Advances in Cognitive Sciences

fa طراحی یک مدل هوشمند مصنوعی برای ارزیابی عدالت در نظام سلامت در زمان همه‌گیری ویروس کرونا Designing an artificial intelligence model to evaluate justice in the health system during the COVID-19 pandemic مدل سازی شناختی، پردازش سیگنال و تصویربرداری مغز پژوهشي اصیل Research <div style="text-align: justify;">مقدمه: رعایت عدالت در نظام سلامت سبب شده که افراد یک جامعه به طور منصفانه از مراقبت‌های درمانی بهره‌مند شده و همین امر از افزایش بیماری و مرگ‌ومیر جلوگیری می‌کند. هدف از انجام این پژوهش طراحی مدل هوشمند سیستم سلامت مبتنی بر عدالت در نظام سلامت در دوران همه‌گیری ویروس کرونا بود. روش کار: این پژوهش بر اساس هدف پژوهش، یک مطالعه کاربردی و بر اساس شیوه گردآوری داده‌ها از نوع توصیفی_تحلیلی بود. در این پژوهش ابتدا با مطالعه‌ ادبیات پژوهش‌های پیشین و استفاده از نظر خبرگان، عوامل تاثیرگذار بر عدالت در نظام سلامت استخراج شد. سپس پرسشنامه‌ای طراحی گردید که داده‌های 109 نفر از بیماران مبتلا به ویروس کرونا که به مراکز درمانی شهر اصفهان مراجعه کرده‌ بودند جمع‌آوری شد. سپس با استفاده از آمار استنباطی، اعتبار روایی و پایایی پرسشنامه تایید شده و داده‌های حاصل از پرسشنامه در مدل هوشمند مبتنی بر شبکه Hopfield به کار گرفته شده و وضعیت رعایت عدالت در نظام سلامت با استفاده از مدل فوق سنجیده ‌شد. سپس دقت مدل با روش ماتریس درهم‌ریختگی ارزیابی شد. یافته‌ها: نتایج پیش‌بینی شبکه عصبی Hopfield برای رکوردهای مختلف نشان می‌دهد که بیشترین میزان برچسب تخصیص داده شده توسط شبکه، مربوط به برچسب "نسبتا عادلانه" و "کمی عادلانه" است که در کل میزان رضایتمندی از عدالت در سطح مطلوبی قرار ندارد. نتیجه‌گیری: با استفاده از بررسی نتایج این پژوهش وضعیت رعایت عدالت در دوران همه‌گیری ویروس کرونا با استفاده از مدل هوشمند بررسی شد و بر تاکید بیشتر بر پارامترهای رعایت عدالت در نظام سلامت اشاره دارد.  </div> <pre style="text-align:justify"> Introduction</pre> <pre style="text-align:justify"> The issue of justice is a cognitive and complex issue, including many dimensions. Observance of justice in the health system has resulted in individuals in a community receiving fair medical care, which in turn prevents an increase in morbidity and mortality. Healthcare justice encompasses various aspects, including the equitable distribution of resources and facilities, ensuring access to diverse care services, providing care responsive to patient needs, fair financing and allocation of healthcare costs based on financial capacity, efficient management, and the implementation of impartial decision-making processes in resource allocation (6, 7).</pre> <pre style="text-align:justify"> Notably, one of the basic aspects in the evaluation of healthcare services is the issue of justice, which has become more important during the COVID-19 pandemic (8). This research tried to measure patients' assessment of justice in the health system during the COVID-19 pandemic, which is an essential challenge in Iran and involves all parts of the health system, by designing an intelligent model using Hopfield's neural network. An intelligent model means that the data is stored with minimal user intervention and automatically processing and extracting patterns and connections between a subset of data from these hidden, valuable, valid, novel, and understandable patterns for system learning (similar to experience human learning). Moreover, it can be used to predict the future (9).</pre> The design of an intelligent model during the COVID-19 pandemic can be used as a model for future research to design and improve models in other pandemics. <pre style="text-align:justify"> Methods</pre> <pre style="margin-bottom:16px; text-align:justify"> This research is applied research based on the method of data collection, which is descriptive-analytical. First, by studying the literature of the previous studies and using the experts' opinions, factors affecting justice in the health system were extracted. Then, a questionnaire was designed, and the data of 109 patients with COVID-19 who visited the medical centers of Isfahan city, Iran, were collected. After confirming the validity and reliability of the questionnaire, data obtained from the questionnaire was used in the intelligent model based on the Hopfield network, and the status of justice in the health system was measured. In order to design an intelligent model, Python programming language was used in the Anaconda environment and CRISP-DM methodology. CRISP-DM methodology consists of six phases: In the problem recognition phase, the problem and the purpose of the problem are explained first. The present research aims to design an intelligent model of a health system based on justice. In the data recognition stage, the data from the questionnaires were collected in an Excel file. This data had 25 columns and 109 rows. The questionnaire of the current research was published online on social networks such as LinkedIn, WhatsApp, Telegram and the like to the people who were infected with