10.3.1 Detecting Diseases and Diagnosis

One of the dominant machine learning usage and applications in the healthcare industry is the detection and analysis of diseases, infections, and diseases. Various diseases, such as tumors, are rough to detect during the early stages, dangerous genetic diseases and infections caused by pathogens, etc. IBM Watson Genomics is a key illustration of how incorporating perceptive computing with genome structure and cancer sequencing can aid in constructing a fast identification and diagnosis.

10.3.2 Drug Detection and Manufacturing

One of the main scientific usage and applications of machine learning in healthcare is the early-stage drug detection process. Research and advance technologies, such as advanced sequencing and precision medicine, can support in discovering alternative procedures for the remedy of multifactorial illnesses. Machine learning practices implicate unsupervised-based learning, which can recognize outlines in data to proceed toward next-generation exploration and advancements. Microsoft developed Hanover by using machine learning-based technologies for several ingenuity-based AI systems for cancer conduct and personalizing medication amalgamation for severe myeloid leukemia.

10.3.3 Medical Imaging Analysis and Diagnosis

With machine learning, deep learning, and computer vision technologies, Microsoft developed InnerEye for image analytics and everything on image diagnostic tools. Medical image detection deals with the difficulty of detecting a certain portion/element in a medical image. As medical images are volumetric, efficient parsing is observed through AI-driven diagnostic methods.

10.3.4 Personalized/Adapted Medicine

Personalized/adapted conducts can not only be further operative by combining specific health with prophetic analytics but are also suitable are for advanced research and improved disease assessment/treatment. IBM has developed Watson Oncology to understand patient medical history to aid generate several treatment opportunities.

10.3.5 Behavioral Modification

Behavioral modification is a crucial measure of protective medicine, and machine learning-based models are researching the arenas of cancer inhibition and identification that allows us to understand our comatose conduct and make crucial modifications.

10.3.6 Maintenance of Smart Health Data

With the advent of machine learning models such as document/image classification, OCR recognition techniques and smart health records in healthcare a lot of time, energy, and money is saved. Google’s Cloud Vision API is using machine learning-based handwriting recognition technology to develop the next generation smart and intelligent, health record maintenance system.

10.3.7 Clinical Trial and Study

Machine learning has numerous potential applications and importance in the area of clinical trials and research. Using ML-based predictive analytics, we can classify prospective clinical trial observations, and it can aid researchers to pull a pool from an extensive variety of data sources, such as preceding doctor appointments, social media, and other related vital information.

10.3.8 Crowdsourced Information Discovery

Crowdsourcing is altogether the indignation in the health field these days, letting researchers and medical practitioners contact a vast volume of information given by individuals based on their interests. Apple’s ResearchKit permits users to access collaborative apps, which apply machine learning constructed facial recognition to identify and treat Autism and Parkinson’s illness. IBM newly aligned with Medtronic to decipher, collect, and deliver real-time diabetes data founded on the crowdsourced data. With the progressions being made in the Internet of things, the healthcare industry is still identifying new and novel ways to handle tough-to-diagnose scenarios and support the overall progress of analysis, diagnosis, and medication.

10.3.9 Enhanced Radiotherapy

One of the ultimately preferred applications of machine learning in healthcare exists in the field of radiology. Medical image analysis has several distinct variables, which are interpreted with complex equations based on trained machine learning-based models. These models study from the huge training set of different available samples to easily diagnose and find the diseases. Google’s DeepMind Health is allowing efficient classification of objects into various categories, such as normal/abnormal, lesion/non-lesion, usual/unusual, etc. to develop efficient machine learning algorithms.

