Deep learning model for predicting and detecting overlapping symptoms of cardiovascular diseases in hospitals of UAE
Deep learning (DL) is a subdomain of machine learning (ML) representing exponentially growing potential in the field of medicine, helping to classify information, new diseases, phenotyping, and intricate decision-making. The DL algorithm and technique of an ML domain is often meditated by a variety...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Karadeniz Technical University
2012
|
| Subjects: | |
| Online Access: | http://irep.iium.edu.my/94421/7/94421_Deep%20learning%20model%20for%20predicting%20and%20detecting%20overlapping.pdf http://irep.iium.edu.my/94421/ https://www.turcomat.org/index.php/turkbilmat/article/view/11580/8501 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Deep learning (DL) is a subdomain of machine learning (ML) representing exponentially growing potential in the field of medicine, helping to classify information, new diseases,
phenotyping, and intricate decision-making. The DL algorithm and technique of an ML domain is often meditated by a variety of neural networks (NN). DL module has been augmented by ongoing developments in computer tools as well as techniques. The use of this learning technique has been increased in various domains such as e-commerce, banking and finance, as well as for speech and feature recognition to learn and classify intricate information. There is no medical literature on the strengths and weaknesses of DL. DL strengths comprises of its potential to automatically diagnose clinical appearances, enhance decision-making, recognizing the phenotypes, and effectively select treatment methodology to complex diseases. The DL algorithm that can be well matched to chemistry is the hemodynamic and electrophysiological parameters catalogues that are effectively captured over continuous periods by wearable
machines, as well as the division of images into captured images or pictures (Nagueh, 2016).
However, DL has number of weaknesses as well, including difficulty in interpreting its examples (the 'black-box' criticism), its requirement for multiple training data, no specificity in design, no data-usefulness in training, limited use in experimental models, and so on. Hence, the best
clinical applications of DL require considerate problem solving solution, selection of the most suitable DL algorithms and information, and defining balance of outcome. This review updates the existing state of DL for cardiac physicians and researchers and gives diverse professions to escalate the pitfalls, close challenges, and opportunities for the currently available new area. |
|---|