the COVID-19 disease and referred to the medical centers of Isfahan city, and some of the questionnaires were also given in person by referring to Al-Zahra hospital. In the data preparation stage, irrelevant and redundant fields were first removed in order to clean the data. Excess spaces and words have been eliminated from the text. Additionally, a 5-point Likert scale was utilized to respond to the survey questions. For this purpose, in order to create uniformity and convert qualitative propositions into quantitative ones, a number was assigned to each of the answers. For example, the word "Very good" was assigned the number 5, "Good" the number 4, "Average" the number 3, "Weak" the number 2, and "Very weak" the number 1. Then, the names of the columns were changed and replaced with English names. In the modeling phase, we modeled the data using the Hopfield neural network, and in the evaluation phase, the accuracy of the model was evaluated using the confusion matrix method. Moreover, according to the requirements, the development and deployment phase can be as simple as producing reports and as complex as the development of repeatable data mining processes.</pre> <pre style="text-align:justify"> Results</pre> <pre style="text-align:justify"> Hopfield's recurrent neural network was trained and subsequently, it successfully identified the desired label corresponding to each spike within the dataset. For this purpose, the data was saved in a new CSV file called dataset_with_result.csv. In this file, a column called predicted label was created, which was created according to the specified patterns for each row. The labels "Very unfair", "Quite unfair", "Slightly fair", "Quite fair," "Very fair," or "Very fair" are divided. Then, this study divided the dataset based on the last prediction column of the CSV file and counted the number of data in each of the prediction labels using the count method. The number of labels predicted in table 1 is as follows:</pre> <pre style="text-align:center"> Table 1. Number of data in each label</pre> <table align="center" class="MsoTableGrid" dir="rtl" style="border-collapse:collapse; border:none"> <tbody> <tr> <td style="border-bottom:1px solid black; width:104px; padding:0in 7px 0in 7px; border-top:1px solid black; border-right:1px solid black; border-left:1px solid black" valign="top">Label number</td> <td style="border-bottom:1px solid black; width:113px; padding:0in 7px 0in 7px; border-top:1px solid black; border-right:none; border-left:1px solid black" valign="top">Label name</td> <td style="border-bottom:1px solid black; width:198px; padding:0in 7px 0in 7px; border-top:1px solid black; border-right:none; border-left:1px solid black" valign="top">Number of data in each label</td> </tr> <tr> <td style="border-bottom:1px solid black; width:104px; padding:0in 7px 0in 7px; border-top:none; border-right:1px solid black; border-left:1px solid black" valign="top">1</td> <td style="border-bottom:1px solid black; width:113px; padding:0in 7px 0in 7px; border-top:none; border-right:none; border-left:1px solid black" valign="top">Very unfair</td> <td style="border-bottom:1px solid black; width:198px; padding:0in 7px 0in 7px; border-top:none; border-right:none; border-left:1px solid black" valign="top">10</td> </tr> <tr> <td style="border-bottom:1px solid black; width:104px; padding:0in 7px 0in 7px; border-top:none; border-right:1px solid black; border-left:1px solid black" valign="top">2</td> <td style="border-bottom:1px solid black; width:113px; padding:0in 7px 0in 7px; border-top:none; border-right:none; border-left:1px solid black" valign="top">Quite unfair</td> <td style="border-bottom:1px solid black; width:198px; padding:0in 7px 0in 7px; border-top:none; border-right:none; border-left:1px solid black" valign="top">2</td> </tr> <tr> <td style="border-bottom:1px solid black; width:104px; padding:0in 7px 0in 7px; border-top:none; border-right:1px solid black; border-left:1px solid black" valign="top">3</td> <td style="border-bottom:1px solid black; width:113px; padding:0in 7px 0in 7px; border-top:none; border-right:none; border-left:1px solid black" valign="top">Slightly Fair</td> <td style="border-bottom:1px solid black; width:198px; padding:0in 7px 0in 7px; border-top:none; border-right:none; border-left:1px solid black" valign="top">28</td> </tr> <tr> <td style="border-bottom:1px solid black; width:104px; padding:0in 7px 0in 7px; border-top:none; border-right:1px solid black; border-left:1px solid black" valign="top">4</td> <td style="border-bottom:1px solid black; width:113px; padding:0in 7px 0in 7px; border-top:none; border-right:none; border-left:1px solid black" valign="top">Quite fair</td> <td style="border-bottom:1px solid black; width:198px; padding:0in 7px 0in 7px; border-top:none; border-right:none; border-left:1px solid