Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Artificial intelligence in diabetes care

Artificial intelligence in diabetes care IntroductionMedical artificial intelligence (AI) is moving forward at considerable pace. Promising research ideas are surfacing in clinical waters; AI is automating the national 111 triage service and has exhibited dermatologist‐level performance at identifying suspicious skin lesions, a task where experts frequently disagree . The present article explores how machine learning, a prominent branch of AI, may be set to transform diabetes care.What is machine learning?Machine learning finds relationships within data; for example, a linear regression proposes a ‘line of best fit’ relationship. Such simple representations may struggle, however, to describe more complex data, as shown in Fig. . When clinical guidelines are based on simple models, they might fail to make allowances for the enormous variability found in a person's genome, medical history and their behaviours. By contrast, AI research has developed machine learning techniques that can model very large datasets with multiple, non‐linear relationships. For example, a ‘neural network’ represents data as vast numbers of interconnected neurons, analogous to the human brain; therefore, such models can approach clinical problems in a similar fashion to a clinician by integrating numerous sources of disparate information and providing a personalized solution.(a) and (b) Machine learning reveals important relationships in complex data: an http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Diabetic Medicine Wiley

Artificial intelligence in diabetes care

Buch, V.; Varughese, G.; Maruthappu, M.
Diabetic Medicine , Volume 35 (4) – Jan 1, 2018
Download PDF
3 pages

Loading next page...
1
 
/lp/wiley/artificial-intelligence-in-diabetes-care-xdPCmilSPV

References (11)

Publisher
Wiley
Copyright
Diabetic Medicine © 2018 Diabetes UK
ISSN
0742-3071
eISSN
1464-5491
DOI
10.1111/dme.13587
Publisher site
See Article on Publisher Site

Abstract

IntroductionMedical artificial intelligence (AI) is moving forward at considerable pace. Promising research ideas are surfacing in clinical waters; AI is automating the national 111 triage service and has exhibited dermatologist‐level performance at identifying suspicious skin lesions, a task where experts frequently disagree . The present article explores how machine learning, a prominent branch of AI, may be set to transform diabetes care.What is machine learning?Machine learning finds relationships within data; for example, a linear regression proposes a ‘line of best fit’ relationship. Such simple representations may struggle, however, to describe more complex data, as shown in Fig. . When clinical guidelines are based on simple models, they might fail to make allowances for the enormous variability found in a person's genome, medical history and their behaviours. By contrast, AI research has developed machine learning techniques that can model very large datasets with multiple, non‐linear relationships. For example, a ‘neural network’ represents data as vast numbers of interconnected neurons, analogous to the human brain; therefore, such models can approach clinical problems in a similar fashion to a clinician by integrating numerous sources of disparate information and providing a personalized solution.(a) and (b) Machine learning reveals important relationships in complex data: an

Journal

Diabetic Medicine – Wiley

Published: Jan 1, 2018

There are no references for this article.