3D affine registration using teaching-learning based optimization

Jhala

2016

IEEE Systems Journal

In this paper, seven different metaheuristic optimization algorithms developed between 2005 and 2012 are applied to optimize the robot trajectory for a three-revolute (3R) robotic arm. These algorithms include the artificial bee colony (ABC) algorithm, biogeography-based optimization, gravitational search algorithm, cuckoo search algorithm (CS), firefly algorithm, bat algorithm, and teaching-learning-based optimization (TLBO). This work presents the optimization problem with the objective to plan a trajectory, which can minimize joint travelling time, joint travelling distance, and total joint Cartesian lengths, simultaneously. Nine different design variables are considered for the intermediate joint angles, joint velocities, the end-effector angle for the final configuration, the time from initial to intermediate configuration, and the time from intermediate to final configuration. Two different experiments are conducted to investigate the effect of different considered metaheuristics: 1) for the free workspace and 2) for the workspace with obstacle. All these algorithms are compared based on the algorithm ability to find the best solutions, mean solutions, standard deviation, computational effort, and the convergence of the solutions. Statistical investigation is also performed by using the Friedman rank test and the Holm-Sidak multiple comparison t-test to check the significance of the algorithm for its suitability to the robot trajectory optimization problems. The results show the significance of TLBO, ABC, and CS for the robot trajectory optimization problems. Index Terms-Artificial bee colony (ABC) algorithm, bat algorithm (BA), biogeography-based optimization (BBO), cuckoo search algorithm (CS), firefly algorithm (FA), gravitational search algorithm (GSA), obstacle existence, optimization, robot trajectory, teaching-learning-based optimization (TLBO).

Section: A Tlbomentioning

confidence: 99%

Comparative Study of Different Metaheuristics for the Trajectory Planning of a Robotic Arm

Jhala

2016

IEEE Systems Journal

“…As the images are coming from different persons, we choose to use a rigid registration, allowing correction of the different positions and orientations arising from the clinical exam. Since the natural size of the skulls is different from one person to another, we have avoided using affine registration [15], which risks distorting the estimation volume that will be used later as a parameter for identification. Thereby, we were able to build a new database consisting only of regions of interest, with the same size as the reference box.…”

Section: The Automatic Roi Extraction For the Ct-imagementioning

confidence: 99%

Automatic segmentation of the sphenoid sinus in CT-scans volume with DeepMedics 3D CNN architecture

Souadih

Belaid

Salem

2019

Mèd Technologies J.

Today, researchers are increasingly using manual, semi-automatic, and automatic segmentation techniques to delimit or extract organs from medical images. Deep learning algorithms are increasingly being used in the area of medical imaging analysis. In comparison to traditional methods, these algorithms are more efficient to obtain compact information, which considerably enhances the quality of medical image analysis system. In this paper, we present a new method to fully automatic segmentation of the sphenoid sinus using a 3D (convolutional neural network). The scarcity of medical data initially forced us through this study to use a 3D CNN model learned on a small data set. To make our method fully automatic, preprocessing and post processing are automated with extraction techniques and mathematical morphologies. The proposed tool is compared to a semi-automatic method and manual deductions performed by a specialist. Preliminary results from CT volumes appear very promising.

Advanced Intelligent Systems

Image Registration in Medical Robotics and Intelligent Systems: Fundamentals and Applications

Liu

Singh

Al’Aref

et al. 2019

Medical image registration, by transforming two or more sets of imaging data into one coordinate system, plays a central role in medical robotics and intelligent systems from diagnostics and surgical planning to real‐time guidance and postprocedural assessment. Recent advances in medical image registration have made a significant impact in orthopedic, neurological, cardiovascular, and oncological applications.The recent literature in medical image registration is reviewed, providing a discussion of their fundamentals and applications. Within each section, the registration techniques are introduced, classifying each method based on their working mechanisms, and discussing their benefits and limitations are discussed. Recently, machine learning has had an important impact on the field of image registration, yielding novel methods and unprecedented speed. The validation of registration methods, however, remains a challenge due to the lack of reliable ground truth. Medical image registration will continue to make significant impacts in the area of advanced medical imaging, as the fusion/combination of multimodal images and advanced visualization technology become more widespread.