Digital healthcare (DH) is a multidisciplinary field of research, at the intersection of medical sciences, biology sciences, biochemistry neurosciences, cognitive sciences and informatics. In the last years, various computational intelligence (CI) techniques and methodologies have been proposed by the researchers in order to develop digital knowledge-based systems (DKBS) for different medical and healthcare tasks. These systems are based on artificial intelligence (AI) concepts and theories. Many types of DKBS are in existence today and are applies to different healthcare domains and tasks. The objective of the paper is to presents a comprehensive and up-to-date research in the area of DHI covering a wide spectrum of CI methodological and intelligent algorithmic issues, discussing implementations and case studies, identifying the best design practices, assessing implementation models and practices of AI paradigms in digital healthcare systems .This paper presents some of the CI techniques for managing and engineering knowledge in digital healthcare systems(DHS). Some of the research results and applications of the author and his colleagues that have been carried out in last years are discussed.
Computational intelligence (CI) is the study of intelligent computer algorithms that
improve automatically through experience. CI aims to enable computers to learn
from data and make improvements without any dependence on commands in a
program.CI is an inherently interdisciplinary, includes; neurobiology, information theory,
probability, statistics, AI, control theory, Bayesian methods, physiology and
philosophy [
In the recent years, various AI paradigms and CI techniques have been proposed
by the researchers in order to develop efficient and smart systems in the areas of
health informatics and health monitoring systems.AI and CI offers robust, intelligent
algorithms and smart methods that can help to solve problems in a variety of health
and life sciences areas [
There are numerous examples of successful applications of CI in areas of diagnosis and prevention, prognosis and therapeutic decision making.
CI algorithms are used for the following tasks; (a) discovering new diseases, (b) finding predictive and therapeutic biomarkers, and(c) detecting relationships and structure among the clinical. CI contributes to the enhancement of management and information retrieval processes leading to development of intelligent (involving ontologies and natural language processing) and integrated literature searches.
Moreover, applications of CI in bioinformatics include the following areas of
research ;(a) microarray analysis, chromosome and proteome databases, modeling of
inhibition of metabolic networks(b) signal analysis (echocardiograph images and
electroencephalograph time series). and (c) drug delivery and software for pattern
recognition in biomedical data [
This paper discusses the potential role of the AI and CI approaches, techniques, which are used in developing the intelligent health informatics and health monitoring systems.
The paper discusses the following CI paradigms: (a) bio-inspired computing, (b) analogical reasoning computing, (c) vagueness and fuzzy computing, and (d) deep learning.
Also, the paper presents the challenges as well as the current research directions in the areas of digital health informatics.
Examples of the research performed by the author and his associates at Artificial Intelligence and Knowledge Engineering Research Labs, (ASU-AIKE labs) are also included.
The following cases and diseases are presented; cancer, heart, brain tumor, thrombosis diseases, Pandemic Influenza, Diabetics type-2, and volume visualization of the interior body.
The paper is organized as follows: Section 2 discusses the smart health monitoring paradigms. Section 3 reviews the bio-inspired computing approaches. In section 4, we discuss the analogical reasoning paradigms.
While section 5 discusses the vagueness and Fuzzy computing Section 6 presents our cases and applications, section 7 presents our future research directions and then we conclude in section 8.
SHMS involves deploying CI, information, and networking technologies to aid in
preventing disease, improving the quality of care and lowering overall cost [
ANN is a class of learning algorithm consisting of multiple nodes that communicate
through their connecting synapses. ANN are inspired in biological models of brain
functioning [
They are capable of learning by examples and generalizing the acquired knowledge. Due to these abilities the neural networks are widely used to find out nonlinear relations which otherwise could not be unveiled due to analytical constraints. The learned knowledge is hidden in their structure thus it is not possibly to be easily extracted and interpreted. ANN can be used in the following medical purposes: 1. Modelling: simulating the functions of the brain and neurosensory organs. 2. Signal processing: Bioelectric signal filtering and evaluation 3. System control and checking: Intelligent artificial machine control and checking based on responses of biological or technical systems given to any signals. 4. Classification tasks: Interpretation of physical and instrumental findings to achieve more accurate diagnosis. 5. Prediction: provide prognostic information based on retrospective parameter analysis.
