Research
 The Tufts Initiative for the Forecasting and Modeling of
Infectious Diseases (InForMID) was established to conduct research
and provide a venue for training in the fields of computational
epidemiology, conservation medicine, biostatistics, and
bioinformatics with the emphasis on public health applications. The
mission of the Initiative is to improve the quality of biomedical
research and health care by developing innovative analytical and
computational tools and systems to life-science researchers, public
health professionals, and policy makers. Methodology
At InForMID we use advanced computational and analytical
techniques that expands on information provided by traditional data
visualization and analysis methods, and open doors for researchers
working to better understand an infectious disease outbreak.
 Spatial Spread:
Geographic
Information Systems and Dynamic Mapping
Understanding the geographic spread of a disease is an important
first step to identifying causal and contributing factors, and
preventing further disease transmission. Geographic Information
Systems (GIS) and Dynamic Maps create easy to interpret visuals
of disease incidence rates across space. We have used maps in
our research of how infectious disease and weather events affect
the elderly and will make extensive use of these tools in our
continued studies of factors affecting elderly health and
outbreaks of waterborne illness.
Temporal Patterns:
Seasonality
Assessment & Distributed Time Lags
Periodic patterns in infectious disease are among the best-known
and worst understood phenomena in the study of disease dynamics. InForMID researchers are on the forefront of research regarding
these seasonal behaviors. Our researchers use new regression
techniques to retain and quantify seasonal fluctuations,
in order to more accurately examine the influence of these
variations on severity and timing of infectious outbreaks over
the years. We have proposed an analytical and conceptual
framework for the assessment of disease seasonality demonstrated
by seasonal patterns of enterically transmitted and respiratory diseases. These
methods are combined with a new time-series analysis
tool-Temporal Exposure Response Surface (TERS) Data
Visualization. This new technique offers a three-dimensional
picture of disease spread revealing magnitude, duration and
shape of the epidemic curve of an infectious outbreak in
association with the level of exposure. We demonstrated the use
of TERS-plots as an advanced visualization tool for syndromic
surveillance systems. It also allowed us to detect a secondary
spread of cryptosporidiosis infection, as well as shorter
average incubation time, in the elderly compared to general
population.
We introduced the Distributed Lag Model, an approach for
modeling time-distributed effects with respect to an incubation
period of infection through an analysis of the association
between ambient temperatures and enterically-transmitted
infections.
 Improved Surveillance:
Outbreak and Disease Signature
We've used mathematical modeling to describe the concept of
‘outbreak signature', as an improved tool for surveillance and
early detection of infectious disease. Building on this work, we
used mathematical models to create a hypothetical outbreak
scenario and defined the idea of a population-specific ‘disease
signature' combining all three elements of disease dynamics-the
temporal and spatial spread of outbreaks in and across
populations. A disease signature allows for a higher level of
differentiation between disease-spread scenarios in different
populations, leading to a new practical definition for
‘outbreak'. This model can be used to understand, estimate, and,
in some cases, correct for, reporting error inherent in
traditional disease surveillance.
|
