![Show Menu](styles/mobile-menu.png)
![Page Background](./../common/page-substrates/page0493.png)
INFORMS Nashville – 2016
491
WE23
108-MCC
Surveillance of Diseases with Big Data
Sponsored: Health Applications
Sponsored Session
Chair: Qingpeng Zhang, City University of Hong Kong,
83 Tat Chee Ave, Kowloon Tong, TX, 00000, Hong Kong,
qingpeng.zhang@cityu.edu.hk1 - Patient-specific Depression Monitoring By Selective Sensing
Ying Lin, University of Washington,
linyeliana.ie@gmail.com,
Shuai Huang, Shan Liu
Development of personalized health surveillance is enabled by sensing and
information technologies. Scaling it up to large scale depression population needs
a seamless combination of data analysis and sensing strategy design. We
developed a selective sensing method to capture individual depression
progressions from acquired health data and optimally allocate sensing resources to
high-risk individuals by exploiting the similarities of their progression trajectories.
The proposed method can lead to efficient and cost-effective monitoring of
depression population.
2 - An Optimization Approach To Concussion Management
Gian Gabriel Garcia, University of Michigan,
garciagg@umich.edu,
Mariel Sofia Lavieri
We apply data-driven optimization to improve concussion diagnosis for athletes
suspected of concussion and apply dynamic programming to solve the sequential
decision-making problem of optimal return-to-play management for athletes who
have concussions.
3 - Biosurveillance Of Climate Sensitive Mosquito-borne Diseases
Using Online Social Media
Kusha Nezafati, University of Texas, Dallas, TX,
United States,
Kusha.Nezafati@utdallas.edu, Yulia Gel,
L. Leticia Ramirez Ramirez
Chikungunya is a mosquito-borne virus that is transmitted by the same type of
mosquito as dengue and zika. Chikungunya is relatively well documented in Asia,
Africa, and the Indian subcontinent. In 2014 the first confirmed case of
chikungunya virus has been reported in the Americas, and since then its spread
has been attracting a lot of attention from the health care professionals. However,
the data on chikungunya still remain relatively scarce which makes forecasting of
its epidemiological curve a very challenging task. In this talk we discuss predictive
utility and limitations of online social media, particularly, Google trend, as a
proxy for unavailable data on chikungunya.
4 - Semantic Social Network Analysis Of Online Health Communities
Ronghua Xu, City University of Hong Kong,
ronghuaxu2-c@my.cityu.edu.hk,Qingpeng Zhang
The understanding of how people use online health communities/groups (OHCs)
to discuss health-related topics is critical to the the effective use of social media to
provide social support. In this research, we collected a comprehensive dataset of
mental health related OHC in China, and developed a set of methods to model
and analyze the information spread and semantic patterns of users’ discussions.
The results unveiled the unique topological features and semantic patterns of
mental health related OHCs, with managerial insights of how to utilize OHCs to
provide social support to complement conventional offline approaches.
WE24
109-MCC
Operations Research in Healthcare Management
in Chile
Sponsored: Health Applications
Sponsored Session
Chair: Jorge Vera, Professor, Pontificia Universidad Catolica de Chile,
Vicuna Mackenna 4860, Macul, Santiago, 7820436, Chile,
jvera@ing.puc.cl1 - Physiotherapy Treatment Appointments Scheduling Using
An MDP-based System
Sergio Maturana, Pontificia Universidad Catolica de Chile,
smaturan@ing.puc.cl,Ignacio Lazo
Scheduling physiotherapy treatment appointments in a hospital faced with a very
high demand is complex. The current system in a Chilean hospital results in many
patients waiting long times before their treatments can begin. This hospital has
three types of specialists: a physiatrist and two types of therapist. Before the
therapy can begin, patients must see the physiatrist, who indicates the
appropriate treatment. We propose a scheduling system, based on a Markov
Decision Process (MDP), which determines how to assign the patients to the
physiatrist and how to distribute the patient’s sessions within a planning horizon
in order to reduce waiting times and assure that sessions are evenly spread.
2 - Chemotherapy Treatment Scheduling In Public Health
Juan-Carlos Ferrer, Professor, Pontificia Universidad
Catolica de Chile, Santiago, Chile,
jferrer@ing.puc.clThis research addresses a real scheduling problem for chemotherapy patients at a
Chilean public Hospital. We divide the problem into two subproblems, scheduling
patients on an infinite time horizon and daily patient scheduling. The benefits of
both stages are evaluated for a real case in the Hospital s Chemotherapy Unit
using simulation in the first stage and solving the model to optimality in the
second one. We evaluate potential opportunities for efficiency through a
sensitivity analysis of key resources.
3 - A Hierarchical Solution Approach For Bed Capacity Planning
Under Uncertainty In The Healthcare Service Industry
Ana Celeste Batista, PhD Candidate, Pontificia Unversidad
Catolica de Chile, Santiago, Chile,
abatista@uc.cl, Jorge Vera
Effective capacity planning under uncertainty ensures robust and consistent plans
in time. The health sector is a service system of great relevance to consider better
methods for inter-temporal decisions, since poor planning affects directly the
welfare of people. This work presents a hierarchical multistage model applied to
beds planning. We propose a solution method based on the formulation and
solved using stochastic optimization. The problem is to determine the availability
of beds that minimizes overall patient welfare loss as a function of waiting time.
From the solution we propose policies allowing better decisions in different
planning horizons.
WE25
110A-MCC
Logistics V
Contributed Session
Chair: Tayo Fabusuyi, University of Michigan and Carnegie Mellon
University, 5520 Baywood Street, Floor #3, Pittsburgh, PA, 15206,
United States,
Fabusuyi@umich.edu1 - Minimizing Customer Waiting Time In A Unit-load Warehouse
Mahmut Tutam, PhD Student, UARK, 1359 N Leverett Avenue,
Apt 31, Fayetteville, AR, 72703, United States,
mtutam@uark.edu,John A White
Although the number of delivery trucks arriving at a unit-load warehouse per
unit time may well be Poisson distributed, the time required to perform rectilinear
round-trip travel between the dock and a uniformly distributed point in the
storage region is not exponentially distributed. However, it can be approximated
using a k-Erlang distribution. Using the Method of Moments, the value of k is
determined. The analysis includes one or more docks distributed along one wall
of a rectangular-shaped warehouse. The dimensions of the warehouse that
minimize customer waiting time are determined for a given storage area.
2 - The Mode Most Traveled: Parking Implications And
Policy Responses
Tayo Fabusuyi, University of Michigan and Carnegie Mellon
University, 5520 Baywood Street, Floor #3, Pittsburgh, PA, 15206,
United States,
Fabusuyi@umich.edu,Robert Hampshire,
Zhen Qian
Driving to work alone continues to be the travel mode most utilized by
commuters. Using the US Census public use microdata sample (PUMS) dataset
from the Pacific region of the U.S., we examine why this trend persists by
generating travel mode profiles for representative individuals. In addition to the
profiles, we examine the marginal effects on travel mode choice for selected
explanatory variables at their representative values. The empirical exercise
provides insights on the influence policy measures may have in shaping
individuals’ travel mode preferences and how this will impact on demand for
parking spaces.
WE25