SVM are new learning-by example paradigm for classification and regression
problems [
Deep learning is a branch of AI covering a spectrum of current exciting research
and industrial innovation that provides more efficient algorithms to deal with
largescale data in healthcare, recommender systems, learning theory, robotics, games,
neurosciences, computer vision, speech recognition, language
processing,humancomputer interaction, drug discovery, biomedical informatics, act [
GA follows the lead of genetics, reproduction, evolution and the survival for the
fittest theory by Darwin. GA is a class of machine learning algorithm that is based on
the theory of evolution [
EC is an approach to the design of learning algorithms that is structured along the
lines of the theory of evolution. A collection of potential solutions for a problem
compete with each other. The best solutions are selected and combined with each
other according to a kind of ‘survival of the fittest’ strategy. GA are a well-known
variant of evolutionary computation [
DNA computing is essential computation using biological molecules rather than
traditional silicon chips. In recent years, DNA computing has been a research tool for
solving complex problems. Despite this, it is still not easy to understand. The main
idea behind DNA (Deoxyribo Nucleic Acid) computing is to adopt a biological (wet)
technique as an efficient computing vehicle, where data are represented using strands
of DNA. Even though a DNA reaction is much slower than the cycle time of a
siliconbased computer, the inherently parallel processing offered by the DNA process plays
an important role. This massive parallelism of DNA processing is of particular
interest in solving NP-complete or NP-hard problems [
It is not uncommon to encounter molecular biological experiments which involve 6
× 1016/ml of DNA molecules. This means that we can effectively realize 60,000 Tera
Bytes of memory, assuming that each string of a DNA molecule expresses one
character. The total execution speed of a DNA computer can outshine that of a
conventional electronic computer, even though the execution time of a single DNA molecule
reaction is relatively slow. A DNA computer is thus suited to problems such as the
analysis of genome information, and the functional design of molecules (where
molecules constitute the input data) [
DNA computing will solve that problem and serve as an alternative technology.
DNA computing is also known as molecular computing. It is computing using the
processing power of molecular information instead the conventional digital
components. It is one of the non-silicon-based computing approaches. DNA has been shown
to have massive processing capabilities that might allow a DNA-based computer to
solve complex problems in a reasonable amount of time [
ARC provides both a methodology for problem solving and a cognitive model of
people [
Most commonly application of CBR used in developing expert systems
technology. In CBR expert systems, the system can reason from analogy from the past cases.
This system contains what is called “case-memory” which contains the knowledge in
the form of old cases (experiences). CES solves new problems by adapting solutions
that were used for previous and similar problems [
According to Kolodner [
Rough set theory was proposed as a new approach to vague concept description
from incomplete data. The rough set theory is one of the most useful techniques in
many real-life applications such as medicine, pharmacology, engineering, banking
and market analysis. This theory provides a powerful foundation to reveal and
discover important structures in data and to classify complex objects. The theory is very
useful in practice, e.g. in medicine, pharmacology, engineering, banking, financial
and market analysis [
Fuzzy logic allows one to express knowledge in a rule format that is close to a natural language expression. The difference between this fuzzy rule and the Booleanlogic rules we used in our forward- and backward-chaining examples is that the clauses “temperature is hot” and “humidity is sticky” are not strictly true or false. Clauses in fuzzy rules are real-valued functions called membership functions that map the fuzzy set “hot” onto the domain of the fuzzy variable “temperature” and produce a truth-value that ranges from 0.0 to 1.0 (a continuous output value, much like neural networks).
Reasoning with fuzzy rule systems is a forward-chaining procedure. The initial
numeric data values are fuzzified, that is, turned into fuzzy values using the
membership functions. Instead of a match and conflict resolution phase where we select a
triggered rule to fire, in fuzzy systems, all rules are evaluated, because all fuzzy rules
can be true to some degree (ranging from 0.0 to 1.0). The antecedent clause truth
values are combined using fuzzy logic operators (a fuzzy conjunction or and
operation takes the minimum value of the two fuzzy clauses).Next, the fuzzy sets specified
in the consequent clauses of all rules are combined, using the rule truth values as
scaling factors. The result is a single fuzzy set, which is then defuzzified to return a crisp
output value. More technical details and applications can be found in the recent book
of Voskuhl [
Cancer is a group of more than 200 different diseases. From the medical point of
view, Cancer occurs when cells become abnormal and keep dividing and forming
either benign or malignant tumors. Cancer has initial signs or symptoms if any is
observed, the patient should perform complete blood count and other clinical
examinations. To specify cancer type, patient needs to perform special lab-tests. In our
research Labs, we have performed an interesting research for developing intelligent
systems for thyroid cancer diagnoses. .This research are based on using case-based
and rule-based reasoning, ontological engineering, and artificial neural networks [
Figure 4 shows an examples for the encoded rules for cancer diagnosis..
