RESUMEN

Abstract This paper aims to introduce an innovative approach to semantic segmentation by leveraging a convolutional neural network (CNN) for predicting the shape and pose parameters of the left ventricle (LV). Our approach involves a modified U-Net architecture with a regression layer as the final stage, as opposed to the traditional classification layer. This modification allows us to predict all the shape and pose parameters of a statistical shape model, including rotation, translation, scale, and deformation. The adapted U-Net is trained using data from a point distribution model (PDM) of the LV. The experimental results demonstrate a mean Dice coefficient of 0.82 on good quality images, and 0.66 including mean and low-quality images. Our approach successfully overcomes a common issue encountered in CNN-based semantic segmentation. Unlike the inaccurate pixel classification that often leads to unwanted blobs, our CNN generates statistically valid shapes. These shapes hold significant potential in initializing other methods, such as active shape models (ASMs). Our novel CNN-based approach provides a novel solution for semantic segmentation, offering shapes and pose parameters that can enhance the accuracy and reliability of subsequent medical image analysis methods.

Resumen Este artículo tiene como objetivo introducir un enfoque innovador para la segmentación semántica utilizando una red neuronal convolucional (CNN) para predecir los parámetros de forma y posición del ventrículo izquierdo (VI). Nuestro enfoque implica una arquitectura U-Net modificada con una capa de regresión como etapa final, en contraposición a la capa de clasificación tradicional. Esta modificación nos permite predecir todos los parámetros de un modelo estadístico de formas que incluyen rotación, traslación, escala y deformación. La red convolucional se entrena utilizando datos de un modelo de distribución de puntos (PDM) del VI. Los resultados experimentales muestran un coeficiente Dice promedio de 0.82 para imágenes de buena calidad y de 0.66 cuando se incluyen imágenes de calidad media y baja. Nuestro enfoque supera con éxito un problema común en la segmentación semántica basada en CNNs. A diferencia de la clasificación inexacta de píxeles que a menudo conduce a elementos no deseados (blobs), nuestra CNN genera formas estadísticamente válidas. Estas formas tienen un gran potencial para inicializar otros métodos, como los modelos de forma activa (ASMs). En resumen, nuestro enfoque basado en CNN proporciona una solución innovadora para la segmentación semántica, ofreciendo formas y parámetros de posición que pueden mejorar la precisión y confiabilidad de otros métodos de análisis del VI.

RESUMEN

Quantification of brain growth is crucial for the assessment of fetal well being, for which ultrasound (US) images are the chosen clinical modality. However, they present artefacts, such as acoustic occlusion, especially after the 18th gestational week, when cranial calcification appears. Fetal US volume registration is useful in one or all of the following cases: to monitor the evolution of fetometry indicators, to segment different structures using a fetal brain atlas, and to align and combine multiple fetal brain acquisitions. This paper presents a new approach for automatic registration of real 3D US fetal brain volumes, volumes that contain a considerable degree of occlusion artefacts, noise, and missing data. To achieve this, a novel variant of the coherent point drift method is proposed. This work employs supervised learning to segment and conform a point cloud automatically and to estimate their subsequent weight factors. These factors are obtained by a random forest-based classification and are used to appropriately assign nonuniform membership probability values of a Gaussian mixture model. These characteristics allow for the automatic registration of 3D US fetal brain volumes with occlusions and multiplicative noise, without needing an initial point cloud. Compared to other intensity and geometry-based algorithms, the proposed method achieves an error reduction of 7.4% to 60.7%, with a target registration error of only 6.38 ± 3.24 mm. This makes the herein proposed approach highly suitable for 3D automatic registration of fetal head US volumes, an approach which can be useful to monitor fetal growth, segment several brain structures, or even compound multiple acquisitions taken from different projections.

Asunto(s)

RESUMEN

Analysis of cardiac images is a fundamental task to diagnose heart problems. Left ventricle (LV) is one of the most important heart structures used for cardiac evaluation. In this work, we propose a novel 3D hierarchical multiscale segmentation method based on a local active contour (AC) model and the Hermite transform (HT) for LV analysis in cardiac magnetic resonance (MR) and computed tomography (CT) volumes in short axis view. Features such as directional edges, texture, and intensities are analyzed using the multiscale HT space. A local AC model is configured using the HT coefficients and geometrical constraints. The endocardial and epicardial boundaries are used for evaluation. Segmentation of the endocardium is controlled using elliptical shape constraints. The final endocardial shape is used to define the geometrical constraints for segmentation of the epicardium. We follow the assumption that epicardial and endocardial shapes are similar in volumes with short axis view. An initialization scheme based on a fuzzy C-means algorithm and mathematical morphology was designed. The algorithm performance was evaluated using cardiac MR and CT volumes in short axis view demonstrating the feasibility of the proposed method.

Asunto(s)

RESUMEN

A new method to address the problem of shadowing in fetal brain ultrasound volumes is presented. The proposed approach is based on the spatial composition of multiple 3-D fetal head projections using the weighted Euclidean norm as an operator. A support vector machine, which is trained with optimal textural features, was used to assign weighting according to the posterior probabilities of brain tissue and shadows. Both phantom and real fetal head ultrasound volumes were compounded using previously reported operators and compared with the proposed composition method to validate it. The quantitative evaluations revealed increases in signal-to-noise ratio ≤35% and in contrast-to-noise ratio ≤135% using real data. Qualitative comparisons made by obstetricians indicated that this novel method adequately recovers brain tissue and improves the visibility of the main cerebral structures. This may prove useful both for fetal monitoring and in the diagnosis of brain defects. Overall this new approach outperforms spatial composition methods previously reported.

Asunto(s)

RESUMEN

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.

Asunto(s)

RESUMEN

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.

Asunto(s)

RESUMEN

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.

Asunto(s)