3 PhD / postdoc positions in the DIACON project:
Control Algorithms for Semi- and Fully-Automated Pen and Pump Insulin
Administration

3 PhD/postdooc positions are available. The positions will be staffed as
soon as qualified candidates are available – preferably no later than
January 2009.

DIACON (Diabetes Control) is a research project that concerns development
of “Control Algorithms for Semi- and Fully-Automated Pen and Pump Insulin
Administration” to patients with type I diabetes. The project is funded by
NABIIT - the Danish Strategic Research council. The partners in the
project are DTU Informatics, DTU-Biosys, Hvidovre Hospital, and Novo
Nordisk A/S. In addition we collaborate with University of California at
Santa Barbara and Uppsala University.

In this project we develop and test new technologies for treatment of
diabetes. The objective of the project is to construct systems that can
regulate and stabilize blood glucose levels in patients with type I
diabetes. These systems are variations of the so-called Artificial
Pancreas. The Artificial Pancreas consists of a subcutaneous glucose
sensor, a control algorithm, and an insulin pump. The sensor measures
glucose levels under the skin. This signal is used by the control
algorithm to compute the correct amount of insulin to inject in order to
keep the blood glucose at its target level.  The computed amount is
injected using the insulin pump. Glucose sensors and insulin pumps are
already available. The project and the 3 PhD/Post Doc positions focus on
development of the automatic control technology needed for the Artificial
Pancreas.

The projects for the 3 positions (PhD or Post Doc) require significant
collaboration. At a later stage a medical PhD will be associated to the
projects in order to connect to clinical tests at Hvidovre Hospital.  The
projects are defined as follows:



Project 1: Computational Systems Physiology
(Physiological Models, Simulator)
--------------------------------------------
The purpose of this project is to construct a virtual type I diabetes
patient. The virtual patient, i.e. a simulator representing the
glucose-insulin metabolism in man, must be developed and used for initial
test of the insulin dosing concepts. Existing physiological compartmental
PK/PD models for glucose-insulin levels in the blood, subcutaneous insulin
and glucose diffusion, gut carbon hydrate absorption, and physical
exercise models will be used to construct the virtual patient. The
simulator will be used to test the algorithms developed in project 2 and
3. Population modeling is used to reflect the inter-individual variations.
The starting point is existing models for Type 1 and Type 2 diabetes as
well as models from the literature. The virtual patient will be validated
using clinical data.

The candidate for project 1 must have an interest in simulation,
mathematical modelling of dynamic systems, and physiology. Previous
experience in one of the topics is desirable. In addition knowledge of
Matlab is beneficial.



Project 2: Adaptation to the Individual (System Identification)
---------------------------------------------------------------
The predictive quality of the models used in model predictive control
(MPC) is critical to the closed-loop performance of the system. Automatic,
robust and safe methods for identifying the predictive model for each
individual patient are necessary. This project concerns construction and
test of algorithms for optimal input design (optimal experimental design)
and parameter estimation in models tailored for predictive control.
Automatic input design, data-generation and parameter-estimation continue
until the predictive quality of the predictive-control-relevant model is
sufficiently accurate. This decision is based on statistical tests for
validation. The models identified serve as basis for fault detection,
state estimation (including, smoothing, filtering and prediction), and
control. Several families of models and uncertainty representations will
be investigated.

Candidates who have an interest in a research education in time series
analysis, stochastic adaptive control, model predictive control, and
systems identification should apply for this PhD/Post Doc position.



Project 3: Control Systems and Predictive Control Algorithms
------------------------------------------------------------
Project 3 concerns development of the control system and the predictive
control algorithms for the artificial pancreas. The PhD/Post Doc must
develop, implement, analyze and test various predictive controllers using
both linear and nonlinear models. Also robust MPC leading to second-order
cone programs must be investigated. The overall purpose of this project is
to come with a model-based controller that can be tested on the virtual
patient developed in project 1, and ultimately tested in clinical practice
at Hvidovre Hospital.

The PhD/Post Doc should have an interest in doing research within control
and systems theory, model predictive control, and convex optimization.



Qualifications
--------------
Candidates should have a master's degree in chemical engineering, medicine
and technology, biotechnology, applied mathematics, science or a similar
degree with an academic level equivalent to the master's degree in
engineering (e.g. M.Sc. in chemistry, physics, mathematics, computer
science). The candidates must have an interest in at least one of the
topics: mathematical modelling, numerical optimization, numerical solution
of systems of differential equations, time series analysis, control and
system theory, and scientific programming. The candidates must be fluent
in oral and written English and have strong team-work skills.

Post Doc candidates should have a PhD in a relevant subject, e.g. control
and systems theory.

Terms of Employment:
--------------------
The scholarships for the PhD degree are subject to academic approval, and
the candidates will be enrolled in one of the general degree programmes of
DTU. Information about the general requirements for enrolment and the
general planning of the scholarship studies is included in the general
rules of DTU (see
www.dtu.dk/English/education/PhD_Education.aspx).

The salary and appointment terms are consistent with the current rules for
PhD degree students and are negotiated in accordance with the current
collective agreement for Danish University faculty members.

The place of employment is Department of Informatics and Mathematical
Modelling at DTU (the Technical University of Denmark located 10 km north
of Copenhagen).

Further Information
Further information is available at www.imm.dtu.dk or by contacting John
Bagterp Jørgensen by e-mail, [log in to unmask] or tel.: +45 4525 3088 or cell
phone +46 (0)70 9807440.

Application
Applications should be submitted electronically by e-mail to John Bagterp
Jørgensen, [log in to unmask]

Preferably, all application material is collected as one pdf-file.
Applications must include a curriculum vitae and documentation of a
completed master's degree (phd degree for post doc applicants).

All interested candidates irrespective of age, gender,  race, religion or
ethnic background are encouraged to apply.

Application deadline:  August 22nd 2008 at 12.00 noon.



===========================================
John Bagterp Jørgensen, PhD
Assistant Professor of Scientific Computing

DTU Informatics
Building 305 Office 109
Richard Petersens Plads
Technical University of Denmark
DK-2800 Kgs Lyngby

Phone : +45 45253088
E-mail: [log in to unmask]