A Generalized Overview of the Biomedical Image Processing From the Big Data Perspective

Introduction

Computer aided Diagnostic systems have changed the face of the biomedical data analysis. In general manual investigations are sometimes error prone and time consuming due to the inherent limitations of the human experts (Robb & Hanson, 1990; R. A. Shaikh, Li, Khan, & Memon, 2016). Therefore computer aided Diagnostic methods are very useful to detect various diseases accurately and within the stipulated amount of time (Doi, 2007). Accuracy and time is very precious from the perspective of the diagnostic industry because it is directly associated with the health of the patients. There is no provision to compromise with the accuracy and the precision of the obtained results because it can leads to the wrong treatment which can be harmful for the patients (Doi, 2007; Kohn & Furuie, 1991). Computer aided diagnostic systems have changed the face of the biomedical data analysis. In general manual investigations are sometimes error prone and time consuming due to the inherent limitations of the human experts (Shah et al., 2018). Therefore computer aided Diagnostic methods are very useful to detect various diseases accurately and within the stipulated amount of time. Accuracy and time is very precious from the perspective of the diagnostic industry because it is directly associated with the health of the patients. There is no provision to compromise with the accuracy and the precision of the obtained results because it can leads to the wrong treatment which can be harmful for the patients (Bischof et al., 1999; Dolatabadi, Khadem, & Asl, 2017).

Manual investigation of the biomedical data generally takes considerable amount of time. Moreover various hidden patterns are may not be discoverable by the human experts all the time because of the inherent limitations of the humans (Endsley & Kiris, 1995; Podgurski et al., 2003). In general amount of data which is generated in the medical industry is quite huge and may be very difficult to be processed by the digital systems with limited resources (Breton, Medina, & Montagnat, 2003; Lim, De Heras Ciechomski, Sarni, & Thalmann, 2003). Generally the type of data is not homogeneous in nature. The speed off the data generation is quite high. To handle all these constraints, some efficient data processing methods are required to effectively process data in real time. Big data analytics is one of the major advancements in the field of data science that helps to process data in real time. Real time data processing also demands intelligent algorithms that can efficiently understand the acquired data, remove noises and interpret the data by considering the inconsistencies which can be present sometime (Ilyasova, Kupriyanov, Paringer, & Kirsh, 2018; Luo, Wu, Gopukumar, & Zhao, 2016; Nair & Ganesh, 2016; Neshatpour et al., 2016; Tchagna Kouanou et al., 2018). Big data handling methods are therefore required to overcome all these barriers. Big Data Analytics has several applications including market research, healthcare, agriculture, weather prediction, satellite data analysis etc (Murdoch & Detsky, 2013; Sin & Muthu, 2015).