TY - CHAP
T1 - Deep learning in biomedical informatics
AU - Hung, Che Lun
N1 - Publisher Copyright:
© 2023 Elsevier Inc. All rights reserved.
PY - 2022/1/1
Y1 - 2022/1/1
N2 - With the massive influx of multimodal data in the last decade, the role of data analytics in health informatics has grown rapidly. Deep learning (DL) is defined as a technology based on artificial neural networks (ANNs), which have recently emerged as a powerful ML tool. The rapid increase in computing power, fast data storage, and parallelization, as well as predictive capabilities and the ability to generate automatically optimized advanced functions and semantic interpretation from input data, all contribute to the rapid adoption of this technology. This chapter introduces the latest developments in the use of deep learning in health informatics and makes an important analysis of the relative advantages, potential shortcomings, and future prospects of the technology. This chapter mainly focuses on the key applications of DL in the fields of computational biology, drug design, medical imaging, pervasive sensing, medical informatics, and public health. Artificial intelligence (AI) has become a trend in recent years, opening up an entirely new era of research in various fields. The demand for AI in healthcare has increased in both the academia and industry; therefore, the potential benefits of its applications have been proven. Previous studies have attempted to implement AI methods on medical images, electronic health records, molecular characteristics, and a variety of lifestyles. Researchers used data aggregated from multiple data sources to train models that mimic what clinicians do when they see patients and help in decision-making through results and interpretations. It included how to read clinical images, predict results, discover the connection between genotype and phenotype or phenotype and disease, analyze treatment responses, and track lesions or structural changes (e.g., hippocampal volume reduction). In addition, predictive research (e.g., disease or readmission predictions) and correlation and pattern recognition research have been extended to early warning systems with risk scores and overall pattern research and population care (such as predictive care for an entire population).
AB - With the massive influx of multimodal data in the last decade, the role of data analytics in health informatics has grown rapidly. Deep learning (DL) is defined as a technology based on artificial neural networks (ANNs), which have recently emerged as a powerful ML tool. The rapid increase in computing power, fast data storage, and parallelization, as well as predictive capabilities and the ability to generate automatically optimized advanced functions and semantic interpretation from input data, all contribute to the rapid adoption of this technology. This chapter introduces the latest developments in the use of deep learning in health informatics and makes an important analysis of the relative advantages, potential shortcomings, and future prospects of the technology. This chapter mainly focuses on the key applications of DL in the fields of computational biology, drug design, medical imaging, pervasive sensing, medical informatics, and public health. Artificial intelligence (AI) has become a trend in recent years, opening up an entirely new era of research in various fields. The demand for AI in healthcare has increased in both the academia and industry; therefore, the potential benefits of its applications have been proven. Previous studies have attempted to implement AI methods on medical images, electronic health records, molecular characteristics, and a variety of lifestyles. Researchers used data aggregated from multiple data sources to train models that mimic what clinicians do when they see patients and help in decision-making through results and interpretations. It included how to read clinical images, predict results, discover the connection between genotype and phenotype or phenotype and disease, analyze treatment responses, and track lesions or structural changes (e.g., hippocampal volume reduction). In addition, predictive research (e.g., disease or readmission predictions) and correlation and pattern recognition research have been extended to early warning systems with risk scores and overall pattern research and population care (such as predictive care for an entire population).
KW - Bioinformatics
KW - Convolutional neural networks
KW - Deep learning
KW - Drug discovery
KW - Health informatics
KW - Machine learning
KW - Medical image analysis
KW - Medical Informatics
UR - http://www.scopus.com/inward/record.url?scp=85151207608&partnerID=8YFLogxK
U2 - 10.1016/B978-0-323-85796-3.00011-1
DO - 10.1016/B978-0-323-85796-3.00011-1
M3 - Chapter
AN - SCOPUS:85151207608
SN - 9780323901413
SP - 307
EP - 329
BT - Intelligent Nanotechnology
PB - Elsevier
ER -