Dr. Héctor Gabriel Acosta Mesa
ABSTRACTS DE PUBLICACIONES
INDEXED
Aceto-White Temporal Pattern Classification using k-NN to Identify Precancerous Cervical Lesion in Colposcopic Images (2009)
After Pap smear test, colposcopy is the most used technique to diagnose cervical cancer due to its higher sensitivity and specificity. One of the most promising approaches to improve the colposcopic test is the use of the aceto-white temporal patterns intrinsic to the color changes in digital images. However, there is not a complete understanding of how to use them to segment colposcopic images. In this work, we used the classification algorithm k-NN over the entire length of the aceto-white temporal pattern to automatically discriminate between normal and abnormal cervical tissue, reaching a sensitivity of 71% and specificity of 59%.
The neural basis of visuospatial perception in Alzheimer's disease and healthy elderly comparison subjects: An fMRI study (2009)
The neural basis of visuospatial deficits in Alzheimer's disease is unclear.We wished to investigate the neural basis of visuospatial perception in patients with Alzheimer's disease compared with healthy elderly comparison subjects using functional magnetic resonance imaging (fMRI). Twelve patients with AD and thirteen elderly comparison subjects were investigated. The patients were recruited from the local clinic and comparison subjects were from spouses and community. All participants underwent fMRI whilst viewing visuospatial stimuli and structural MRI, and findings were analysed using voxel-based morphometry. The comparison group activated V5, superior parietal lobe, parieto-occipital cortex and premotor cortices. The AD group demonstrated hypoactivation in the above regions and instead showed greater activation in inferior parietal lobule and activated additional areas. There was no structural atrophy above and beyond that found globally in patients in the identified regions of BOLD activation. To our knowledge, this is the first study to explore the neuroanatomy of perception of depth and motion in Alzheimer's disease. These specific functional deficits in AD provide evidence for an underlying patho-physiological basis for the clinically important symptom of visuospatial disorientation in patients with AD.
Discovering interobserver variability in the cytodiagnosis of breast cancer using decision trees and Bayesian networks. (2009)
We evaluate the performance of two decision tree procedures and four Bayesian network classifiers as potential decision support systems in the cytodiagnosis of breast cancer. In order to test their performance thoroughly, we use two real-world databases containing 692 cases and 322 cases collected by a single observer and 19 observers, respectively. The results show that, in general, there are considerable differences in all tests (accuracy, sensitivity, specificity, PV+, PV_ and ROC) when a specific classifier uses the single-observer dataset compared to those when this same classifier uses themultipleobserver dataset. These results suggest that different observers see different things: a problemknown as interobserver variability. We graphically unveil such a problem by presenting the structures of the decision trees and Bayesian networks resultant from running both databases.
Diagnosis of Breast Cancer Using Bayesian Networks: A case study. (2007)
We evaluate the effectiveness of seven Bayesian network classifiers as potential tools for the diagnosis of breast cancer using two real-world databases containing fine-needle aspiration of the breast lesion cases collected by a single observer and multiple observers, respectively. The results show a certain ingredient of subjectivity implicitly contained in these data: we get an average accuracy of 93.04% for the former and 83.31% for the latter. These findings suggest that observers see different things when looking at the samples in the microscope; a situation that significantly diminishes the performance of these classifiers in diagnosing such a disease.
Bayesian Model Combination and its Application to Cervical Cancer Detection (2006)
We have developed a novel methodology to combine several models using a Bayesian approach. The method selects the most relevant attributes from several models, and produces a Bayesian classifier which has a higher classification rate than any of them, and at the same time is very efficient. Based on conditional information measures, the method eliminates irrelevant variables, and joins or eliminates dependent variables; until an optimal Bayesian classifier is obtained. We have applied this method for diagnosis of precursor lesions of cervical cancer. The temporal evolution of the color changes in a sequence of colposcopy images is analyzed, and the resulting curve is fit to an approximate model. In previous work we develop 3 different mathematical models to describe the temporal evolution of each image region, and based on each model to detect regions that could have cancer. In this paper we combine the three models using our methodology and show very high accurracy for cancer detection, superior to any of the 3 original models.
Diagnosis of Chronic Idiopathic Inflammatory Bowel Disease Using Bayesian Networks (2006)
In this paper, we evaluate the effectiveness of four Bayesian network classifiers as potential tools for the histopathological diagnosis of chronic idiopathic inflammatory bowel disease (CIIBD) using a database containing endoscopic colorectal biopsies. CIIBD is the generic term for referring to two ailments known as Crohn’s disease and ulcerative colitis. The results show that the defined histological attributes, considered relevant in the medical literature for the diagnosis of CIIBD, are very good for the distinction between normal samples and CIIBD samples (Crohn’s disease and ulcerative colitis combined into a single category) but less good for the explicit distinction between Crohn’s disease and ulcerative colitis. The findings suggest an intrinsic impossibility of selecting a set of features for achieving good balance for both sensitivity and specificity for Crohn’s disease and ulcerative colitis.
