-
Notifications
You must be signed in to change notification settings - Fork 2
/
main.tex
156 lines (130 loc) · 6.29 KB
/
main.tex
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
\documentclass[10pt]{article}
\renewcommand{\sfdefault}{lmss}
\usepackage{helvet} % PSNFSS Font, in every TeX distribution
%\usepackage{avant} % PSNFSS Font, in every TeX distribution
% % % Extended \ssfamily (Sans) fonts; load extraFonts option from TeX
%\usepackage[scaled=0.88]{berasans}% package has a handy scaling option
%\usepackage{libris} % a nice, almost handwritten calligraphic look
%\usepackage{biolinum} % included with the {Libertine} font package
%\usepackage{iwona}
\usepackage{paratype}
\usepackage{amsmath}
\usepackage{textcase}
\usepackage{fullpage, setspace, parskip, titlesec}
\usepackage[section]{placeins}
\usepackage[dvipsnames]{xcolor}
\usepackage{breakcites, lineno, hyphenat, rotating, etoolbox}
\usepackage{ graphicx}
\usepackage{latexsym, textcomp, longtable, tabulary, tabularx}
\usepackage{booktabs,array,multirow}
\usepackage{amsfonts,amsmath,amssymb}
\usepackage[utf8]{inputenc}
\usepackage[english]{babel}
\usepackage{xspace, caption}
\usepackage[toc,symbols]{glossaries}
\usepackage{bm}
\usepackage[automake]{glossaries-extra}
\usepackage[round]{natbib}
\usepackage[left=0.75in]{geometry}
\usepackage[ampersand]{easylist}
\usepackage{enumitem}
\usepackage{pifont, makecell}
\usepackage[disable]{todonotes}
\usepackage{float}
\usepackage{lscape}
\usepackage{tcolorbox}
\usepackage{titlesec}
\usepackage{latexcolors}
\tcbuselibrary{most}
\usepackage{epigraph}
\usepackage{svg}
\usepackage{multicol}
\PassOptionsToPackage{hyphens}{url}
\usepackage[colorlinks]{hyperref}
\usepackage{url}
% Counter for information panels
\newcounter{panel}
\newenvironment{panel}[1][]{\refstepcounter{panel}\par\medskip
\textbf{Panel~\thepanel. #1} \rmfamily}{\medskip}
\def\panelautorefname{Panel}
% Colors for references and links
\definecolor{darkblue}{RGB}{0,0,130}
\definecolor{darkgreen}{RGB}{0,80,20}
\hypersetup{colorlinks = true, linkcolor = darkblue, citecolor = darkgreen, urlcolor= RedViolet}
%Enable using math mode with the package acronym
\DeclareRobustCommand{\mL}{\mathcal{L}}
\glsenablehyper
% Orcid logo and links for authors
\newcommand{\orcid}[1]{\href{https://orcid.org/#1}{}}
% Use a dash in bullet points, not an ugly dot
\renewcommand\labelitemi{-}
% Graphics extensions
\AtBeginDocument{\DeclareGraphicsExtensions{.png,.PNG,.jpg,.JPG,.jpeg,.JPEG}}
% Document style changes
\renewenvironment{abstract}
{{\noindent\bfseries{\abstractname}\par\nobreak}\footnotesize}
{\bigskip}
\titlespacing{\section}{0pt}{*3}{*1}
\titlespacing{\subsection}{0pt}{*2}{*0.5}
\titlespacing{\subsubsection}{0pt}{*1.5}{0pt}
\setlength\parindent{24pt}
%Bibliography and glossary
\bibliographystyle{unsrtnat}
\providecommand\citet{\cite}
\providecommand\citep{\cite}
\providecommand\citealt{\cite}
\renewcommand\cite{\citep}
\newglossary[odg]{oalgo}{old}{odn}{\gls{ohd}-GAN Acronyms}
\setabbreviationstyle[acronym]{long-short}
\glsxtraddallcrossrefs
\makeglossaries
\include{glossary_definitions}
% Document header
\title{ Synthetic Observational Health Data with GANs: from slow adoption to a boom in medical research and ultimately digital twins?}
\author{
Georges-Filteau, Jeremy \href{https://orcid.org/0000-0002-0352-6468}{\includegraphics[width=10pt]{assets/orcid.png}}\\[0.2cm]
\small \textit{Radboud University, The Hyve}\\
\small\texttt{jeremy@thehyve.nl}
\and
Leonard Wee \href{https://orcid.org/0000-0003-1612-9055}{\includegraphics[width=10pt]{assets/orcid.png}}\\[0.2cm]
\small \textit{Maastricht Universitair Medisch Centrum}\\
\small\texttt{leonard.wee@maastro.nl}
\and
Dekker, Andre \href{https://http://orcid.org/0000-0002-0422-7996}{\includegraphics[width=10pt]{assets/orcid.png}}\\[0.2cm]
\small \textit{Maastricht Universitair Medisch Centrum,}\\
\small \textit{ Maastricht University, MAASTRO Clinic}\\
\small \texttt{andre.dekker@maastro.nl}
\and
Cirillo, Elisa \href{https://orcid.org/0000-0002-0241-7833}{\includegraphics[width=10pt]{assets/orcid.png}}\\[0.2cm]
\small \textit{The Hyve}\\
\small \texttt{elisa@thehyve.nl}
}
\begin{document}
\maketitle
\vspace{-1em}
\begingroup
\let\center\flushleft
\let\endcenter\endflushleft
\maketitle
\endgroup
\selectlanguage{english}
\glsresetall
\begin{abstract}
After being collected for patient care, \gls{ohd} can further benefit patient well-being by sustaining the development of \gls{hi} and medical research.A vast potential is unexploited because of the private nature of patient-related data and regulations to protect it.\\
\glspl{gan} have recently emerged as a groundbreaking way to train generative models that produce realistic synthetic data. They have revolutionized practices in multiple domains such as self-driving cars, fraud detection, digital twin simulations in industrial sectors, and in medical imaging.\\
The digital twin concept could readily apply to modelling and quantifying disease progression. In addition, \glspl{gan} poses many capabilities relevant to common problems in healthcare: lack of data, class imbalance, rare diseases, and preserving privacy. Unlocking open access to privacy-preserving \gls{ohd} could be transformative for scientific research. In the midst of COVID-19, the healthcare system is facing unprecedented challenges, many of which of are data related for the reasons stated above.\\
Considering these facts, publications concerning GAN applied to \gls{ohd} seemed to be severely lacking. To uncover the reasons for this slow adoption, we broadly reviewed the published literature on the subject. Our findings show that the properties of \gls{ohd} were initially challenging for the existing \gls{gan} algorithms (unlike medical imaging, for which state-of-the-art model were directly transferable) and the evaluation synthetic data lacked clear metrics.\\
We find more publications on the subject than expected, starting slowly in 2017, and since then at an increasing rate. The difficulties of \gls{ohd} remain, and we discuss issues relating to evaluation, consistency, benchmarking, data modelling, and reproducibility.
\end{abstract}
\include{1-introduction}
\include{2-methods}
\include{3-results}
\include{4-discussion}
\include{5-sourcecode}
\include{6-conclusion}
\pagebreak
\printglossary[type=oalgo]
\printglossary[type=\acronymtype]
\pagebreak
\bibliography{biblio}
\end{document}