10.3.10 Outbreak/Epidemic Prediction

Artificial intelligence-based technologies are nowadays being set to use in observing and predicting outbreaks/epidemics around the world. Data scientists have access to a huge amount of data gathered from satellites, social media, websites and clinics. Artificial neural networks aid to collate this data, predict and forecast everything from malaria epidemics to SARS-CoV-2. Predicting and forecasting these outbreaks/epidemics is especially useful in third-world nations as they have a deficiency in essential medical infrastructure and informative systems. ProMED-mail is a cloud-based epidemic reporting platform that observes evolving diseases, developing scenarios of epidemics and delivers outbreak intelligence in real time [19–23].

10.4.1 Role of Edge in Reshaping Healthcare

Edge computing and artificial intelligence promise to create healthcare provision cheaper, well, easier, and improved for everyone. Currently, infectious diseases and epidemics are on the rise. Uncertainties about COVID-19 disease have instigated people to put off regular examinations and treatments, constructing chronic situations worse and guaranteeing an additional 20% rise in additional mortality. According to the report of the World Health Organization, we are now facing a universal deficiency of 7 million healthcare workers and this number could reach more than 13 million by 2040.

To meet several demands and undergo reforming transformation in healthcare, the edge is providing instant solutions as shown in Figure 10.2. Currently, edge computing strategies are being used to control and monitor patients remotely, systematize the automation to deliver healthcare, take along artificial intelligence to advance the speed and precision of diagnoses, track the medicine/vaccine supply chain, etc. That is because edge AI puts data/information/statistics, analytics/decision system, and processing/implementation control where it is most required such as in the ambulance, hospital, operating room, homes, and host bodies.

10.4.2 How AI Powers the Edge

We cannot discourse about edge computing-powered healthcare without also discoursing about artificial intelligence. It is not sufficient to gather data from smart devices, patients, smart vehicles, caregivers, etc., it is also essential to examine it and retort in real time. Progressively, that practice is being accomplished by smart machines using edge technology. According to the survey of IoT security firm Zingbox, the average hospital bed takes 10 to 15 linked devices. Similarly, a survey by Optum—the technology division of UnitedHealth Group tells that 40% of healthcare administrators plan to install AI to study data collected by the IoT devices to monitor the health or diseases or condition of a patient. In more contemporary times, AI has an extent throughout the healthcare industry with its digital landscape. From being on mainframes, workstations and computers, PDAs to the cloud computing and now to the “edge computing.” It is this final part that is specifically enthralling. Edge AI is, therefore, well thought-out to be the ensuing chapter in the progression of AI. Conferring to the newest market research testimony of MarketsandMarkets the edge technology will be worth more than $15 billion in the coming 5 years. The prevalent assumption of healthcare IoT devices is the key reason that is motivating its demand. It benefits healthcare administrators to get in trace with their patients in real time [24–28].

10.5.1 Rural Medicine

Providing value healthcare to secluded rural areas has been a great challenge in past. Even today, with discoveries in telemedicine and additional easily accessible health data, medical providers have resisted providing fast, excellent attention to people who live far from hospitals and observe inadequate net access. Normal healthcare databases challenge substantial experiments here along with their connectivity difficulties. But the edge computing applications overcome all these difficulties with their novel approaches. IoT-based healthcare devices can extend the grasp of existing networks, permitting the medical workforce to contact critical patient data even in areas with insufficient connectivity. This use case on Edge AI, showing its capacity to significantly spread the best of health services.

10.5.2 Autonomous Monitoring of Hospital Rooms— A Case Study

One of the USPs of AI, in all-purpose, is prospective to automate things. AI processes can gather data using a congregation of sensors and investigate it to retort most suitably. Edge AI proceeds this up a notch. Computer vision, data analytics, and artificial intelligence models enable autonomous observing of hospital rooms, infrastructure, and patients. Take for example fall discovery. Many IoT-enabled wearable devices these days arise with the capacity to sense if a person falls unexpectedly with specific hardware. Processes of edge AI in these wearable devices can be competent to sense falls in an immediate and even communicate to caregivers. In maximum cases, this is treated as a life-saving option. Ex: Fall detection option available on Apple Watch.

