Towards Clinically Relevant Explanations for Type-2 Diabetes Risk Prediction
Abstract.
With the increased need for AI models to be more explainable, there have been efforts to apply these techniques in high-precision domains like healthcare. From our past user studies and interaction with medical experts, we find that they appreciate explanations to help them understand model predictions better. To this end, we have begun to explore techniques to generate user-centered explanations that are contextualized by authoritative literature from clinical practice guidelines and backed by AI model predictions and post-hoc explanations. We present and discuss our methodology and initial results for this approach for a type-2 diabetes risk prediction use case.
1. Introduction
ACM’s consolidated article template, introduced in 2017, provides a consistent LaTeX style for use across ACM publications, and incorporates accessibility and metadata-extraction functionality necessary for future Digital Library endeavors. Numerous ACM and SIG-specific LaTeX templates have been examined, and their unique features incorporated into this single new template.
If you are new to publishing with ACM, this document is a valuable guide to the process of preparing your work for publication. If you have published with ACM before, this document provides insight and instruction into more recent changes to the article template.
The “acmart” document class can be used to prepare articles
for any ACM publication — conference or journal, and for any stage
of publication, from review to final “camera-ready” copy, to the
author’s own version, with very few changes to the source.
2. Background
As noted in the introduction, the “acmart” document class can
be used to prepare many different kinds of documentation — a
double-blind initial submission of a full-length technical paper, a
two-page SIGGRAPH Emerging Technologies abstract, a “camera-ready”
journal article, a SIGCHI Extended Abstract, and more — all by
selecting the appropriate template style and template parameters.
This document will explain the major features of the document class. For further information, the LaTeX User’s Guide is available from https://www.acm.org/publications/proceedings-template.
2.1. Use Case
The primary parameter given to the “acmart” document class is
the template style which corresponds to the kind of publication
or SIG publishing the work. This parameter is enclosed in square
brackets and is a part of the documentclass command:
\documentclass[STYLE]{acmart}
Journals use one of three template styles. All but three ACM journals
use the acmsmall template style:
-
•
acmsmall: The default journal template style. -
•
acmlarge: Used by JOCCH and TAP. -
•
acmtog: Used by TOG.
The majority of conference proceedings documentation will use the acmconf template style.
-
•
acmconf: The default proceedings template style. -
•
sigchi: Used for SIGCHI conference articles. -
•
sigchi-a: Used for SIGCHI “Extended Abstract” articles. -
•
sigplan: Used for SIGPLAN conference articles.
2.2. Datasets
In addition to specifying the template style to be used in formatting your work, there are a number of template parameters which modify some part of the applied template style. A complete list of these parameters can be found in the LaTeX User’s Guide.
Frequently-used parameters, or combinations of parameters, include:
-
•
anonymous,review: Suitable for a “double-blind” conference submission. Anonymizes the work and includes line numbers. Use with the\acmSubmissionIDcommand to print the submission’s unique ID on each page of the work. -
•
authorversion: Produces a version of the work suitable for posting by the author. -
•
screen: Produces colored hyperlinks.
This document uses the following string as the first command in the source file:
\documentclass[sigconf,authordraft]{acmart}
3. Methods
Modifying the template — including but not limited to: adjusting
margins, typeface sizes, line spacing, paragraph and list definitions,
and the use of the \vspace command to manually adjust the
vertical spacing between elements of your work — is not allowed.
Your document will be returned to you for revision if modifications are discovered.
3.1. Risk Prediction Model
3.2. Post-hoc Explainers
3.3. Guideline Contextualization
4. Results and Evaluation
The “acmart” document class requires the use of the
“Libertine” typeface family. Your TeX installation should include
this set of packages. Please do not substitute other typefaces. The
“lmodern” and “ltimes” packages should not be used,
as they will override the built-in typeface families.
4.1. Initial Results
4.2. Expert Feedback
5. Conclusion
The title of your work should use capital letters appropriately -
https://capitalizemytitle.com/ has useful rules for
capitalization. Use the title command to define the title of
your work. If your work has a subtitle, define it with the
subtitle command. Do not insert line breaks in your title.
If your title is lengthy, you must define a short version to be used
in the page headers, to prevent overlapping text. The title
command has a “short title” parameter:
\title[short title]{full title}
6. Acknowledgments
This work is supported by IBM Research AI through the AI Horizons Network.
7. Appendices
If your work needs an appendix, add it before the
“\end{document}” command at the conclusion of your source
document.
Start the appendix with the “appendix” command:
\appendix
and note that in the appendix, sections are lettered, not numbered. This document has two appendices, demonstrating the section and subsection identification method.
8. SIGCHI Extended Abstracts
The “sigchi-a” template style (available only in LaTeX and
not in Word) produces a landscape-orientation formatted article, with
a wide left margin. Three environments are available for use with the
“sigchi-a” template style, and produce formatted output in
the margin:
-
•
sidebar: Place formatted text in the margin. -
•
marginfigure: Place a figure in the margin. -
•
margintable: Place a table in the margin.