Artificial Neural Networks in Medical Sciences

Abstract

Introduction

I. INTRODUCTION

The use of artificial intelligence has currently been largely utilised in medical diagnostic programs. these strategies are adaptive processing algorithms which might be capable of examine multiple and heterogeneous kinds of clinical information a good way to combine them into categorised outputs. in this examine, we in short review studies papers bearing on the problem and talk the idea, opportunities and applicability of artificial neural network techniques for scientific diagnosis. This study focuses on clinical researches from 2010 to 2021. Most of the papers studied in this review in particular targeted on bodily illnesses regardless of mental sicknesses, the utility of ANN in mental and comorbid disorders have no longer been thoroughly studied, ANN fashions and algorithms specifically don't forget homogeneous input records resources and now not heterogeneous enter records resources, and ANN models on multi-target output systems are few in comparison to single-output ANN fashions.

analysis is one of the fundamental tasks of all docs, and its importance for humans cannot be overemphasized. In analysis, medical doctors are tasked with inferring sure sicknesses or formulating treatments based on specific signs and symptoms and symptoms or observations and expertise. Defined diagnosis as "the system of assigning a label to a disorder or different trouble via examining observations and signs and symptoms." absolutely positioned, diagnostics is classed by using experts as a rigorous, complicated and multifaceted task this is related to many issues. some of those troubles encompass disorder signs and symptoms that are not unique to simplest one disease and occasionally overlap with the signs and symptoms of different diseases with a indistinct or inadequate description of ways and what the patient feels within the body because of memory or voice loss; have extreme mental conditions that impair cognitive abilities; ailment incidence; and hazard elements consisting of age, intercourse, and body mass index (BMI), which affect each the sample of dependencies between signs and diseases. Others encompass wrong and premature interpretation of data provided by using sufferers and pointless delays and mistakes in the evaluation of laboratory effects, loss of device, bad technicalities and unbalanced ratio of sufferers to docs inside the fitness quarter in particular in growing countries, and so on. no longer enormously, they claim that fifty% diagnoses are incorrect.

This paper takes a look and review based on clinical research papers related to the software of artificial neural network (ANN) strategies. especially, it makes a study of applicable literary works.

II. LITERATURE STUDY

Conclusion

The research papers studied above gives us the conclusion of growing interest in computer science, especially artificial neural networks (ANNs), whose models and algorithms are becoming standard tools for decision support systems and expert systems. ANN is undoubtedly a powerful tool to help doctors, other healthcare professionals, and stakeholders implement diagnosis, prognosis, and other enforcement measures. The various advantages of ANN as found by the studies are listed below: 1) flexible and prone to adaptiveness regarding a system 2) Can process huge amounts of medical data 3) Overlooking of information which is relevant is reduced significantly 4) Diagnosis of diseases in a timely manner 5) processing of datasets from various sources not limited to symptoms etc. This will be an important computational model in the development of modern clinical decision support and expert systems. The studies reviewed here suggest that ANNS has been shown to be useful in the satisfactory medical diagnosis of both physical and mental illness. Their use has improved the accuracy and reliability of the diagnosis, which in turn has improved patient satisfaction. Protecting patient privacy in all the methods discussed in the above paper is a major issue that needs to be addressed. Also, despite the widespread use of ANN and other intelligent computational algorithms, this tool is only considered as an intermediary in the final decision of the physician with the ultimate responsibility for critically assessing ANN\'s output. LIMITATIONS OF ANN This study shows that limitations are still there in all of these methods and ANN models which are employed to the solving of the problem of diagnosis of diseases. a) Hardware Dependency: Due to its structure, artificial neural networks require processors with parallel computing power. Therefore, the version depends on the device. b) Unexplained Network Function: This is the main problem with ANN. If ANN provides an exploratory solution, the reason and method will not be given. This reduces confidence in the network. c) Ensuring Proper Network Structure: There are no specific rules for determining the structure of artificial neural networks Appropriate network structure is realized by experience and trial and error. d) Difficult To Show Problems On The Network: ANN can handle numerical information. Before deploying to ANN, you need to convert the problem to a number. The display mechanism determined has a direct impact on network performance. This depends on the skill of the user. e) Network Duration Unknown: The network is reduced to a certain value of the error on the sample means that the training has been completed. The value does not give us optimum results.

References

[1] Baxt, W. G (1991) artificial neural network applied in diagnosing myocardial infarction. [2] Eladia Maria Pena Mendez, Artificial neural Networks in medical diagnosis [3] Egba. Anwaity Fraser, Okonkwo, Obikwelu R. Artificial Neural Networks for Medical Diagnosis: A Review of Recent Trends [4] IHPME, University of Toronto, Toronto Health Economics Technology Assessment University Health Network Uses of artificial neural networks in health care organizations with scope review method [5] The University of Warwick, UNITED KINGDOM Artificial Neural Networks in Health Care [6] Noura Ajam Heart Diseases Diagnoseds using Artificial Neural Network [7] B. Tarle, S. Jena Artificial Neural Network with Classified Algorithms for Diagnosis of Heart Disease. [8] A. Muñoz,E. Palomo, . Jeréz-Calero Use of an ANN Valuation of MTF & Melatonin on Children with ADHD.

Copyright

Copyright © 2022 Abhishek Nath, Aayush Kumar, Dr. Manjot Kaur Bhatia . This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.