black" valign="top">54</td> </tr> <tr> <td style="border-bottom:1px solid black; width:104px; padding:0in 7px 0in 7px; border-top:none; border-right:1px solid black; border-left:1px solid black" valign="top">5</td> <td style="border-bottom:1px solid black; width:113px; padding:0in 7px 0in 7px; border-top:none; border-right:none; border-left:1px solid black" valign="top">Very fair</td> <td style="border-bottom:1px solid black; width:198px; padding:0in 7px 0in 7px; border-top:none; border-right:none; border-left:1px solid black" valign="top">14</td> </tr> </tbody> </table> <pre> </pre> <pre style="text-align:justify"> The prediction results of the Hopfield neural network for different records show that the highest amount of labels assigned by the network is related to "Quite fair" and "Slightly fair" labels, indicating that the level of satisfaction with justice is not at a favorable level.</pre> <pre style="text-align:justify"> Moreover, in order to evaluate the results related to the accuracy of the model, the confusion matrix was used. The elements of the main diameter of the matrix represent the prediction of the Hopfield neural network in each of the classes, and the network has determined all the data to be the most similar pattern from the set of training patterns.</pre> <pre style="text-align:justify"> Conclusion</pre> <pre style="text-align:justify"> The study employed an intelligent model to examine the state of justice throughout the Covid-19 pandemic, highlighting the need for greater focus on equitable parameters within the healthcare system. Considering that most of the labels detected by the Hopfield network were related to "Quite fair" and then "Slightly fair", more and better attention is paid to the discussion of justice during the COVID-19 epidemic. Undeniably, the largest number of predicted labels indicates that the state of justice in the health system is not at an optimal level. More attention should be paid to the components affecting justice in the health system used in the smart model, such as the appropriate distribution of resources and facilities such as medicine, equipment medicine, access to medical centers, sufficient human resources in medical centers, access to COVID-19 diagnostic tests, and the like. The improvement of the disease is effective, the treatment process is faster, and the death rate and cases of this virus are reduced.</pre> <pre style="margin-bottom:16px; text-align:justify"> The topic of justice within the health system emerged prominently during the coronavirus pandemic, highlighting a critical social crisis. However, this concern is not unique to COVID-19; it surfaces with each pandemic or epidemic that strikes. Evaluating the degree of justice in our health system is both important and necessary. Thus, this study lays the groundwork for future research aimed at assessing justice through innovative methods.</pre> <pre style="text-align:justify"> Ethical Considerations</pre> <pre style="text-align:justify"> Compliance with ethical guidelines</pre> <pre style="margin-bottom:16px; text-align:justify"> This study was conducted following ethical principles, such as the consent of participants, respect for the confidentiality of patient information, and freedom to leave the research process.</pre> <pre style="text-align:justify"> Authors' contributions</pre> <pre style="margin-bottom:16px; text-align:justify"> Saeed Setayeshi: Involved in research design. Fatemeh Chelongar: Involved in collecting and analyzing data. All the authors were involved in writing and editing the manuscript.</pre> Funding <pre style="margin-bottom:16px; text-align:justify"> No financial assistance has been received from any organization.</pre> Acknowledgments <pre style="margin-bottom:16px; text-align:justify"> This article is the result of the first author's Master's thesis in Information Technology Management at Allameh Tabataba'i University. The authors would like to thank all the people who contributed to this research.</pre> <pre> Conflict of interest</pre> The authors reported no potential conflict of interest عدالت, نظام سلامت, همه‌گیری, ویروس‌کرونا, شبکه عصبی Hopfield Justice, Health system, Pandemic, COVID-19, Hopfield neural network 128 140 http://icssjournal.ir/browse.php?a_code=A-10-1734-1&slc_lang=fa&sid=1 Fatemeh Chelongar فاطمه چلونگر 100319475328460016110 100319475328460016110 Yes Master of Information Technology Management, Allameh Tabataba'i University, Tehran, Iran کارشناس ارشد، مدیریت فناوری اطلاعات، دانشکده مدیریت و حسابداری، دانشگاه علامه طباطبائی، تهران، ایران Saeed Setayeshi سعید ستایشی Setayesh@aut.ac.ir 100319475328460016111 100319475328460016111 No Professor of Department of Physics and Energy Engineering, Amirkabir University of Technology, Tehran, Iran استاد گروه مهندسی هسته‌ای، دانشکده فیزیک و مهندسی انرژی، دانشگاه صنعتی امیرکبیر، تهران، ایران