Syllabus

Schedule

Course learning modules are divided into weeks. Each week starts on Wednesday at 12:00 am Eastern Time (ET) and closes on Wednesday at 11:59 pm ET, with the exception of Week 8, which ends on Sunday. All assignments must be submitted by 11:59 pm ET on the due date.

Course Schedule at a Glance

Learning Modules	Topics	Assignments and Due Dates
Week 1 2/27 – 3/6	Managing unstructured data in healthcare settings	Introductory Discussion Discussion – Initial post by Sunday 3/3, responses by Wednesday 3/6 Week 1 Assignment – Wednesday 3/6
Week 2 3/6 – 3/13	Advanced machine learning in healthcare	Discussion – Initial post by Sunday 3/10, responses by Wednesday 3/13  Week 2 Assignment – Wednesday 3/13
Week 3 3/13 – 3/20	Exploratory data analysis and visualization for medicare severity diagnosis related group	Discussion -Initial post by Sunday 3/17, responses by Wednesday 3/20  Week 3 Assignment – Wednesday 3/20
Week 4 3/20 – 3/27	Exploratory data analysis and visualization for medicare severity diagnosis related group (continued)	Discussion – Initial post by Sunday 3/24, responses by Wednesday 3/27 Week 4 Assignment – Wednesday 3/27
Week 5 3/27 – 4/3	Understanding comparative effectiveness research	Discussion – Initial post by Sunday 3/31, responses by Wednesday 4/3 Week 5 Assignment – Wednesday 4/3
Week 6 4/3 – 4/10	Using data mining techniques to predict healthcare-associated infections (Part 1)	Discussion – Initial post by Sunday 4/7, responses by Wednesday 4/10 Week 6 Assignment – Wednesday 4/10
Week 7 4/10 – 4/17	Using data mining techniques to predict healthcare-associated infections (Part 2)	Discussion – Initial post by Sunday 4/14, responses by Wednesday 4/17 Week 7 Assignment – Wednesday 4/17
Week 8 4/17 – 4/21	The future of Artificial Intelligence in healthcare	Discussion – Initial post by Friday 4/19, responses by Sunday 4/21

Weekly Course Schedule

Week 1 – Managing unstructured data in healthcare settings

Learning Outcomes

• Describe different types of unstructured data
• Review techniques for extracting unstructured datasets
• Describe text mining, natural language processing and analytics methods used with unstructured datasets
• Distinguish between text and image-based searching

Course Materials

Read:

• Marconi, K., & Lehmann, H. (2014;2015;). Big data and health analytics. Philadelphia, PA: Auerbach Publications. doi:10.1201/b17945
  • Chapter 2

Download:

• Two packages from
  • https://cran.r-project.org/web/packages/SnowballC/index.html
  • https://cran.r-project.org/web/packages/tm/index.html
• Unstructured Dataset

Assignments

Discussion: Introduction

Please use this forum to introduce yourselves to your instructor and your classmates. Briefly share your academic and professional backgrounds. What are you most interested in learning about during this course? What questions do you hope to have answered?

Discussion:

Unstructured data, that is content found in free text fields within the EHR, contain data valuable to healthcare organizations. Natural language processing (NLP) is a valuable tool to mine this data. For this discussion post, investigate the value and potential opportunities for NLP as described within the scientific literature. How can you balance the advantages of unstructured vs. structured data in the application of healthcare analytics? Include at least 2 primary journal sources in your post.

Week 1 Assignment

This exercise will help you to understand how to select, process and identify specific text terms for finding information from large unstructured datasets.

For this week’s assignment:

• Download a file (CSV) with unstructured text data that includes adverse events for the drug warfarin from social media and internet blogs.
• Load and install two packages in R studio (“™” and SnowballC).
• Create a document term matrix of the data.
• From the data, find a list of word terms with 20 or more frequencies and find associate between texts. Remove sparse terms (less than 0.4 sparsity).

