2020
Complexity of brain tumors
Miguel Martín-Landrove, Francisco Torres-Hoyos, Antonio Rueda-Toicen
Physica A 537 (2020) 122696
Tumor growth is a complex process characterized by uncontrolled cell proliferation and invasion of neighboring tissues. The understanding of these phenomena is of vital importance to establish the appropriate diagnosis and therapeutic strategies and starts with the evaluation of their complex morphology with suitable descriptors, such as those produced by scaling analysis. In the present work, scaling analysis is used for the extraction of dynamic parameters that characterize tumor growth processes in brain tumors. The emphasis in the analysis is on the assessment of general properties of tumor growth, such as the Family–Vicsek ansatz, which includes a great variety of ballistic growth models. Results indicate in a definitive way that gliomas strictly behave as it is proposed by the ansatz, while benign tumors behave quite differently. As a complementary view, complex visibility networks derived from the tumor interface support these results and its use is introduced as a possible descriptor in the understanding of tumor growth dynamics.
Retrieving 3D medical data along fitted curved slices and their display
Marco Paluszny, Dany Ríos
BMC Medical Informatics and DecisionMaking (2020) 20:23
Background: Computer tomography and magnetic resonance imaging are usually offered to the clinician in the form of sequences of axial, sagittal and coronal planar cuts. Frequently this does not allow for the full inspection of the morphology of the area of interest, because it is limited by the planarity. Efforts have been made to extract information along curved slices but their planar display is prone to metric deformation.
Methods: We propose a new visualization alternative of 3D medical volumes using curved slices adapted to areas of interest. We use surfaces fitted to specific organs as visualization canvasses. We describe the differential geometry techniques used to build the surfaces that may be isometrically flattened. These are referred to as develpable surfaces.
Results: We show concrete examples deemed useful for the development of clinical and educational tools. Our examples are centered in magnetic resonance data of the rotator cuff muscle complex and computed tomography data of maxillofacial and dental studies. We also look at the extraction and display of information from volumes of aortic aneurysms along transversal surfaces.
Discussion: We look at extensions of the technique and propose further possible clinical use of texturized surfaces in the context of volume navigation.
Conclusions: We presented a technique to extract information from computer tomography and magnetic resonance volumes, using two different texturization techniques. In the cases that the fitting surfaces are chosen to be developable, they may be flattened without distortion. We also discuss how tu use the technique in other visualization tasks such as volume navigation and detection of volumetric features.
Visualization of dental information within CT volumes
.C. González, G. Albrecht, M. Paluszny
Mathematics and Computers in Simulation 173 (2020) 71–84
The use of imaging techniques in odontological practice has been increasing in recent years, and today full volumes are imaged. In this context, we consider the problem of visualizing the maxillofacial bone along curved slices. To this end we employ a special type of developable surfaces which are perfectly adaptable to the considered anatomical structure and allow for planar, distortion-free visualization.
Physical Activity Classification Using an Artificial Neural Networks Based on the Analysis of Anthropometric Measurements
Antonio J. Alvarez, Erika Severeyn, Sara Wong, Héctor Herrera, Jesús Velásquez, Alexandra La Cruz
International Conference on Systems and Information Sciences ICCIS 2020: Systems and Information Sciences pp 60-70
Physical activity is one of the most important factors in leading a healthy life, which has increased the interest in the scientific community to evaluate methods and tools that can help people maintain an exercise routine, such as portable devices that can track the movements of the user and provide an appropriate feedback. Interest has also emerged in assessing the discrimination between physically active and inactive persons through the use of readily available data, which is the aim of this work. In this case, we used an auto-encoder to find the most outstanding characteristics of an anthropometric data set, in order to get the most representative attributes. Then use them to train an Artificial Neural Network (ANN), so that it could learn to identify between a physically active and a sedentary person. The ANN obtained 81% accuracy, 82% precision, 88% recall, 83% F1 score and 0.89 AUC. These results position the ANN as a viable model that could be used as a tool in scenarios such as customer profiling for different interested companies.
Analysis of Receiver Operating Characteristic Curve Using Anthropometric Measurements for Obesity Diagnosis
Erika Severeyn, Jesús Velásquez, Héctor Herrera, Sara Wong, Alexandra La Cruz
International Conference on Systems and Information Sciences ICCIS 2020: Systems and Information Sciences pp 71-80
Today, obesity is a major public health problem. Obesity increases the risk of diabetes, coronary artery disease, stroke, cancer, premature death and contributes substantially the costs to society. Obesity can be diagnosed with body mass index (BMI). According to the World Health Organization, the diagnosis of overweight is made with a BMI ≥25 Kg/m2, and obesity with a BMI ≥30 kg/m2. The diagnosis of obesity has been made using the abdominal circumference, the hip circumference, the thickness of the skin folds and the percentage of body fat (measured directly or indirectly). Besides, the characteristic operating receiver curves (ROC) have been used to find the optimal cut-off points of hip and waist circumference for the diagnosis of obesity. The aim of this study is to evaluate the ability of anthropometric measures for diagnosing overweight and obesity. A database of 1053 subjects with 26 anthropometric measurements was used. For evaluating the predictive ability of anthropometric measures, the area under the ROC curve (AUCROC), the sensitivity (SEN), the specificity (SPE), the negative predictive value (NPV ) and the positive predictive value (PPV) were calculated. The hip circumference was the anthropometric value that best detected overweight/obese subjects with a AUCROC = 0.932 (SEN = 0.871 SPE = 0.855, PPV = 0.536 and NPV = 0.972) and an optimal cut-off point of 97.2 cm for recognition of obesity. The findings reported in this
research suggest that the diagnosis of obesity can be made with anthropometric measurements. In the future, machine learning techniques, such as: k-means, neural networks or support vector machines; will be explored for the detection of overweight and obesity.
