Biomedical subjects

Luciano Silva

Publications and source records attributed to Luciano Silva.

3 recordsLinked to original sources

PubMed2005

Using computer vision to help the determination of the gestational age of newborns.

RATIONALE AND OBJECTIVES: This report presents a computational approach to help the gestational age determination of newborns. Gestational age knowledge is fundamental to guide postnatal treatment and increase survival chances of newborns. However, current methods are invasive and do not generate precise results, mainly because they were developed based on nonpremature populations. MATERIALS AND METHODS: We developed an original and noninvasive method to help determination of gestational age based on information supplied by plantar surface images. These images present many details and patterns, but, to date, have not received attention from the image-processing community. We provide a computational tool with suitable facilities to allow the image analysis, either automatically or user driven. This image-processing tool is presented here. RESULTS: The image-processing tool was developed on a user-driven basis. However, as a quantitative experiment, 186 images were processed without user intervention to observe tool behavior in performing different tasks. Although preliminary, experimental results confirm the relationship between plantar surface features and gestational age. CONCLUSION: A prototype of the FootScanAge System is being used and evaluated by experts in neonatology. By means of digital processing of plantar surface images, some characteristics may be shown. Some hypotheses regarding the method have already been confirmed. Also, we show that some well-known image-processing techniques, if appropriately adapted, lead to suitable results when applied to plantar surface images.

Dermatoglyphics↗

PubMed2005

Precision range image registration using a robust surface interpenetration measure and enhanced genetic algorithms.

This paper addresses the range image registration problem for views having low overlap and which may include substantial noise. The current state of the art in range image registration is best represented by the well-known iterative closest point (ICP) algorithm and numerous variations on it. Although this method is effective in many domains, it nevertheless suffers from two key limitations: It requires prealignment of the range surfaces to a reasonable starting point and it is not robust to outliers arising either from noise or low surface overlap. This paper proposes a new approach that avoids these problems. To that end, there are two key, novel contributions in this work: a new, hybrid genetic algorithm (GA) technique, including hillclimbing and parallel-migration, combined with a new, robust evaluation metric based on surface interpenetration. Up to now, interpenetration has been evaluated only qualitatively; we define the first quantitative measure for it. Because they search in a space of transformations, GAs are capable of registering surfaces even when there is low overlap between them and without need for prealignment. The novel GA search algorithm we present offers much faster convergence than prior GA methods, while the new robust evaluation metric ensures more precise alignments, even in the presence of significant noise, than mean squared error or other well-known robust cost functions. The paper presents thorough experimental results to show the improvements realized by these two contributions.

Algorithms↗

PubMed2004

Range image segmentation into planar and quadric surfaces using an improved robust estimator and genetic algorithm.

This paper presents a novel range image segmentation method employing an improved robust estimator to iteratively detect and extract distinct planar and quadric surfaces. Our robust estimator extends M-estimator Sample Consensus/Random Sample Consensus (MSAC/RANSAC) to use local surface orientation information, enhancing the accuracy of inlier/outlier classification when processing noisy range data describing multiple structures. An efficient approximation to the true geometric distance between a point and a quadric surface also contributes to effectively reject weak surface hypotheses and avoid the extraction of false surface components. Additionally, a genetic algorithm was specifically designed to accelerate the optimization process of surface extraction, while avoiding premature convergence. We present thorough experimental results with quantitative evaluation against ground truth. The segmentation algorithm was applied to three real range image databases and competes favorably against eleven other segmenters using the most popular evaluation framework in the literature. Our approach lends itself naturally to parallel implementation and application in real-time tasks. The method fits well, into several of today's applications in man-made environments, such as target detection and autonomous navigation, for which obstacle detection, but not description or reconstruction, is required. It can also be extended to process point clouds resulting from range image registration.

Algorithms↗