Edge AI can also be of immense support in the observing of important signs. Medical devices intended to record significant data like heart rate, body temperature, breathing rate, blood sugar, blood pressure, hypertension, etc. can influence AI to notice any deviation in an instant. These devices could then alert the hospital staff/caregivers and they can take necessary action. For the patient, this is not only precarious but also progresses the overall treatment and medication [29–31].

10.6.1 Fuzzy Logic—Outline

Nevertheless, the enormous improvement and increasing the elderly people implies more resources for aftercare, paramedical care and natural assistance in their habits. The fuzzy logic was introduced with the Lotfi Zadeh’s proposal of fuzzy in 1965. The term fuzzy refers to the things that are not clear and or are vague, which provides very valuable flexibility for reasoning, that resembles human reasoning in many ways, and hence we can consider uncertainties and inaccuracies of any situation. The term Crisp value means a precise can be referred to a Boolean logic system. In fuzzy logic, there are intermediate values present that are partially true and partially false. Fuzzy sets can be considered as an extension of classical sets. It swallows partial membership, which means that it contains elements that have varying degrees of membership in the set. Classical set contains elements that satisfy precise properties of membership which fuzzy set contains elements thatsatisfy imprecise properties of membership.

A fuzzy set in the universe U can be defined as a set of ordered pairs and it can be represented mathematically as

= Membership of Y in  and

The way of representing a Fuzzy set as:

 is the contribution or importance of y. Fuzzy logic being a computing approach on “degree of truth,” rather than conventional Boolean logic classification “true or false” (1 or 0) the one modern computing requirement is needed. The applications of fuzzy logic are numerous. In real world scenarios like weather forecasting business decision making, an impeccable role in healthcare support systems and medicinal making artificial intelligence and neural network which are in trend today, fuzzy logic can be used vibrantly used in all these fields and gaining more popularity. The reason why fuzzy logic applies so well in real life is because in most of the scenarios, we cannot have a distinct value, rather based on an approximation. For example, the no. of red blood cells in your blood can not be actually counted but we can have an estimated value, or the growth of bacteria in a particular environment can only be predicted, but cannot be counted and have the correct value as well. Fuzzy logic works on the logic based on Membership and degree of membership which ranges from 0 to 1. Depending on the membership, function depends on different membership values ranging from 0 to 1, which will decide the impact of that particular characteristic nature or behavior of a member.

10.6.2 Fuzzy Logic-Based Smart Healthcare

A single disease may be symptomatic in different forms depending on the patient’s health conditions and inheritance. In the same way, a single symptom may correspond to different diseases. This description of diseases entitles in linguistic terms that are not exact and may be vague. To handle this inaccuracy and uncertainty, we need an efficient mechanism. Medical diagnosis is fundamentally a pattern grouping phenomenon. Depending on some input delivered by a patient, a professional provides inference based on its information and knowledge, which is usually warehoused in a binary form, and lastly, the outcome is considered, i.e., either the infected person is suffering from a definite disease or not. There are various properties and facilities in fuzzy set theory, and due to which, it is made suitable for medical diagnosis. The result and discovery from the study have revealed that the artificial intelligence-based technique with fuzzy logic can add a reliable outcome to advise the disease [32–34].

10.6.3 Medical Diagnosis Using Fuzzy Logic for Decision Support Systems

In this chapter, medical diagnosis support system will be discussed which is generated using fuzzy logic that focuses on the medical support system. With the advent of fuzzy logic to the model physician diagnosis decisions, it will have much more accuracy than the conventional decisions that are being considered earlier. Some of the fuzzy algorithmic approaches for the decision support systems are multicriteria decision methods and aggregation approaches using the fuzzy approach.

