Page 213 - Handbook of Deep Learning in Biomedical Engineering Techniques and Applications

Chapter 7 Early detection and diagnosis using deep learning 203

2.2 Challenges and solutions
The exhilarating potential of AI [11] in the medical sector has
been extensively conveyed, with possible claims that crossway
many diverse areas of medication [12,13]. This potential has
been saluted as Medicare systems worldwide that fight to bring
the better efficiency profoundly working on cultivating better un-
derstanding of care by totally focusing on refining the well-being
of inhabitants, tumbling per capita costs of healthcare [14], and
completely focusing on improvement of the workelife balance
of different people as this is the main reason of diseases to enter
and stay.
Nonetheless, the power that AI holds is not yet acknowledged
by the healthcare sector till date, which can be proved with the
help of the existing clinical information and by the price assis-
tants that have ascended from actual practice of AI algorithms
in the healthcare sector. This topic not only highlights all the
recent challenges and limitations that the healthcare sector is
facing with the so much hyped technology called AI but also
focuses on how to transform the limitations that are faced by
the industry to find valid solutions that can help researchers in
making a difference in the systems. The number of researches
is going on everyday in the field of AI that can help in different
applications of medical sector in which different algorithms are
applied for different solutions such as cancer, diabetes, skin dis-
eases, chest radiographs [15e18], and retinal images [19,20] which
can be easily tested and predicted. Not only diseases but also
genomics construal can be done via AI.

2.2.1 Retroactive versus forthcoming trainings
Though prevailing trainings have incorporated huge records of
patients with widespread criterion in contradiction of proficient

Figure 7.5 Cluster analysis.