How Good Are the Bayesian Information Criterion and the Minimum Description Length Principle for Model Selection? A Bayesian Network Analysis. (2006)
The Bayesian Information Criterion (BIC) and the Minimum Description Length Principle (MDL) have been widely proposed as good metrics for model selection. Such scores basically include two terms: one for accuracy and the other for complexity. Their philosophy is to find a model that rightly balances these terms. However, it is surprising that both metrics do often not work very well in practice for they overfit the data. In this paper, we present an analysis of the BIC and MDL scores using the framework of Bayesian networks that supports such a claim. To this end, we carry out different tests that include the recovery of gold-standard network structures as well as the construction and evaluation of Bayesian network classifiers. Finally, based on these results, we discuss the disadvantages of both metrics and propose some future work to examine these limitations more deeply.
Digital Image Processing of Functional Magnetic Resonance Images to Identify Stereo-sensitive Cortical Regions Using Dynemic Global Stimuli. (2004)
Functional magnetic resonance images (fMRI) were analyzed to investigate the cortical regions involved in stereoscopic vision using red/green anaglyphs to present random dot stereograms. Two experiments were conducted both of which required high attentional demands. In the first experiment the subjects were instructed to follow the path of a square defined by depth and moving in the horizontal plane contrasted with a similar sized square defined by a slight difference in luminance. Three main regions were identified V3A, V3B and BA7. To test that the observed activations were not produced by the pursuit eye movements, a second experiment required the subjects to fixate whilst a shape was presented in different random orientations. Our results suggests that areas V1, V3A and precuneus are involved in stereo disparity processing. We hypothesise that the activation of the V3B region was produced by the second order motion component induced by the spatio-temporal changes in disparity.
REFEREED
On the Possibility of Reliably Constructing a Decision Support System for the Cytodiagnosis of Breast Cancer.(2007)
We evaluate the performance of three Bayesian network classifiers as decision support system in the cytodiagnosis of breast cancer. In order to test their performance thoroughly, we use two real-world databases containing 692 cases collected by a single observer and 322 cases collected by multiple observers respectively. Surprisingly enough, these classifiers generalize well only in the former dataset. In the case of the latter one, the results given by such procedures have a considerable reduction in the sensitivity and PV- tests. These results suggest that different observers see different things: a problem known as interobserver variability. Thus, it is necessary to carry out more tests for identifying the cause of this subjectivity.
Clasificación de patrones temporales para caracterizar lesiones cervico uterinas en imágenes colposcópicas . (2007)
En el presente trabajo se propone una metodología para analizar y clasificar patrones temporales extraídos de imágenes colposcópicas, para caracterizar lesiones cervico uterinas. Las imágenes colposcópicas han sido adquiridas con luz blanca las cuales se han representado en diversos espacios de color para identificar con cual de ellos se obtiene una mejor caracterización de las series temporales. El enfoque de aprendizaje supervisado fue elegido para realizar la clasificación de las series temporales. La clasificación se realizó utilizando el algoritmo kvecinos mas cercanos. El método kfold cross validation fue utilizado para evaluar el desempeño del clasificador. Los resultados preliminares obtenidos en este trabajo en proceso, son alentadores alcanzando una sensibilidad de 73% y una especificidad de 59%, mismas que se esperan mejorar al incluir un mayor número de casos.
Comparison of the Performance of Seven Classifiers as Effective Decision Support Tools for the Cytodiagnosis of Breast Cancer: A Case Study. (2007)
We evaluate the performance of seven classifiers as effective potential decision support tools in the cytodiagnosis of breast cancer. To this end, we use a real-world database containing 692 fine needle aspiration of the breast lesion cases collected by a single observer. The results show, in average, good overall classification performance in terms of five different tests: accuracy of 93.62%, sensitivity of 89.37%, specificity of 96%, PV+ of 92% and PV- of 94.5%. With this comparison, we identify and discuss the advantages and disadvantages of each of these approaches. Finally, based on these results, we give some advice regarding the selection on the classifier depending on the user’s needs.
Modeling Aceto-White Temporal Patterns to Segment Colposcopic Images. (2007)
Colposcopy test is the second most used technique to diagnose cervical cancer disease. Some researchers have proposed to use temporal changes intrinsic to the colposcopic image sequences to automatically characterize cervical lesion. Under this approach, every single pixel on the image is represented as a Time Series of length equal to the sampling frequency times acquisition points. Although this approach seems to show promising results, the data analysis procedures have to deal with huge data set that rapidly increase with the number of cases (patients) considered in the analysis. In the present work, we perform principal component analysis (PCA) to reduce the dimensionality of the data in order to facilitate similarity measures for classification and clustering. The importance of this work is that we propose a model to parameterize the dynamics of the system using an efficient representation getting a 1.11% data compression ratio and similarity on clustering of 0.78. The feasibility of the proposed model is shown testing the similarity of the clusters generated using the k-means algorithm over the raw data and the compressed representation of real data.
Assessing Cervical Cancer Lesion Predictability Using Aceto-White Temporal Patterns with Bayesian Network Learning. (2006)
Colposcopic test is the second most used technique to diagnose cervical cancer disease. In a previous work, we propose a methodology analysis to automatically segment the colposcopic images using the temporal patterns, which produces a compact representation to facilitate similarity measures for classification. In the present work, we used different Bayesian network algo-rithms to assess their predictability scores to perform classification of different temporal patterns related to precursor lesions of cervical cancer. The aim of this work is to show evidence of the viability of this machine learning framework to segment colposcopic images.