Demystification of Deep Learning-Driven Medical Image Processing and Its Impact on Future Biomedical Applications

Demystification of Deep Learning-Driven Medical Image Processing and Its Impact on Future Biomedical Applications

R. Udendhran (Department of Computer Science and Engineering, Bharathidasan University, India) and Balamurugan M. (Department of Computer Science and Engineering, Bharathidasan University, India)
DOI: 10.4018/978-1-7998-3591-2.ch010
OnDemand PDF Download:
No Current Special Offers


The recent growth of big data has ushered in a new era of deep learning algorithms in every sphere of technological advance, including medicine, as well as in medical imaging, particularly radiology. However, the recent achievements of deep learning, in particular biomedical applications, have, to some extent, masked decades-long developments in computational technology for medical image analysis. The methods of multi-modality medical imaging have been implemented in clinical as well as research studies. Due to the reason that multi-modal image analysis and deep learning algorithms have seen fast development and provide certain benefits to biomedical applications, this chapter presents the importance of deep learning-driven medical imaging applications, future advancements, and techniques to enhance biomedical applications by employing deep learning.
Chapter Preview


Researchers calm that deep learning, Quantum Computing and Internet of Things will revolutionize the world similar the way electricity did a century ago. This chapter presents the important opportunities as well as challenges experienced in medical image applications. Generally, biomedical imaging and healthcare industry works under the rule of doctor-patient confidentiality, however, this becomes a challenge for biomedical industry with the integration of deep Learning, for instance:

  • Will the data be safe after entering into the system.

  • What will happen to the patients’ profile and data?

  • What factors contributes to the accountability and integrity of automated decision making of deep learning driven image interpretation and the machines utilization of data?

Figure 1.

An example of Multi-modality for 3-D image of the brain which presents multi-information


Information about the same patient will be generated and entered into the database through various tools in own proprietary data formats are employed. In this case, if any flawed user input are entered and imperfect database design leads to data inconsistency and redundancy (Würfl et al (2018)).

In order to preclude data redundancy, inconsistency problems and less expenditure to create effective database structures before they are deployed. But in some cases, database which does not have effective data structures that can suffer from these problems, a process of database normalization should be implemented. The purpose of database normalization is to re-modify tables such a way that the relations among them are logical, so that database is scalable without any anomalies and avoid data redundancy, inconsistency problems and less expenditure. It is recommendable to design the database in the OLTP format which are highly normalized which avoids data duplication errors. In such cases, healthcare server and computers can employ Linux operating system which is a powerful multi-user operating system which allows several users to access it simultaneously. Linux precludes any changes when employed in the mainframes as well as servers, however, it needs to address the need for security since an attacker can corrupt confidential data, in order to solve this challenge Linux has classified the authorization into ownership and permissions. The reasons for the network failure and other problems in healthcare information systems maybe due to the IP Address and Network Card issues since more than one computers allocated to the same IP address in the data center could cause these kinds of network problems, generally, network card links computers and problems may arise from network card, another reason can be weak radio signals in parts of locations. Drop-in Internet connections can be considered as another potential problem for network issues and the firewall settings must be checked. Before you start the troubleshooting, we need to check if all the hardware are switched on and working well. And make sure the router is not switched off and all switches are in correct positions.By proper power cycling the modems, routers and systems can provide a solution for solving simple network problems. For example, if the administrator encounters network failure in healthcare database server, type “ipconfig” in the command prompt in the terminal. Now check if computer’s IP address starts with 169, this means the system will not receive a valid IP address. You can solve this problem by typing “ipconfig /release” followed by “ipconfig /renew” to get a new valid IP address. By using the command “nslookup” we can perform a DNS check. Review database logs. Check all the database logs are working properly sometimes the database maybe full or malfunctioning, it could be reason for the problems that flow on and affect your network performance.

In health and insurance,a central semantic store approach can be deployed which concentrates on logging as well as storing all the rules employed by the database integration process in a single centralized repository. The reason for this approach is that data sources are updated and new ones which that included do not fall outside data integration rules as shown in figure 2.

Complete Chapter List

Search this Book: