Innovative Applications in Smart Cities

Alberto Ochoa, Genoveva Vargas-Solar, Javier Alfonso Espinosa Oviedo, Alberto Ochoa, Genoveva Vargas-Solar, Javier Alfonso Espinosa Oviedo

• Publisher: CRC Press
• Year: 2021
• Print ISBN: 9780367820961
• eBook ISBN: 9781000462265
• Edition: 1
• Topic: Business
• Subtopic: Biology
• Index: Business

This book is a compilation of chapters on scientific work in novel and innovative reference that compiles interdisciplinary perspectives about diverse issues related with Industry 4.0 and smart cities in different ways, i.e., intelligent optimisation, industrial applications in the real world, social applications and technology applications with a different perspective about existing solutions. Chapters review research in improving optimisation in smart manufacturing, logistics of products and services, optimisation of different elements in the time and location, social applications to enjoy our life of a better way and applications that increase daily life quality. This book covers applications of Industry 4.0; applications to improve the life of the citizens in a smart city; and finally, welfare of the working-age population and their expectations in their jobs correlated with the welfare-work relationship.

PART I
Daily Life in a Smart City

CHAPTER-1
Segmentation of Mammogram masses for Smart Cities Health Systems

Paula Andrea Gutiérrez-Salgado, Jose Mejia,* Leticia Ortega, Nelly Gordillo, Boris Mederos and Alberto Ochoa-Zezzatti

* Corresponding author: [email protected]

One of the fundamental aspects of smart cities is an improvement in the health sector, by providing its citizens with better care and prevention and detection of diseases. Breast cancer is one of the most common diseases and the one with the highest incidence in women worldwide. In smart cities, to improve the quality of life of its citizens, especially for women, is to diagnose breast tumors in shorter periods with simpler and automated methods. In this chapter, a new deep learning architecture is proposed to segment breast cancer tumors.

1. Introduction

According to the World Health Organization (WHO), breast cancer is one of the most common diseases with the highest incidence in women worldwide, about 522 thousand deaths are estimated annually with data collected in 2012 (OMS, 2019). In smart cities, as regards to the health sector, it seeks to improve the quality of life of its citizens (Kashif et al., 2020; Abdelaziz et al., 2019; Rathee et al., 2019), thus there is a need to diagnose breast tumors in shorter periods with simpler and automated methods that can produce accurate results. The most common method to an early diagnostic is through mammographic images, however, these images usually have noise and low contrast which can cause the doctor to have difficulty classifying different tissues. In some mammogram images, malignant tissues and normal dense tissues are presented, but it is difficult to contrast between them by applying simple thresholds when automatic methods are used (Villalba, 2016). Because of these problems, it is necessary to develop various approaches that can correctly identify the malignant tissues, which represent higher intensity values compared to background information and other regions of the breast. Also, regions where some normal dense tissues have intensities similar to the tumor region have to be excluded (Singh et al., 2015).

Universidad Autónoma de Ciudad Juárez, Av. Hermanos Escobar, Omega, 32410 Cd Juárez, Chihuahua, México.

The interpretation of a mammogram is usually difficult, sometimes it depends on the experience of medical staff. In approximately 9% of the cancers detected, tumors were visible on mammograms obtained from two years earlier (Gutiérrez et al., 2015). The key factor for early detection is the use of computerized systems. The segmentation of tumors takes a very important role in the diagnosis and timely treatment of breast cancer. Currently, there are methods to delimit tumors using artificial neuronal networks (Karianakis et al., 2015; Rafegas et al., 2018) and deep learning networks (Hamidinekoo et al., 2018; Goodfellow et al., 2016), but there is the possibility of improving them.

In this chapter, a new architecture with aims to segment mammary tumors in mammograms using deep neural networks is proposed.

2. Literature Review

There exist several diagnostic methods to perform timely detection, the use of mammography being the method most used by medical staff because of the effective and safe results of the method. The examination is carried out by firm compression of the breast between two plates, using ionizing radiation to obtain images of breast tissue, which can be interpreted as benign or malignant (Marinovich et al., 2018). ...

Table of contents

• Cover
• Title Page
• Copyright Page
• Preface
• Contents
• Prologue
• Part I: Daily Life in a Smart City
• Part II: Applications to Improve a Smart City
• Part III: Industry 4.0, Logistics 4.0 and Smart Manufacturing
• Index