First of all, we should consider a basic normative model for the decision theory: for the set of alternatives with feasible options, X, the set of states relevant to the context of the experiment, the resulting scenarios E, and utility function u that order the spaces of events with respect the desirability. The highest yields of utility will lead to the decision concerning uncertainty and possibility. The feasible alternatives may be explicitly defined with implicit enumerable constraints. The developing model can be extended to work on different objectives. The tradeoff is the decision making point toward the other objective. The real-world scenarios may have the add-on vague semistructured situations that may arise like:

• ✓ The alternative cannot be determined crisp, though it can be gained to a certain degree of membership,
• ✓ The situation may not be represented mathematically, but can be explained and represented by vague ideas and linguistic terminologies.
• ✓ The utility depends on several categorical decisions and risk comportment
• ✓ The functional dependencies are unable to depict.

We restrict our discussion to one major facility of a multiattribute decision model (MADAM). In turn, to provide the comprehension of the fuzzy model’s area, let us understand the possible degree of desirability. The aggregation approach will be usually consisting of two phases.

• ✓ Phase 1: It is of the judgment among all goals and decision alternatives.
• ✓ Phase 2: It is the ranking of the decision alternatives to the summarized judgments.

A multiattribute decision-making system usually consists of a set of alternatives. These alternatives will be treated as variables and will be compared with respect to multiple, and using conflict attributes.

The aim to determine the ranking with respect to all the attributes in a given scenario is based on classical utility theory and performance with respect to each attribute and its importance against the objectives. The problem must be assessed before aggregating the subjective assessments together to determine the overall performance of each among all attributes in the situation. The overview of these processes is shown in Figure 10.3. After identifying the alternatives and attributes for the multiattribute decision making model, the pairwise comparison and assessment of the performance of those attributes will be computed and then the computation of alternative performance using fuzzy expert analysis will take place. Along with that we also compute the alternative importance of the fuzzy expert analysis. The consecutive processes will take place to the final multiattribute alternatives discrete ranking after deriving the fuzzy utilities of alternatives. In this process, the degree of dominance of the utilities and alternatives will be used for decision making for computing the raking.

10.6.4 Applications of Fuzzy Logic in Healthcare

The complicated medical procedures result from the lack of proper and accurate information, imprecise information and contradicting nature raising the advanced developments in diagnostic procedures and treatments. To address these issues, fuzzy logic can alone be used in a hybrid manner with other methods like neuro-fuzzy and fuzzy-NN and fuzzy Bayesian applications. It has a wide scope of applications in Medical Research. Ranking of risk factors and test results and their performance is important for decision making process. Clustering is used for the split of data (referred to as clusters) to derive the hidden patterns for the objects with similar properties and different behaviors. Medical Image processing has a vital role in medical decision making. 2-Dim and d3-Dim images are the main sources of magnetic resonance image processing, digital mammography and positron, emission tomography and topography tests. The images will generally be textures with noise and the acquisition of inaccuracies in the expected portions. In these areas, using the clustering techniques is more useful to make the precise decision with imprecise conditions and vagueness [35–37].

