Online Course – Certified Professional Internship in Informed Clinical Decision Making Using Google Deep Learning.

Learn how to apply deep learning to electronic health records. Discover the path from data mining in clinical databases to clinical decision support systems.

Suggested by: Coursera (What is Coursera?)

Start your Coursera.com free trial today
קורס אונליין - התמחות מקצועית מוסמכת בעשיית החלטות קליניות מושכלות באמצעות למידת עומק של Google.

Professional Certificate

Intermediate level

No prior knowledge required

Time to complete the course

7-day free trial

No unnecessary risks

Skills you will acquire in the course

  • Data mining from clinical databases
  • Deep Learning in Electronic Medical Records
  • Explainable deep learning models for healthcare applications
  • Clinical decision support systems
  • EHR preprocessing
  • Building deep learning models
  • General explanations of models
  • Local explanations of models

What you will learn in the course

Courses for which the course is suitable

  • Clinical Data Analyst
  • Deep learning model developer
  • Healthcare Software Engineer
  • Clinical Decision Support Systems Specialist
  • Researcher in the field of artificial intelligence in healthcare
  • Medical Information Systems Analyst
  • Medical Algorithm Developer
  • Medical data mining expert
  • Developer of technological solutions for healthcare systems
  • Medical Predictive Analyst

Internship – Series of 5 courses

This internship is designed for learners with programming experience who are interested in expanding their skills in applying deep learning to electronic medical records, with an emphasis on how to translate their models into clinical decision support systems.

Main topics

  • Data mining from clinical databases: ethics, the MIMIC III database, the International Classification of Diseases System, and definition of common clinical outcomes.
  • Deep learning in electronic medical records: from descriptive to predictive analyses.
  • Explainable deep learning models for healthcare applications: what they are and why they are needed.
  • Clinical decision support systems: generalization, information bias, ‘fairness’, clinical utility and privacy of artificial intelligence algorithms.

Hands-on Learning Project

Learners have the opportunity to select and perform an exercise based on datasets extracted from MIMIC-III, which combines knowledge from:

  • Mining information from clinical databases to query the MIMIC database.
  • Deep learning in electronic medical records for EHR preprocessing and building deep learning models.
  • Deep learning models are explicable to explain the models’ decision.

Options to choose from

  • Feature Importance in Replacement on the MIMIC Dataset for Critical Care: The technique is applied to both logistic regression and LSTM models. The resulting explanations are general explanations of the model.
  • LIME on the MIMIC dataset for critical care: The technique is applied to both logistic regression and LSTM models. The resulting explanations are local explanations of the model.
  • Grad-CAM on the MIMIC dataset for critical care: GradCam is applied to an LSTM model predicting mortality. The resulting explanations are local explanations of the model.

Details of the courses that make up the specialization

Data mining from clinical databases – CDSS 1

Course 1
20 hours
4.8 (13 ratings)

What will you learn?

  • Understand the schema of publicly accessible EHR databases (MIMIC-III)
  • Become familiar with the use of the International Classification of Diseases (ICD)
  • Extract and visualize theoretical statistics from clinical databases
  • Understand and extract key clinical outcomes such as mortality and length of stay

Skills you will acquire

  • Data mining from clinical databases
  • Electronic medical records
  • Theoretical statistics
  • Ethics in electronic medical records
  • International Classification of Diseases

Deep Learning in Electronic Medical Records – CDSS 2

Course 2
31 hours

What will you learn?

  • Train deep learning architectures such as multilayer perceptrons, convolutional neural networks, and recurrent neural networks for classification
  • Validate and compare different machine learning algorithms
  • Deploy electronic health records and represent them as time series data
  • Data subtraction and coding strategies

Skills you will acquire

  • Global and local explanations
  • Explainable machine learning models
  • Attention mechanisms
  • Interpretation vs. Explanation
  • Neutral models and specific models

Explainable Models for Deep Learning in Healthcare – CDSS 3

Course 3
30 hours
4.6 (15 ratings)

What will you learn?

  • Encode global explanatory methodologies into time series classification categories
  • Encode local explanation methodologies for deep learning such as CAM and GRAD-CAM
  • Understand the proof methods for deep learning networks
  • Integrate attention into recurrent neural networks and visualize attention weights

Skills you will acquire

  • Recurrent neural network
  • Convolutional neural network
  • Data encoding and autoencoders
  • EHR and deduction process
  • Deep learning and verification

Clinical Decision Support Systems – CDSS 4

Course 4
8 hours

What will you learn?

  • Evaluation of clinical decision support systems
  • Bias, Calibration, and Fairness in Machine Learning Models
  • Decision Curve Analysis and Audience-Centered Clinical Support Systems
  • Privacy issues in clinical decision support systems

Skills you will acquire

  • Privacy issues in clinical support systems
  • Distortion and fairness in machine learning models
  • Calibration in machine learning models
  • Clinical support systems
  • Audience-focused clinical support systems

Rooftop Mission – CDSS 5

Course 5
2 hours

What will you learn?

This course is a capstone assignment that requires you to apply the knowledge and skills you have learned throughout the internship. In this course, you will choose one of the areas and complete the assignment to pass.

Related products