Heart disease is a vital health care problem affecting millions of people. Figure 5 shows Types of Heart Diseases. Expert system (ES) is a consultation intelligent system that contains the knowledge and experience of one or more experts in a specific domain that anyone can tap as an aid in solving problems.
At our research unit, we have developed two versions of expert systems for heart diseases diagnosis.
The first one uses the rule-based reasoning while the second one uses case-based reasoning. The first version is composed of three components: knowledge base, user interface and computational model.
The knowledge was gathered from expert doctors in EL-Maadi Military Egyptian hospital, Egyptian Health Insurance Institute and medical books. We have built the system’s knowledge base for the 24 heart diseases and it is composed of 24 facts and 65 rules.
The system is implemented in Visual Prolog and has been tested for 13 real experiments (patients). The experimental results have shown 76.9% accuracy in estimating the right conclusion [30].
The Second version of the expert system uses CBR methodology. We have represented the knowledge in the form of frames and built the case memory for 4 heart diseases namely; mistral stenosis, left-sided heart failure, left-sided heart failure, stable angina pectoris and essential hypertension.
The system has been implemented in visual prolog for Windows and has trained set of 42 cases for Egyptian cardiac patients and has been tested by another 13 different cases.
Each case contains 33 significant attributes resettled from the statistical analysis performed to 110 cases.
The system has been tested for 13 real cases. The experimental results have shown 100% accuracy in estimating the correct results for using nearest neighbor algorithm and this percentage is dropped to 53.8% in case of using the induction algorithm. The systems are able to give an appropriate diagnosis for the presented symptoms, signs and investigations done to a cardiac patient with the corresponding certainty factors. It aims to serve as doctor diagnostic assistant and support the education for the undergraduate and postgraduate young physicians.
At our research unit, ASU-AIKE labs, a rough set-based medical system for mining patient data for predictive rules to determine thrombosis disease was developed [31,32] this system aims to search for patterns specific/sensitive to thrombosis disease. This system reduced the number of attributes that describe the thrombosis disease from 60 to 16 significant attribute in addition to extracting some decision rules, through decision applying decision algorithms, which can help young physicians to predict the thrombosis disease.
In our research group we developed a hybrid classifier that integrates the strengths of genetic algorithms and decision trees. The algorithm was applied on a medical database of 20 MB size for predicting thrombosis disease [33]. The results show that our classifier is a very promising tool for thrombosis disease prediction in terms of predictive accuracy.
Pandemic influenza has great potential to cause large and rapid increases in deaths and serious illness. The first major pandemic influenza H1N1 is recorded in 19181919, which killed 20-40 million people and is thought to be one of the most deadly pandemics in human history. In 1957, a H2N2 virus originated in China, quickly spread throughout the world and caused 1-4 million deaths worldwide. In 1968, an H3N2 virus emerged in Hong Kong for which the fatalities were 1-4 million [34]. In recent years, novel H1N1 influenza has appeared. Novel H1N1 influenza is a swineorigin flue and is often called swine flu by the public media.