1. 1. Yu, K.H., Beam, A.L., Kohane, I.S., Artificial intelligence in healthcare. Nat. Biomed. Eng., 2, 719–731, 2018.
2. 2. Jiang, F., Jiang, Y., Zhi, H. et al., Artificial intelligence in healthcare: Past, present and future. Stroke Vasc. Neurol., 2, 4, 230–243, 2017.
3. 3. Davenport, T. and Kalakota, R., The potential for artificial intelligence in healthcare. Future Healthc. J., 6, 2, 94–98, 2019.
4. 4. Amin, S.U. and Hossain, M.S., Edge intelligence and internet of things in healthcare: A survey. IEEE Access, 9, 45–59, 2021.
5. 5. Keshavarzi, A. and van den Hoek, W., Edge intelligence—On the challenging road to a trillion smart connected IoT devices. IEEE Des.Test, 36, 2, 41–64, 2019.
6. 6. Liu, Y., Peng, M., Shou, G., Chen, Y., Chen, S., Toward edge intelligence: Multiaccess edge computing for 5G and internet of things. IEEE Internet Things J., 7, 8, 6722–6747, 2020.
7. 7. Du, Y., Wang, Z., Leung, V.C.M., Blockchain-enabled edge intelligence for IoT: Background, emerging trends and open issues. Future Internet, 13, 248, 1–21, 2021.
8. 8. Rausch, T. and Dustdar, S., Edge intelligence: The convergence of humans, things, and AI. 2019 IEEE International Conference on Cloud Engineering (IC2E), pp. 86–96, 2019.
9. 9. Satpathy, S., Mangla, M., Mohanty, S.N., Potluri, S., GA-based iterative optimization system to supervise adaptive workflows in cloud environment, in: Emerging Technologies in Data Mining and Information Security. Advances in Intelligent Systems and Computing, vol. 1300, A.E. Hassanien, S. Bhattacharyya, S. Chakrabati, A. Bhattacharya, S. Dutta (Eds.), Springer, Singapore, 2021.
10. 10. Potluri, S., Mangla, M., Satpathy, S., Mohanty, S.N., Detection and prevention mechanisms for DDoS attack in cloud computing environment. 2020 11th International Conference on Computing, Communication and Networking Technologies (ICCCNT), pp. 1–6, 2020.
11. 11. Malladi, A. and Potluri, S., A Study on technologies in cloud-based design and manufacturing. Int. J. Mech. Prod. Eng. Res. Dev., 8, 187–192, 2018.
12. 12. Gunturi, M., Kotha, H., Potluri, S., An Iot based solution for health monitoring using a body-worn sensor enabled device. J. Adv. Res. Dyn. Control Syst., 10, 646–651, 2018.
13. 13. Potluri, S. and Avinash, M., Health record data analysis using wireless wearable technology device. J. Adv. Res. Dyn. Control Syst. (JARDCS), 10, 9, 696– 701, 2018.
14. 14. Ravikrishna, B. and Singh, H.P., Big data classification using hybrid parallel Lfr-Cm at cloud environment. IJAST, 29, 08, 6469, May 2020.
15. 15. Potluri, S., Sunaina, S., Neha, P., Govind, C., Raghavender, J., Gupta, V.M., A secure cloud infrastructure towards smart healthcare: IoT based health monitoring. Cloud Secur. Technique. Appl., 63, 63–82, 2021.
16. 16. Mohanty, S.N., Mohammad, G.B., Potluri, S., Ramya, P., Lavanya, P., Next generation cloud security: State of the art machine learning model. Cloud Secur. Technique. Appl., 125, 125–144, 2021.
17. 17. Potluri, S., Sarkar, A., Yasin, E.T., Mohanty, S.N., IoT enabled cloud based healthcare system using fog computing: A case study. J. Crit. Rev., 7, 6, 1068– 1072, 2020.
18. 18. Mohanty, S.N., Potluri, S., Prakash, V.B., Srinath, B., Manjunath, B., Cloud security concepts, threats and solutions: Artificial intelligence based Approach. Cloud Secur. Technique. Appl., 1, 1–20, 2021.
19. 19. Bhardwaj, R., Nambiar, A.R., Dutta, D., A study of machine learning in healthcare. 2017 IEEE 41st Annual Computer Software and Applications Conference (COMPSAC), pp. 236–241, 2017.
20. 20. Callahan, A. and Shah, N.H., Chapter 19 - Machine learning in healthcare, in: Advances in Clinical Informatics, A. Sheikh, K.M. Cresswell, Wright, D.W. Bates (Eds.), pp. 279–291, Academic Press, Canada, 2017.