Deliverables:

• List of words or terms with 20 or more frequencies (either in a txt file or csv file)

Week 2 – Advanced machine learning in healthcare

Learning Outcomes

• Review the concepts in machine learning with examples
• Explore differences between machine learning, deep learning and artificial intelligence
• Summarize the statistical modelling behind forecasting and prediction modelling
• Assess the pros and cons of using artificial intelligence with Big Data

Course Materials

Read:

• Fogel, A. L., & Kvedar, J. C. (2018). Artificial intelligence powers digital medicine. Npj Digital Medicine, 1(1) doi:10.1038/s41746-017-0012-2
• Beam, A. L., & Kohane, I. S. (2018). Big data and machine learning in health care. Jama, 319(13), 1317-1318. doi:10.1001/jama.2017.18391

Watch:

• TEDx Talks. (May 18, 2016). Artificial Intelligence in Healthcare – It’s about Time | Casey Bennett | TEDxNashville. Retrieved from https://www.youtube.com/watch?v=3LkbUxqGTfo.
• TEDx Talks. (Mar 13, 2018). Using Big Data to Improve Healthcare Services | Tiranee Achalakul | TEDxChiangMai. Retrieved from https://www.youtube.com/watch?v=7t75CNC34vU.

Assignments

Discussion

After reading the two articles on “Artificial Intelligence powers digital medicine” and “Big data and machine learning in Healthcare,” please discuss the following topics

1. Artificial intelligence is a concept that was known since 1960’s. This concept is gaining popularity in healthcare. What are your views on the applications of AI in healthcare? Please also discuss any potential pros and cons regarding using AI in healthcare.
2. What would be some of the limitations to fulfil the requirements of using machine learning using Big data in healthcare?
3. Describe the limitations of using machine learning techniques with healthcare related data? How might you overcome these limitations?

Week 2 Assignment

Write a 2-page (excluding references) paper on the following topic:

Identify and reference two primary journal articles in which an artificial intelligence (AI) method has already been used in some sort of healthcare model or application.

Identify and describe the statistical methods used. Given your knowledge of AI and statistical testing, do the authors’ research methods appear to be sound? Describe in your own words the future opportunities for research and application of AI in healthcare.

Please use APA formatting to cite at least 2 primary journal references (use additional articles, as needed). Use 12-point font, Arial, single spaced.

Deliverables:

• Two-page paper in MS word format

Week 3 – Exploratory data analysis and visualization for medicare severity diagnosis related group

Learning Outcomes

• Identify the economic impacts of MS-DRGs on healthcare delivery
• Summarize the analytical techniques used to evaluate coded healthcare data
• Modify parameters in the data during data preparation stage

Course Materials

Read:

• Marc, D. & Sandefer, R. (2016). Data analytics in healthcare research: tools and strategies. Chicago, Illinois: AHIMA Press.
  • Chapter 6

Download:

• Chapter6_ipps2011.csv 
  Chapter6_ipps2012.csv

Assignments

Discussion

One federal initiative to improve healthcare quality is that payors (insurance companies and CMS) may not reimburse hospitals for the cost of care for some hospital acquired conditions (HACs.)

This policy is controversial, as variation in coding practices at hospitals may lead to differences in the inclusion and position of HACs in the list of codes used for Medicare Severity Diagnosis‐Related Group (MS‐DRG) assignment. Please share your ideas about the pros and cons of this practice and discuss how healthcare organizations might comply to operate in a successful manner.

Week 3 Assignment

This assignment may seem familiar to you; however, you will be doing the work in R, rather than Excel. Your final goal is to revise the column headers in R Studio, rather than Excel, as you did in HIN-770. You will use the two data files Chapter6_ipps2011.csv and Chapter6_ipps2012.csv files. In this module, you will first import the files into R studio console. You will examine different headers and the overall organization for both files. Then, you will use Table 6.1 from the book to change names of headers of the files to process files for data visualization.

Deliverables:

• An R script which contains the code to change the column headers

Week 4 – Exploratory data analysis and visualization for medicare severity diagnosis related group (continued)

Learning Outcomes

• Apply exploratory data analysis and data visualization techniques for presenting aggregate data
• Explore multiple opportunities for visualizing the same dataset
• Identify techniques to explore different methods of visualization for a better understanding of the data

Course Materials

Read:

• Marc, D. & Sandefer, R. (2016). Data analytics in healthcare research: tools and strategies. Chicago, Illinois: AHIMA Press.
  • Chapter 6

Assignments

Discussion

Conduct a simple internet search for several visualization tools and techniques that use R. List and provide names and links for at least two of the tools that you find — one standalone tool, and another that is a plugin/package to R. List the tools that you investigated. Describe the features of the tools you’ve chosen and the rationale for selecting these two tools over the other tools reviewed. As you read your peers’ posts, respond to at least two others by identifying concerns or advantages of using such a tool in your organization.