In the absence of reliable pandemic detection systems, computational intelligence
techniques and paradigms have become an important smart software tools for both
policymakers and the general public [
Actually, recent estimates place the diabetes population in Sudan at around one million – around 95% of whom have type 2 diabetes and patients with diabetes make up around 10% of all hospital admissions in Sudan [38,39]. Mostly, Type 2 diabetes is strongly connected with obesity, age, and physical inactivity [39]. Most medical resources reported that 90 to 95% of diabetic is diagnosed as type 2. So, a successful intelligent control of patient food for treatment purpose must combines patient interesting food list and doctors’ efficient treatment food list. In addition, many rural communities in Sudan have extremely limited access to diabetic diet centers. People travel long distances to clinics or medical facilities, and there is a shortage of medical experts in most of these facilities. This results in slow service, and patients end up waiting long hours without receiving any attention. Hence the expert systems paradigm can play a very important role in such cases where medical experts are not readily available. At our research ASU-AIKE labs, Ain Shams University, Egypt, we design and implement of an intelligent medical expert system for diabetes diet that intended to be used in Sudan [38,39]. The expert system provides the patients with medical advices and basic knowledge on diabetes diet. The development of such system went through a number of technical stages, namely; requirements analysis, food knowledge acquisition, formalization, design and implementation. Visual prolog was used for designing the graphical user interface and the implementation of the system. The system is a promising helpful smart tool that reduces the workload for physicians and provides a more comfort for diabetic patients.
MVV brings profound changes to personal health programs and clinical healthcare delivery. It’s seeks to reveal internal structures hidden by the skin and bones, as well as to diagnose and treat disease. Visual representation of the interior body is a key element in medicine. There are many techniques for creating it; such as magnetic resonance imaging, computed tomography, and ultra-sound [40]. The past few decades have witnessed an increasing number of new techniques being developed for practical clinical image display. One approach is to render the data interactively using a specialized multi-processor hardware support. Since these devices are not cheap, they are not widely used in practice. Another alternative is to use volume visualization [41]. In our research labs, we performed a comprehensive study for the recent intelligent techniques and algorithms used for medical data visualization [41]. These techniques cover filtering, segmentation, classification and visualization as well as the intelligent software supporting medical volume visualization. The study reveals hybrid techniques are the best approach for medical image segmentation is the best. In our future, work we are looking to develop mobile based intelligent system using direct volume rendering texture mapping technique with bones data sets.
Currently, we are working on the following applications; (a) Determining the appropriate computational intelligent paradigm and model capable to classify Alzheimer disease and to build computer aided diagnostic system capable of detecting Alzheimer's at any stage [42.43]. (b) Big medical data analytics of intensive care unit data using machine learning and computational intelligence techniques [44,45]. (c) Developing an intelligent medical imaging evaluation system to help doctors diagnose pneumonia caused by the novel corona virus. The system capable of detecting corona at any stage.
The development of robust intelligent medical decision support systems is a very difficult and complex process that raises a lot of technological and research challenges that have to be addressed in an interdisciplinary way. The development of robust intelligent healthcare systems is a very difficult and complex process that raises a lot of technological and research challenges that have to be addressed in an interdisciplinary way. This paper analyzes the main paradigms and applications of the computational intelligence (CI) in healthcare from the artificial intelligence perspective.