Analysis of Anthropometric Measurements Using Receiver Operating Characteristic Curve for Impaired Waist to Height Ratio Detection
Erika Severeyn, Alexandra La Cruz, Sara Wong, Gilberto Perpiñan
The International Conference on Advances in Emerging Trends and Technologies ICAETT 2020: Advances in Emerging Trends and Technologies pp 167-178
Metabolic dysfunctions such as obesity, insulin resistance, metabolic syndrome, and glucose tolerance are strongly related to each other. The presence of any of them in a person translates into a high risk of diseases such as diabetes, heart failure, and cardiovascular disease. Anthropometric measurements such as body circumferences, body folds, and anthropometric indices such as waist-height ratio (WHtR) and body mass index (BMI) have been widely used in the study of metabolic diseases. This study aims to look for relationships between WHtR and anthropometric measurements such as subcutaneous folds and body circumferences.For this purpose, a database of 1863 subjects was used, 16 anthropometric variables were measured for each participant in the database and the BMI was calculated. The receiver operating characteristic (ROC) curves were used to assess the ability of BMI and each anthropometric measurement was used to diagnose BMI impairment. The findings reported in this research strongly suggest that the diagnosis of WHtR deficiency can be made from circumferences, skinfolds, and BMI. In this study, the anthropometric measures that best detect subjects with WHtR deficiency were BMI, subscapular skinfold, supra iliac skinfold, and arm circumference with a high probability of detecting normal WHtR-deficient subjects. Abdominal circumference is one of the areas that have the most direct relationship with cardiac metabolic risk, however, the findings of this study open the possibility of studying accumulated fat tissue in the arms and back as areas that could also indicate a risk of metabolic dysfunction.
Classification of Impaired Waist to Height Ratio Using Machine Learning Technique
Alexandra La Cruz, Erika Severeyn, Sara Wong, Gilberto Perpiñan
The International Conference on Advances in Emerging Trends and Technologies ICAETT 2020: Advances in Emerging Trends and Technologies pp 179-190
Metabolic dysfunctions are a set of metabolic risk factors that include abdominal obesity, dyslipidemia, insulin resistance, among others. Individuals with any of these metabolic dysfunctions are at high risk of developing type 2 diabetes and cardiovascular disease. Several parameters and anthropometric indices are used to detect metabolic dysfunctions, such as waist circumference and waist-height ratio (WHtR). The WHtR has an advantage over the body mass index (BMI) since the WHtR provides information on the distribution of body fat, particularly abdominal fat. Central fat distribution is associated with more significant cardio-metabolic health risks than total body fat. Machine learning techniques involve algorithms capable of predicting and analyzing data, increasing our understanding of the events being studied. k-means is a clustering algorithm that has been used in the detection of obesity. This research aims to apply the k-means grouping algorithm to study its capability as an impaired WHtR classifier. Accuracy (Acc), recall (Rec), and precision (P) were calculated. A database of 1863 subjects was used; the database consists of fifteen (15) anthropometric variables and two (2) indices; each anthropometric variable was measured for each participant. The results reported in this research suggest that the k-means clustering algorithm is an acceptable classifier of impaired WHtR subjects (Acc = 0.81, P = 0.83, and Rec = 0.73). Besides, the k-means algorithm was able to detect subjects with overweight and fatty tissue deposits in the back and arm areas, suggesting that fat accumulation in these areas is directly related to abdominal fat accumulation.
Comparison of cell contour closing methods in microscopy images
Manuel G. Forero, Alexandra La Cruz, Jorge Español, Diego Urrego
Proc. SPIE 11510, Applications of Digital Image Processing XLIII, 115101N (21 August 2020)
Cell counting and tracking approaches are widely used in microscopy image processing. Cells may be of different shapes and may be very crowded or relatively close together. In both cases, the correct identification of each cell requires the detection and tracking of its contour. But, this is not always possible due to noise, image blurring from signal degradation during the acquisition process and staining problems. Generally, cell segmentation approaches use filtering techniques, Hough transform, combined with morphological operators to address this problem. However, usually, not all contours can be closed. Therefore, heuristic contour closing techniques have been employed to achieve better results. Despite being necessary, no comparative studies on this type of methods were found in the literature. For that reason, this paper compares three approaches to contour tracking and closing. Two of them use one end of a contour as a starting point and trace a path along the edge of the cell seeking to find another endpoint of the cell. This is done using the first or second ring of neighboring pixels around the starting point. The heuristics used are based on region growing taking the information from the first or second ring of neighboring pixels and keeping the direction along the plotted path. The third method employs a modification of Dijkstra's algorithm. This approach employs two seed points located at each possible end of the contour. This paper presents a description of these techniques and evaluates the results in microscopy images.
Prediction of Abnormal Body Fat Percentage by Anthropometrics Parameters Using Receiver Operating Characteristic Curve
Erika Severeyn, Sara Wong, Héctor herrera,, Alexandra La Cruz, Jesús Velásquez, Mónica Huerta
2020 IEEE ANDESCON, Quito, 2020, pp. 1-6
The World Health Organization has defined obesity ‟as the abnormal or excessive fat accumulation that represents a risk to health". Although obesity is characterized by an excessive amount of body fat, it is commonly measured using body mass index which is unable to differentiate between elevated body fat content and increased lean mass. The indicator that best predicts obesity is the one that quantify adipose tissue and, therefore, the estimation of body fat percentage (BFP). Skinfolds have been used to measure the BFP, based on the Siri and Brozec formula. There are no official cut-off points for BFP, as the associated data is relatively insufficient worldwide. Studies agreed that fewer than 25% in men and 30% in women are commonly used as normal BFP. The aim of this study is to evaluate the capability of the anthropometrics variables to discriminate subjects with abnormal BFP. A database of 1053 subjects with 28 anthropometrics measures was used. Area under the receiver operating characteristic curves (AUC ROC ), sensibility (SEN), specificity (SPE) and negative predictive value (NPV) was calculated to evaluate the predictive ability of anthropometric variables measured. Three circumferences (Arm, waist and hip) and four skinfolds (calf, suprailiac, abdominal and thigh) were the variables with the best abnormal BFP detection capability, with an AUC ROC >0.800 (SEN>0.760 and SPE>0.673). Having a high probability of detecting subjects with normal BFP (NPV>0.970). Easier variables to acquire, such as waist, arm, and hip circumferences, could be used in low-income countries where it is not easy to have a plicometer.
Study of Basic Reproduction Number Projection of SARS-CoV-2 Epidemic in USA and Brazil
Erika Severeyn, Sara Wong, Héctor herrera,, Alexandra La Cruz, Jesús Velásquez, Mónica Huerta
2020 IEEE ANDESCON, Quito, 2020, pp. 1-6
In December 2019, a group of patients presented a diagnosis of pneumonia of unknown etiology in Hubei Province, Wuhan, China. By January 2020, authorities around the world faced a new coronavirus (SARS-CoV-2). By August 2020, the two countries with the highest number of SARS-CoV-2 infections are the USA and Brazil. The transmission rate of a virus is studied from the basic reproduction number (R 0 ). The SIR model is the simplest compartmental epidemiological model (Susceptible, Infectious and Recovered). The SIR model can be used to estimate R 0 by fitting the curve of the infected compartment to the experimental curve of infected subjects per day. The aim of this work is to study the projection of the R 0 of SARS-CoV-2 in the USA and Brazil. For this purpose, five experiments were performed by adjusting the SIR model curve of infected compartment to experimental data at five time intervals (the first 14, 28, 42, 56 and 187 days for the USA data, and 177 days for Brazil data). In the first two time intervals the R 0 varied between 5.46 and 7.75 for the USA data and 1.84 and 4.29 for Brazil data, and in the last three time intervals the R 0 decreased to 1.05 for the USA data and 1.01 for Brazil data, suggesting that the social distancing measures implemented in both countries were able to decrease the infection spreading. The differences in the R 0 values of the five experiments imply that R 0 also depends on the preventive measures implemented to face the pandemic.
A Novel Method for Reconstructing CT Images in GATE/GEANT4 with Application in Medical Imaging: A Complexity Analysis Approach
Neda Gholami, Mohammad Mahdi Dehshibi, Andrew Adamatzky, Antonio Rueda-Toicen, Hector Zenil, Mahmood Fazlali, David Masip
Journal of Information Processing Vol.28 161–168 (Feb. 2020)
For reconstructing CT images in the clinical setting, ‘effective energy’ is usually used instead of the total X-ray spectrum. This approximation causes an accuracy decline. We proposed to quantize the total X-ray spectrum into irregular intervals to preserve accuracy. A phantom consisting of the skull, rib bone, and lung tissues was irradiated with CT configuration in GATE/GEANT4. We applied inverse Radon transform to the obtained Sinogram to construct a Pixel-based Attenuation Matrix (PAM). PAM was then used to weight the calculated Hounsfield unit scale (HU) of each interval's representative energy. Finally, we multiplied the associated normalized photon flux of each interval to the calculated HUs. The performance of the proposed method was evaluated in the course of Complexity and Visual analysis. Entropy measurements, Kolmogorov complexity, and morphological richness were calculated to evaluate the complexity. Quantitative visual criteria (i.e., PSNR, FSIM, SSIM, and MSE) were reported to show the effectiveness of the fuzzy C-means approach in the segmenting task.
