurations and 10 different routes were performed using the proposed methodology. RankMOEA clearly outperformed NSGA-II and SPEA2. The robustness of this approach was evaluated with the simulation of different sensor masks and sensor noise. The scheme reported was also combined with the wavefront-propagation algorithm for global path planning.

Computer assisted biopsy of breast tumors

In this paper we report our preliminary results of the development of a computer assisted system for breast biopsy. The system is based on tracked ultrasound images of the breast. A three dimensional ultrasound volume is constructed from a set of tracked B-scan images acquired with a calibrated probe. The system has been designed to assist a radiologist during breast biopsy, and also as a training system for radiology residents. A semiautomatic classification algorithm was implemented to assist the user with the annotation of the tumor on an ultrasound volume. We report the development of the system prototype, tested on a physical phantom of a breast with a tumor, made of polivinil alcohol.

Resection Simulation with Local Tissue Deformations for Computer Assisted Surgery of the Prostate

We present a three-dimensional anatomical and deformable model of the prostate for prostatectomy simulation. The model was build from a set of ultrasound images with the prostate contour, automatically annotated, with a technique based on a genetic algorithm and principal components analysis. The model simulates resection operations and local tissue deformations during virtual resectoscope interaction. A mass-spring method is used to model tissue deformation due to surgical tool interaction. Through 3D mesh modification and updating of the nodes of the mesh, the model is able to show in real time, resections and local tissue deformations produced by the user. The anatomical model is designed to assist the surgeon (in conjunction with an optical tracker) to perform Transurethral Resections of the Prostate (TURP) by showing in real time the position of the resectoscope inside the body of the patient and the deformation of the prostate shape during resection.

Motion estimation and segmentation in CT cardiac images using the Hermite transform and active shape models

Considering the importance of studying the movement of certain cardiac structures such as left ventricle and myocardial wall for better medical diagnosis, we propose a method for motion estimation and image segmentation in sequential Computed Tomography images. Two main tasks are tackled. The first one consists of a method to estimate the hearts motion based on a bio-inspired image representation model. Our proposal for optical flow estimation incorporates image structure information extracted from the steered Hermite transform coefficients that is later used as local motion constraints in a differential estimation approach. The second task deals with cardiac structure segmentation in time series of cardiac images based on deformable models. The goal is to extend active shape models (ASM) of 2D objects to the problem of 3D (2D + time) cardiac CT image modeling. The segmentation is achieved by constructing a point distribution model (PDM) that encodes the spatio-temporal variability of a training set. Combination of both motion estimation and image segmentation allows isolating motion in cardiac structures of medical interest such as ventricle walls.

Improved collision detection algorithm for soft tissue deformable models

The virtual 3D models of human organs in surgery simulation must reproduce the complex viscoelastic deformable behavior of living soft tissue during interaction with virtual surgical tools. Collision detection is a crucial task for modelling and real-time simulation of surgical interactions. Few techniques for collision detection between moving bodies that involve tissue cutting and deformations have been reported. In this paper, we present an improved Quinlan algorithm for collision detection suitable for real-time tissue cutting and deformation simulation. We also present an example of the use of the algorithm in a prostate surgery simulation system.