How to develop machine learning models for healthcare

  title={How to develop machine learning models for healthcare},
  author={Po-Hsuan Cameron Chen and Yun Liu and Lily H. Peng},
  journal={Nature Materials},
Rapid progress in machine learning is enabling opportunities for improved clinical decision support. Importantly, however, developing, validating and implementing machine learning models for healthcare entail some particular considerations to increase the chances of eventually improving patient care. 
Prediction of Diabetes Risk based on Machine Learning Techniques
Various research outcomes of machine learning methodologies used in diabetes mellitus are analyzed and how the efficiencies obtained could be helpful in future perspective of a predictive diabetes model designing.
Machine Learning for Clinical Predictive Analytics
In this chapter, a brief overview of applying machine learning techniques for clinical prediction tasks is provided, and some of the most common machine learning algorithms for prediction problems are outlined.
Use of deep learning to develop continuous-risk models for adverse event prediction from electronic health records.
This protocol can be used by interdisciplinary teams with programming and clinical expertise to build deep-learning prediction models with alternate data sources and prediction tasks to predict various clinical and operational outcomes from structured electronic health record (EHR) data.
Machine Learning Implementation in Clinical Anesthesia: Opportunities and Challenges.
As pioneers of patient safety, now is the time to consider how anesthesiologists should interact with, define the authors' relationship to, and guide implementation of novel ML systems, including clinical anesthesia.
Machine Learning Models for Inpatient Glucose Prediction
The aim was to review the clinical evidence for the role of machine learning–based models in predicting hospitalized patients’ glucose trajectory and develop and validate accurate machine learning algorithms for prediction of hypoglycemia and hyperglycemia in the hospital.
Insights from teaching artificial intelligence to medical students in Canada
Clinical artificial intelligence (AI) applications are rapidly developing but existing medical school curricula provide limited teaching covering this area. Here we describe an AI training curriculum
Representation Learning for Electronic Health Records
This chapter introduces the background of learning representations and reasons why the authors need good EHR representations in machine learning for medicine and healthcare, and explains the commonly-used machine learning and evaluation methods for representation learning using a deep learning approach.


Scalable and accurate deep learning with electronic health records
A representation of patients’ entire raw EHR records based on the Fast Healthcare Interoperability Resources (FHIR) format is proposed, and it is demonstrated that deep learning methods using this representation are capable of accurately predicting multiple medical events from multiple centers without site-specific data harmonization.
Discrimination and Calibration of Clinical Prediction Models: Users’ Guides to the Medical Literature
This Users’ Guide will help clinicians understand the available metrics for assessing discrimination, calibration, and the relative performance of different prediction models to help clinicians make optimal use of existing prediction models.
Understanding and using sensitivity, specificity and predictive values
The basic knowledge to calculate sensitivity, specificity, positive predictive value and negative predictive value is discussed and how to use these measures in day-to-day clinical practice is provided.
Eyeing cardiovascular risk factors
Smoking status, blood pressure, age and other cardiovascular risk factors can be predicted from retinal images by using deep learning.
Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support: 4th International Workshop, DLMIA 2018, and 8th International Workshop, ML-CDS 2018, Held in Conjunction with MICCAI 2018, Granada, Spain, September 20, 2018, Proceedings
This paper presents UNet++, a new, more powerful architecture for medical image segmentation where the encoder and decoder sub-networks are connected through a series of nested, dense skip pathways, and argues that the optimizer would deal with an easier learning task when the feature maps from the decoder and encoder networks are semantically similar.
Prediction of cardiovascular risk factors from retinal fundus photographs via deep learning
Deep learning predicts, from retinal images, cardiovascular risk factors—such as smoking status, blood pressure and age—not previously thought to be present or quantifiable in these images.
Clinically applicable deep learning for diagnosis and referral in retinal disease
A novel deep learning architecture performs device-independent tissue segmentation of clinical 3D retinal images followed by separate diagnostic classification that meets or exceeds human expert clinical diagnoses of retinal disease.
Pivotal trial of an autonomous AI-based diagnostic system for detection of diabetic retinopathy in primary care offices
FDA authorized the system for use by health care providers to detect more than mild DR and diabetic macular edema, making it, the first FDA authorized autonomous AI diagnostic system in any field of medicine, with the potential to help prevent vision loss in thousands of people with diabetes annually.