21. 21. Siddique, S. and Chow, J.C.L., Machine learning in healthcare communication. Encyclopedia, 1, 1, 220–239, 2021.
22. 22. Waring, J., Lindvall, C., Umeton, R., Automated machine learning: Review of the state-of-the-art and opportunities for healthcare. Artif. Intell. Med., 104, 1–12, 2020.
23. 23. de la Torre, J., Marin, J., Ilarri, S., Marin, J.J., Applying machine learning for healthcare: A case study on cervical pain assessment with motion capture. Appl. Sci., 10, 17, 1–28, 5942, 2020.
24. 24. Ray, P.P., Dash, D., De, D., Edge computing for Internet of Things: A survey, e-healthcare case study and future direction. J. Netw. Comput. Appl., 140, 1–22, 2019.
25. 25. Li, X., Huang, X., Li, C., Yu, R., Shu, L., EdgeCare: Leveraging edge computing for collaborative data management in mobile healthcare systems. IEEE Access, 7, 22011–22025, 2019.
26. 26. Pustokhina, I.V., Pustokhin, D.A., Gupta, D., Khanna, A., Shankar, K., Nguyen, G.N., An effective training scheme for deep neural network in edge computing enabled internet of medical things (IoMT) systems. IEEE Access, 8, 107112–107123, 2020.
27. 27. Rahman, M.A. and Hossain, M.S., An internet of medical things-enabled edge computing framework for tackling COVID-19. IEEE Internet Things J., 8, 21, 15847–15854, 2021.
28. 28. Klonoff, D.C., Fog Computing and Edge Computing architectures for processing data from diabetes devices connected to the medical Internet of Things. J. Diabetes Sci. Technol., 11, 4, 647–652, 2017.
29. 29. Hassan, N., Gillani, S., Ahmed, E., Yaqoob, I., Imran, M., The role of edge computing in internet of things. IEEE Commun. Mag., 56, 110–115, 11, 2018.
30. 30. Alshehri, F. and Muhammad, G., A comprehensive survey of the internet of things (IoT) and AI-based smart healthcare. IEEE Access, 9, 3660–3678, 2021.
31. 31. Sun, L., Jiang, X., Ren, H., Guo, Y., Edge-cloud computing and artificial intelligence in internet of medical things: Architecture, technology and application. IEEE Access, 8, 101079–101092, 2020.
32. 32. Rizwan, A., Zoha, A., Mabrouk, I.B., Sabbour, H.M., Al-Sumaiti, A.S., Alomainy, A., Imran, M.A., Abbasi, Q.H., A review on the state of the art in atrial fibrillation detection enabled by machine learning. IEEE Rev. Biomed. Eng., 14, 219–239, 2021.
33. 33. Sellam, V., Kannan, N., Basha, H.A., An effective fuzzy logic based clustering scheme for edge-computing based internet of medical things systems, in: Cognitive Internet of Medical Things for Smart Healthcare. Studies in Systems, Decision and Control, vol. 311, A.E. Hassanien, A. Khamparia, D. Gupta, K. Shankar, A. Slowik (Eds.), Springer, Cham, 2021.
34. 34. Kakkar, A. and Farshori, A., Collaborative Medical Inventory Resources Using Edge Computing – A Solution to Serve Critical Healthcare Requirements at Public Hospitals. Amity International Conference on Artificial Intelligence (AICAI), pp. 647–652, 2019.
35. 35. Mulimani, M.S. and Rachh, R.R., Edge computing in healthcare systems, in: Deep Learning and Edge Computing Solutions for High Performance Computing. EAI/Springer Innovations in Communication and Computing, A. Suresh and S. Paiva (Eds.), Springer, Cham, 2021.
36. 36. Tsai, Y.T. and Lin, Z.Y., “A Survey Edge Computing Bioinf. Health Informatics*,” 2020 IEEE Int. Conf. Bioinf. Biomedicine (BIBM), 2020, pp. 2203-2208, 2020.
37. 37. Kumar, M.G.S., Dhulipala, V.R.S., Baskar, S., Fuzzy unordered rule induction algorithm based classification for reliable communication using wearable computing devices in healthcare. J. Ambient Intell. Hum. Comput., 12, 3515–3526, Mar 2021.