This paper describes a methodology that extracts morphological features from histological breast cancer images stained for Hematoxilyn and Eosin (H&E). Cellularity was estimated and the correlation between features and the residual tumour size cellularity after a Neo-Adjuvant treatment (NAT) was examined. Images from whole slide imaging (WSI) were processed automatically with traditional computer vision methods to extract twenty two morphological parameters from the nuclei, epithelial region and the global image. The methodology was applied to a set of images from breast cancer under NAT. The data came from the BreastPathQ Cancer Cellularity Challenge 2019, and consisted of 2579 patches of 255×255 pixels of H&E histopatological samples from NAT treatment patients. The methodology automatically implements colour separation, segmentation and morphological analysis using traditional algorithms (K-means grouping, watershed segmentation, Otsu's binarisation). Linear regression methods were applied to determine strongest correlation between the parameters and the cancer cellularity. The morphological parameters showed correlation with the residual tumour cancer cellularity. The strongest correlations corresponded to the stroma concentration value (r = −0.9786) and value from HSV image colour space (r = −0.9728), both from a global image parameters.
This paper describes a methodology that extracts key morphological features from histological breast cancer images in order to automatically assess Tumour Cellularity (TC) in Neo-Adjuvant treatment (NAT) patients. The response to NAT gives information on therapy efficacy and it is measured by the residual cancer burden index, which is composed of two metrics: TC and the assessment of lymph nodes. The data consist of whole slide images (WSIs) of breast tissue stained with Hematoxylin and Eosin (H&E) released in the 2019 SPIE Breast Challenge. The methodology proposed is based on traditional computer vision methods (K-means, watershed segmentation, Otsu’s binarisation, and morphological operations), implementing colour separation, segmentation, and feature extraction. Correlation between morphological features and the residual TC after a NAT treatment was examined. Linear regression and statistical methods were used and twenty-two key morphological parameters from the nuclei, epithelial region, and the full image were extracted. Subsequently, an automated TC assessment that was based on Machine Learning (ML) algorithms was implemented and trained with only selected key parameters. The methodology was validated with the score assigned by two pathologists through the intra-class correlation coefficient (ICC). The selection of key morphological parameters improved the results reported over other ML methodologies and it was very close to deep learning methodologies. These results are encouraging, as a traditionally-trained ML algorithm can be useful when limited training data are available preventing the use of deep learning approaches.
Resum en El uso de los sistemas de captura de imágenes para las computadoras es cada vez más común. Por otro lado, el procesamiento y análisis de video digital puede tener aplicaciones interesantes en el área de la comunicación hombre-máquina. Pensando en este desarrollo científico-tecnológico, este trabajo aborda el problema de la comunicación "amigable" entre la computadora y algunos seres humanos minusválidos que requieren de facilidades, por lo que se desarrolló un lenguaje gestual con el que la persona puede indicar a la máquina algunas acciones simples a seguir por medio de sus manos. Se definió un alfabeto muy limitado de gestos que la máquina puede interpretar y se pensó también en no utilizar guantes instrumentados, ya que son costosos y estorbosos. Por lo tanto, el número de gestos tiene que ser reducido y deben mostrarse relativamente burdos; dada la falta de precisión que se tendrá durante su captura con una cámara digital común y corriente. Una vez realizados y capturados los gestos, se procede a su segmentación con el propósito de aislarlos del resto de la escena captada por la cámara y poder efectuar su interpretación de una manera más sencilla. Los gestos están provistos de movimiento, de tal modo que para aislarlos se puede explotar esta información recurriendo a un algoritmo de segmentación basada en el movimiento. Logrando el aislamiento se lleva a cabo su interpretación de una manera más sencilla y rápida al obtener una máscara binaria del gesto, la cual se proyecta enseguida en dos ejes ortogonales x y y. Los perfiles del gesto obtenidos por la proyección enxyy, se utilizan como una especie de apuntadores; de esta manera se puede buscar en una base de datos previamente estructurada y guardada en memoria. Los resultados obtenidos hasta ahora son satisfactorios, aunque el número de gestos definidos es muy reducido y bastante distintos unos de otros, lo que facilita su interpretación. Por lo anterior este trabajo consiste en ampliar el tam año del alfabeto y aumentar la complejidad de los gestos para tener una aplicación más realista. Descriptores: interfaz gestual, interacción hombre-máquina, interpretación automática de gestos, segmentación de objetos no rígidos, análisis de movimiento de objetos no rígidos, segmentación de objetos en movimiento.
Abstract.A method for graduated scale inspection using computer vision is proposed. We deal mainly with the lens distortion problem in the image acquire device due to its influence in the uncertainty of the graduated scale inspection process. This paper presents an algorithm for image correction by means of camera calibration and distortion compensation. The camera calibration method provides the ideal undistorted coordinates of the system using as input distorted images of a 2D calibration pattern. The distortion compensation stage is implemented using the ideal undistorted coordinates as an unwarped mesh. Then distortion compensation can be applied to any image acquired with the system, improving the inspection procedure. Test results using real data are presented. Also, we describe the image feature extraction approach used in order to automate the process.
This paper describes a method for residual tumour cellularity (TC) estimation in Neoadjuvant treatment (NAT) of advanced breast cancer. This is determined manually by visual inspection by a radiologist, then an automated computation will contribute to reduce time workload and increase precision and accuracy. TC is estimated as the ratio of tumour area by total image area estimated after the NAT. The method proposed computes TC by using machine learning techniques trained with information on morphological parameters of segmented nuclei in order to classify regions of the image as tumour or normal. The data is provided by the 2019 SPIE Breast challenge, which was proposed to develop automated TC computation algorithms. Three algorithms were implemented: Support Vector Machines, Nearest K-means and Adaptive Boosting (AdaBoost) decision trees. Performance based on accuracy is compared and evaluated and the best result was obtained with Support Vector Machines. Results obtained by the methods implemented were submitted during ongoing challenge with a maximum of 0.76 of prediction probability of success.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.