Page 38 - Handbook of Deep Learning in Biomedical Engineering Techniques and Applications

26 Chapter 2 Deep convolutional neural network in medical image processing

In recent times, DL has gained popularity in different research
domains such as computer vision, speech processing, and natural
language processing. This technique is best suited in those fields
where a huge volume of information has to be evaluated and hu-
manlike intelligence is to be applied. As shown in Fig. 2.1, the
amount of data in the medical field is rapidly increasing, and so
the use of DL as an ML tool is becoming a significant part in the
domain of medical image processing. This is very apparent from
the recent discussion on DL outline and future prospective [2]
in which the primary effect of DL in the field of biomedical imag-
ing is explored. DL has been rated among the top advances of 2013
according to an MIT review on technologies [3]. As a diagnostic
technique, medical image analysis has played a vital role for a
long time. Multiple recent advances in safety procedures, hard-
ware design, data storage abilities, and computational resources
have largely added to the work on clinical image processing. In
addition to medical diagnosis, other applications such as abnor-
mality detection, image classification, and segmentation have
also used DL techniques in the present time [2].
The main aim of medical image processing is to assist profes-
sionals and experts to carry out the disease diagnosis and treat-
ment procedure in an efficient way. Computer-aided diagnosis
and computer-aided detection depend on the efficient clinical
imaging process, hence making it important in terms of perfor-
mance as it will straightly affect the step of medical diagnosis
and treatment [4]. In this regard, it necessitates different unique

Figure 2.1 Plot showing the drastic rise in the healthcare data [1].