ABSTRACT
Introduction: The management of medical images has been gaining followers based on the advantages it offers for the diagnosis of diseases, which, like COVID-19, present with clinical manifestations that can be captured in the form of images. Objective: Take advantage of the quasi-periodicity of the principal components (PCs) in the decomposition into PCs of medical images, which represent dermatological manifestations in paucisymptomatic patients of COVID-19. Methods: Here, a set of photos was taken of one of the most frequent patterns in COVID-19, the maculopapular pattern, characterized by an erythmatopapular rash, and compression of one of the medical images was performed. Said compression was carried out in different ways: (1) using two PCs, (2) using both a periodic PC and a non-periodic PC, (3) using two periodic PCs, (4) using a single PC, and (5) using a single periodic PC. Result: The results of this research proved that it is possible to work with acceptable reconstructions of compressed images in the field of dermatology, without losing the quality and characteristics that allow to reach a correct diagnosis. In addition, this achievement permits to correctly classify many diseases without fear of being wrong. Conclusion: With the method presented, the use of a robust medical image compression technique that could be very useful in the field of health is proposed. The images allow the diagnosis of diseases such as COVID-19 in paucisymptomatic patients, understanding them allows minimizing their weight without losing quality, which facilitates their use and storage.
Introducción: El empleo de imágenes médicas en el diagnóstico de enfermedades ha ido ganando adeptos. Un ejemplo es la COVID-19 que cursa con manifestaciones clínicas dermatológicas. Objetivo: Aprovechar la cuasi-periodicidad de los componentes principales de la descomposición en imágenes médicas, que representan manifestaciones dermatológicas en pacientes paucisintomáticos de COVID-19. Métodos: Se tomó un conjunto de fotografías de uno de los patrones más frecuentes en la COVID-19 (el patrón maculopapular), caracterizado por un exantema eritematopapular, y se realizó la compresión de una de las imágenes médicas. Dicha compresión se realizó de diferentes formas: (1) usando dos componentes principales, (2) usando tanto un componente principal periódico como no periódico, (3) dos componentes principales periódicos, (4) un único componente principal, y (5) un solo componente principal periódico. Resultados: Es posible trabajar con reconstrucciones aceptables de imágenes comprimidas en el campo de la dermatología, sin perder la calidad y características que permitan llegar a un diagnóstico correcto. Además, este logro permite clasificar correctamente muchas enfermedades sin miedo a equivocarse. Conclusiones: Con el método presentado se propone el uso de una técnica robusta de compresión de imágenes médicas que podría ser de gran utilidad en el campo de la salud. Las imágenes permiten el diagnóstico de enfermedades como la COVID-19 en pacientes paucisintomáticos, con cuya compresión se minimiza su peso sin perder la calidad, lo que facilita su uso y almacenamiento.
Subject(s)
Humans , Data Compression/methodsABSTRACT
Nowadays, increasing air-pollution levels are a public health concern that affects all living beings, with the most polluting gases being present in urban environments. For this reason, this research presents portable Internet of Things (IoT) environmental monitoring devices that can be installed in vehicles and that send message queuing telemetry transport (MQTT) messages to a server, with a time series database allocated in edge computing. The visualization stage is performed in cloud computing to determine the city air-pollution concentration using three different labels: low, normal, and high. To determine the environmental conditions in Ibarra, Ecuador, a data analysis scheme is used with outlier detection and supervised classification stages. In terms of relevant results, the performance percentage of the IoT nodes used to infer air quality was greater than 90%. In addition, the memory consumption was 14 Kbytes in a flash and 3 Kbytes in a RAM, reducing the power consumption and bandwidth needed in traditional air-pollution measuring stations.