Week 4 Assignment

In the previous week, you prepared the two datasets. This week, you will use R to explore different data visualizations. Follow the prompts in the book to complete this assignment.

• Install the package “ggplot2”. Use the simple bar chart to explore data of the drg2012_2 data frame.
• Use the zoom function to explore different aspects of the bar chart.
• Use different geoms showing varying heights. This is called the dot plot. Through the commands outlined in the chapter, show average payments from non-medicare sources for non- cervical spinal fusion by state for 2012.
• Use a heatmap to show state-level differences in other MS-DRGs with similar, high non-Medicare payments.

Deliverables:

Screenshots of:

• Bar chart (Step 1)
• Dot plot (Step 2)
• Dot plot showing non-cervical spinal fusion by state for 2012 (Step 3)
• Heatmap showing average non-Medicare payments and average non-Medicare payments (Step 4)

Week 5 – Understanding comparative effectiveness research

Learning Outcomes

• Compare outcomes across hospitals by a variety of organizational characteristics
• Identify predictors of cardiovascular diseases within hospitals, using logistic regression
• Identify and manipulate variables within datasets to process important variables
• Determine the effect of emergency services in the likelihood of a hospital scoring above or below the national average for heart failure 30-day mortality rate (using logistic regression)

Course Materials

Read:

• Marc, D. & Sandefer, R. (2016). Data analytics in healthcare research: tools and strategies. Chicago, Illinois: AHIMA Press.
  • Chapter 9

Download:

• Dataset file: Comparative.csv

Assignments

Discussion

Comparative research is an integral part of hospitals’ operations. Find two examples in the literature that describe how comparative research was used to improve a hospital’s performance. What can institutions who have failing performance do to improve their standing?

Week 5 Assignment

For this week’s assignment, you will use logistic regression in R.

• Use the Comparative.csv file to build a pivot table.
• Follow the instructions in Chapter 9, beginning with Step 3 (p. 140) through “Analyzing the Data in R Studio” (p. 143).
• Interpret your results in a Microsoft Word file (approximately ½ page, double-spaced).

Deliverables:

• Logistic regression output from R Studio (screenshot)
• Interpretation (MS Word file)

Week 6 – Using data mining techniques to predict healthcare-associated infections (Part 1)

Learning Outcomes

• Review the CRISP-DM process for conducting data mining techniques
• Install the data-mining package Rattle for future data mining exercises
• Discuss the challenges of using unstructured data in AI applications

Course Materials

Read:

• Marc, D. & Sandefer, R. (2016). Data analytics in healthcare research: tools and strategies. Chicago, Illinois: AHIMA Press.
  • Chapter 15

Resources:

• Installing Rattle on Windows: https://rattle.togaware.com/rattle-install-mswindows.html
• Installing Rattle on Mac: https://rattle.togaware.com/rattle-install-mac.html
• Explanation of Rattle package: http://eric.univ-lyon2.fr/~ricco/tanagra/fichiers/en_Tanagra_Rattle_Package_for_R.pdf
• miltonmayfield. (Sep 13, 2007). How to install and update packages in R. Retrieved from https://www.youtube.com/watch?v=f1P5jBthaQk
• RStudio and Rattle Installation Assistance: https://bda2020.wordpress.com/r-topics/rstudio-and-rattle-installation-guide/
• Rattle package starter guide: https://cran.r-project.org/web/packages/rattle/vignettes/rattle.pdf

Assignments

Discussion

Healthcare data is sparse and not structured. Critics argue that it is not possible to use healthcare data for Big Data purposes or for Artificial Intelligence. What are the challenges of using unstructured healthcare data in an Artificial Intelligence application? Please support your arguments with published evidence. Here are some resources to help you get started:

• Monga, K., & Singh, H. (2018, Oct 18,). Unstructured data: An important piece of the healthcare puzzle. Retrieved from https://journal.ahima.org/2018/10/18/unstructured-data-an-important-piece-of-the-healthcare-puzzle/
• Pear, Robert. (2018, Mar 30). Artificial Intelligence In Healthcare: Separating Reality From Hype. Forbes. Retrieved from https://www.forbes.com/sites/robertpearl/2018/03/13/artificial-intelligence-in-healthcare/#445628061d75

Week 6 Assignment

Install the “rattle” package. The rattle package installation may be different for MAC and Windows operating systems. Load the package and all other supporting installations that need to be executed to have the package install successfully. Troubleshoot any error messages that prevent a successful installation.

Deliverable:

• A screenshot that documents the successful rattle package installations.

Week 7 – Using data mining techniques to predict healthcare-associated infections (Part 2)

Learning Outcomes

• Construct a decision tree model from publicly available data sets
• Interpret the results of a decision tree model through the CRISP-DM process
• Discuss concerns about the use of AI in healthcare

Course Materials

Read:

• Marc, D. & Sandefer, R. (2016). Data analytics in healthcare research: tools and strategies. Chicago, Illinois: AHIMA Press.
  • Chapter 15

Assignments

Discussion

As a future health informatician, what are your concerns, if any, about the use of AI in healthcare? Base your response on your personal opinion, rather than scholarly evidence. This a chance for you to have a hearty discussion about the future of healthcare.

Week 7 Assignment

Note: For issues with coding, use the search engine of your choice to look for explanations and solutions to the error.

• Follow the instructions in Chapter 15, beginning with Step 3 (p. 248).
• Import the data set [link] into rattle (Step 3, continued)
• Explore the imported dataset using rattle (Step 4)
• Create a graph which shows the correlations among variables (Step 5)
• Build the decision tree underlying the correlations (Step 5, continued)

Deliverables:

• Correlation graphs (screenshot)
  • Note: the graphs may appear in the plot window in R Studio.
• Visual correlation using rattle (screenshot)
• File for decision tree rules (R Script)

Week 8 – The future of Artificial Intelligence in healthcare

Learning Outcomes

• Identify functionalities and opportunities of using Watson with its artificial intelligence based capacity
• Explore different features of Watson within the context of healthcare
• Discuss the challenges and threats that comes with using artificial intelligence

Course Materials

Watch:

• IBM Watson (2019, Jan 2). How Watson works. Retrieved from https://www.youtube.com/playlist?list=PLZDyxLlNKRY9Rqr0xzgTPXqE7DNoDPb-5
• TED. (2015, April 27). What happens when our computers get smarter than we are? | Nick Bostrom. Retrieved from https://www.youtube.com/watch?v=MnT1xgZgkpk

Assignments

Discussion

The Watson videos describe how Watson works within a binary/transactional industry, such as insurance claims. Patient care decision-making is not that orderly. For example, when prescribing medications, one must consider how the new medications interacts with current medications and whether the medication will work for this patient and this symptom.

Based on the video and your readings from this course, do you believe that Watson can be used to address these types of problems? How much credence should we give Nick Bostrom’s concerns about the ethical implications of AI?

No assignment this week

Please note that this week is only 5 days. The class closes on Sunday. No assignments will be accepted after that date.

Student Resources

Online Student Support

Your Student Support Specialist is a resource for you. Please don't hesitate to contact them for assistance, including, but not limited to course planning, current problems or issues in a course, technology concerns, or personal emergencies.

Questions? Visit the Student Support Health Informatics page

APA Style Guide

• https://www.apastyle.org

UNE Libraries:

• Library Access for all students: Your library login ID and password are the same as the ones you use to log into Blackboard/Brightspace.
• MSIN Databases: https://library.une.edu/library-materials/find/research-by-subject/health-informatics/
• Library Tutorial: Find an article
• Library Questions: Ask a librarian or phone library staff at (207) 602-2361 or (207) 221-4330.