The main results based on our analysis. CI offer potentially powerful tools for the development a novel digital healthcare system. The variety of such techniques enabling the design of a robust and efficient IHS. The key to the success of such systems is the selection of the CI technique that best fits the domain knowledge and the problem to be solved. That choice is depending on the experience of the knowledge engineer. And, digital medical decision support systems can benefit from systematic knowledge engineering and structure using techniques from the different disciplines artificial intelligence. AI technologies and techniques play a key role in developing intelligent tools for medical tasks and domains.CI Techniques (e.g. CBR, Data Mining, Rough Set, Ontology) can cope with medical noisy data, sub symbolic data, and complex structure data. In addition, CI offer intelligent computational methods for accumulating, changing and updating medical knowledge in IHS, and in particular learning mechanisms that will help us to induce knowledge from medical information or data. From our comprehensive analysis, one can recommend the following recommendations: 1. The cooperation between physicians and AI communities is essential to produce efficient computing systems for medical purposes. The physicians will have more information to deliver a better service and dynamic guidelines to improve quality and reduce risks. 2. Mobile devices can feed real-time medical data directly to patients and doctors via secure computing networks and IoT. The web based and IoT medical systems can enhance the online education/ learning/training processes. 3. The use of ICT technologies also improves the quality of patient care and reduces clinical risk. At the same time, the patient will be part of the healthcare process, having more information about diseases and access to his/her electronic health record. 4. Public health authorities can get more accurate information and develop dashboards to make better and fast decisions. 5. Hospital management benefits from a more updated meaningful data. This data is used by management systems to delivery KPI (Key Performance Indicators). 29. Salem, A.-B. M., Alfonse, M.: Ontological Engineering Approach for Breast Cancer Knowledge Management, Proceeding of Med-e-Tel The International eHealth, Telemedicine and Health ICT for Education, Networking and Business. Luxembourg. March 30 – April 5 (2009). 30. Salem, A.-B. M., Roushdy, M., HodHod, R. A.: A Case Based Expert System For Supporting Diagnosis Of Heart Diseases. International Journal On Artificial Intelligence and Machine Learning, AIML, Tubungen, Germany, 1, pp.33-39, Dec. (2004). 31. Salem, A.-B. M., Roushdy, M., Mahmoud, S. A.: Mining Patient Data Based on Rough Set Theory To Determine Thrombosis Disease. International Journal On Artificial Intelligence and Machine Learning, AIML, Tubungen, Germany, 1, Dec., pp.27-31 (2004). 32. Salem, A.-B. M., Mahmoud, S. A.: Mining patient Data Based on Rough Set Theory to Determine Thrombosis Disease. In: Proceedings of First Intelligence conference on Intelligent Computing and Information Systems, pp 291-296. ICICIS 2002, Cairo, Egypt, June 24-26 (2002). 33. Salem, A.-B. M., Mahmoud, A. M.: Applying the Genetic Algorithms Approach for Data Mining Classification Task. In: IFIP WG12.6, First IFIP Conference on Artificial Intelligence Applications and Innovations, Toulouse France, August, 22-27 (2004). 34. World Health Organization, http://www.who.int/en/. 35. Timpka, T., Eriksson, H., Gursky, E. A., Nyce, J. M., Morin, M., Jenvald, J., Strömgren, M., Holm, E., Ekberg, J.: Populationbased simulations of influenza pandemics: validity and significance for public health policy. Bull World Health Organ. 87(4), pp. 305-311.
ISSN 0042-9686, (2009). 36. Perez, L., Dragicevic, S. An agent-based approach for modeling dynamics of contagious disease spread. International Journal of Health Geographics, 8:50, (2009). 37. Khalil, K. M., Abdel-Aziz, M., Nazmy, T. T., Salem, A.-B. M.: An Agent-Based Modeling for Pandemic Influenza in Egypt. In: INFOS2010: 7th International Conference on Informatics and Systems, Cairo, Egypt, AOM 24-30, (2010). 38. Ahmed, I. M., Mahmoud A. M.: Development of an Expert System for Diabetic Type-2 Diet. International Journal of Computer Applications, (IJCA), 107(1), pp 13-21, December (2014). 39. Ahmed, I. M., Alfonse, M., Aref, M. M., Salem, A.-B. M.: Daily Meal Planner Expert System for Diabetics Type-2. E-Leader International Journal, 10(2), New York, USA, July (2015). 40. Balazs, C.: Interactive volume-rendering techniques for medical data visualization. Institut fur Computer graphik und Algorithme (2001). 41. Abdallah, Y., Abdelhamid, A.-A., Elarif, T., Salem, A.-B. M.: Intelligent Techniques in
Medical Volume Visualization. Procedia Computer Science 65, 546 – 555 (2015). 42. Kruthika, K. R,et al.: Multistage Classifier-Based Approach for Alzheimer's Disease Prediction and Retrieval. Informatics in Medicine Unlocked, doi: 10.1016/j.imu.12.003 (2018). 43. Solimana, S. A., Mohsen, H., El-Dahshan, S. A., Salem, A.-B. M.: Exploiting of Machine Learning Paradigms in Alzheimer's disease. International Journal of Psychiatry and Psychotherapyhttp://www.iaras.org/iaras/journals/ijpp , ISSN: 2535-0994, 4, pp. 1-11 (2019) 44. Rayan, Z., Alfonse, M., Salem, A.-B. M.: Machine Learning Approaches in Smart Health.
Procedia Computer Science 154, pp. 361–368, ScienceDirect (2019). Available online at www.sciencedirect.com www.elsevier.com/locate/prodedia. 45. Rayan, Z., Alfonse, M., Salem, A.-B. M.: Intensive care unit (icu) data analytics using machine learning techniques. ITHEA® Int. Journal “Information Theories and Applications”, 26(1), pp. 69-82 (2019) http://www.foibg.com/ijita/vol26/ijita-fv26.htm.