UNE Student Academic Success Center

The Student Academic Success Center (SASC) offers a range of services to support your academic achievement, including tutoring, writing support, test prep and studying strategies, learning style consultations, and many online resources. To make an appointment for tutoring, writing support, or a learning specialist consultation, go to une.tutortrac.com. To access our online resources, including links, guides, and video tutorials, please visit:

• UNE Student Academic Success Center

Accommodations

Any student who would like to request, or ask any questions regarding, academic adjustments or accommodations must contact the Student Access Center at (207) 221-4438 or pcstudentaccess@une.edu. Student Access Center staff will evaluate the student's documentation and determine eligibility of accommodation(s) through the Student Access Center registration procedure.

Policies

Technology Requirements

Please review the technical requirements for UNE Online Graduate Programs: Technical Requirements

Turnitin Originality Check and Plagiarism Detection Tool

The College of Professional Studies uses Turnitin to help deter plagiarism and to foster the proper attribution of sources. Turnitin provides comparative reports for submitted assignments that reflect similarities in other written works. This can include, but is not limited to, previously submitted assignments, internet articles, research journals, and academic databases.

Make sure to cite your sources appropriately as well as use your own words in synthesizing information from published literature. Webinars and workshops, included early in your coursework, will help guide best practices in APA citation and academic writing.

You can learn more about Turnitin in the Turnitin Student quick start guide.

Information Technology Services (ITS)

ITS Contact: Toll Free Help Desk 24 hours/7 days per week at 1-877-518-4673

Course Evaluation Policy

Course surveys are one of the most important tools that University of New England uses for evaluating the quality of your education, and for providing meaningful feedback to instructors on their teaching. In order to assure that the feedback is both comprehensive and precise, we need to receive it from each student for each course. Evaluation access is distributed via UNE email at the beginning of the last week of the course.

Attendance Policy

Online students are required to submit a graded assignment/discussion prior to Sunday evening at 11:59 pm ET of the first week of the term. If a student does not submit a posting to the graded assignment/discussion prior to Sunday evening at 11:59 pm ET, the student will be automatically dropped from the course for non-participation. Review the full attendance policy.

Late Policy

Assignments: Late assignments will be accepted up to 3 days late; however, there is a 10% grade reduction (from the total points) for the late submission. After three days the assignment will not be accepted.

Discussion posts: If the initial post is submitted late, but still within the discussion board week, there will be a 10% grade reduction from the total discussion grade (e.g., a 3 point discussion will be reduced by 0.3 points). Any posts submitted after the end of the Discussion Board week will not be graded.

Please make every effort ahead of time to contact your instructor and your student support specialist if you are not able to meet an assignment deadline. Arrangements for extenuating circumstances may be considered by faculty.

Student Handbook Online - Policies and Procedures

The policies contained within this document apply to all students in the College of Graduate and Professional Studies. It is each student's responsibility to know the contents of this handbook.

UNE Online Student Handbook

UNE Course Withdrawal

Please contact your student support specialist if you are considering dropping or withdrawing from a course. The last day to drop for 100% tuition refund is the 2nd day of the course. Financial Aid charges may still apply. Students using Financial Aid should contact the Financial Aid Office prior to withdrawing from a course.

Academic Integrity

The University of New England values academic integrity in all aspects of the educational experience. Academic dishonesty in any form undermines this standard and devalues the original contributions of others. It is the responsibility of all members of the University community to actively uphold the integrity of the academy; failure to act, for any reason, is not acceptable. For information about plagiarism and academic misconduct, please visit UNE Plagiarism Policies.

Academic dishonesty includes, but is not limited to the following:

1. Cheating, copying, or the offering or receiving of unauthorized assistance or information.
2. Fabrication or falsification of data, results, or sources for papers or reports.
3. Action which destroys or alters the work of another student.
4. Multiple submissions of the same paper or report for assignments in more than one course without permission of each instructor.
5. Plagiarism, the appropriation of records, research, materials, ideas, or the language of other persons or writers and the submission of them as one's own.

Charges of academic dishonesty will be reviewed by the Program Director. Penalties for students found responsible for violations may depend upon the seriousness and circumstances of the violation, the degree of premeditation involved, and/or the student’s previous record of violations.  Appeal of a decision may be made to the Dean whose decision will be final.  Student appeals will take place through the grievance process outlined in the student handbook.