Research Interests

	Biomedical and physiological
	systems modelling
	Decision support systems
	Pulmonary physiology
	Mechanical ventilation
	Respiratory Monitoring
	Java development and programming

Project Opportunities                                            to the top

Pulmonary pressure-volume relationship and gas exchange

Collaborators: Bram W Smith, Steve Rees and Steen Andreassen

Current models of lung mechanics and gas exchange act independently to simulate variations in pressure-volume (PV) and ventilation-perfusion (V/Q) properties in the lungs respectively. However, changes in ventilator pressures can cause alveoli recruitment, collapse or over-distension causing V/Q changes in the lungs that are unaccounted for in these models. MMDS has created a compartmental model of the lungs based on a physiological interpretation of lung function and simulating each alveolus individually. This is the first known model of its type and has shown potential for simulating experimentally measured pressure-volume relationships in the lungs.

This project will involve continued development of this alveoli model from its current initial stage, through further development, experimental verification and, if successful, implementation in the clinical setting as a diagnostic tool and as part of the INtellengent VENTilator (INVENT) project at MMDS. It is hoped that the model will simulate how changes in respiratory pressures will affect the ventilation and perfusion parameters such as dead-space, shunt, lung density, pulmonary vascular resistance and lung volumes for healthy and diseased lungs. Animal trials are, planned for September 2005, are intended to produce experimental data of static PV curves for both healthy and diseased lungs that can be used to verify the model.

Students working on this project can expect to gain a physiological understanding of respiratory function and experience in model development, verification and implementation.
 

Investigation of alveolar recruitment using CT scan analysis and static respiratory mechanics

Collaborators: Bram W Smith, Steve Rees, Steen Andreassen, Per Lambert and Anders Larsson

Patients residing in the intensive care unit often depend on ventilator therapy to survive. The process of choosing appropriate ventilator pressures and volumes in these patients is complicated for clinicians. Changes may in some lung regions recruit collapsed alveoli improving gas exchange. At the same time already open alveoli in other lung regions may become over-distended increasing the risk of ventilator induced lung injury (VILI).

CT scans of the lungs can be used to assess patient lung function by measuring the number of open, collapsed and over-distended alveoli. However, this technique is impractical for clinical use, partly due constraints such as cost, availability and patient safety. Measurements of the static pressure-volume (PV) curve provide an alternative, low cost method of assessing lung mechanics using a bedside ventilator. The characteristic shape of the static PV curve has been investigated in several studies to assist setting ventilator pressures and volumes. However, it is not yet clear how the shape of the static PV curve relates to the amount of collapsed and over-distended alveoli.

MMDS has recently obtained results from a study of animals with healthy and damaged lungs, where CT scans have been taken at different pressures and volumes while measuring the static PV curve. This data enables investigation of the relationship between the static PV curve and the number of open, collapsed and over distended alveoli estimated from the CT scans. Thus it will be possible to identify any features of the static PV curve which may assist setting ventilator pressures and volumes. The data includes other lung mechanics measurements for analysis such as the dynamic lung compliance. The project involves analysis of CT scans to identify how PV curve measurements correlate to the amount of collapsed and over-distended alveoli. The eventual aim is to develop a system for suggesting ventilation pressures and volumes. The project can include physiological modeling of respiratory mechanics.

Students working on this project can expect to gain a physiological understanding of respiratory mechanics and experience in image analysis, physiological modeling and feature extraction.

Step to target system for INVENT

Collaborators: Bram W Smith, Steve Rees

Patients residing in the intensive care unit (ICU) often depend on ventilator therapy to survive. The process of choosing appropriate ventilator settings in these patients is complicated by changes in ventilator settings that may improve gas exchange in the lungs but at the same time increase risk for ventilator induced lung injury. For example, it can be necessary to increase ventilator volumes to open up collapsed alveoli and improve gas exchange but high volumes may also injure the lungs. Thus, it can be difficult to obtain the right balance between the positive and negative effects of changing the ventilator settings, and incorrect settings have been shown to increase mortality.

MMDS are developing a decision support system (INVENT), which based on physiological models of gas exchange, lung mechanics and acid base chemistry, can describe the current sate of the patient and simulate effects of changing ventilator settings. INVENT is based on penalty functions, which is decision theory implementing a doctor’s preference for different measures e.g. how high oxygen saturation should be. Using these penalty functions, INVENT gives advice on three ventilator settings: inspired oxygen fraction (FiO2), tidal volume (Vt) and respiratory frequency (f) by simulating the resulting penalties associated with different settings and finding the settings with the smallest total penalty.

Changing Vt can affect alveolar recruitment and collapse resulting in a different respiratory compliance. The INVENT advice depends on a measure of respiratory compliance, thus if this changes the original INVENT advice may be incorrect. It is necessary, before clinical application, to develop a step to target Vt function for INVENT, which in a series of steps alters Vt towards the target and evaluates, whether it is necessary to calculate a new advice. A mathematical model of respiratory mechanics must be applied to evaluate the lung mechanics. The project involves creating a system which implements INVENT with a step to target function for Vt and can communicate with a ventilator. MMDS has a ventilator connected to a mechanical lung, which can be programmed to act as a patient.

Students working on this project can expect to gain a physiological understanding of respiratory mechanics and gain experience in physiological modeling, object oriented system development and JAVA programming.

Full papers and conference proceedings                                                            to the top

 2008

Allerød, Charlotte; Rees, Stephen Edward; Rasmussen, Bodil S.; Karbing, Dan Stieper; Kjærgaard, Søren; Thorgaard, Per; Andreassen, Steen. 2008. A decision support system for suggesting Ventilator settings: retrospective evaluation in cardiac surgery patients ventilated in the ICU. In: Computer Methods and Programs in Biomedicine; Vol. 92, nr. 2, Page.205-212. Journal (Full)

Karbing, Dan Stieper. Kan et beslutningsstøttesystem indstille en respirator?  I: Medicinsk Teknologi og Informatik. 2008 ; vol. 5, nr. 4, s. 12-14. Journal

Karbing, Dan Stieper, Kjærgaard, Søren, Smith, Bram Wallace, allerød, Charlotte, Espersen, Kurt, Andreassen, Steen,Rees, Stephen Edward. Decision support of inspired oxygen fraction using a model of oxygen transport. In: Proceedings of the 17th World Congress of The International Federation of Automatic Control, IFAC 2008, 6-11 July 2008, Seoul, South Korea (CD-ROM). 2008, page 8080-8084. (Conference Proceeding)

Karbing, Dan Stieper; Rees, Stephen Edward. Authors' response to Mohamed F El-Khatib´s and Gassan W Jamaleddine´s letter: Clinical relevance og the PaO2 /FiO2 ratio. In: Critical Care. 2008; vol. 12, nr. 1, page 2.(Letter)

Variation in the PaO2/FiO2 ratio with FiO2: Mathematical and experimental description, and clinical relevance. Dan S. Karbing, Søren Kjærgaard, Bram W. Smith, Steen Andreassen, Stephen E. Rees. Submitted to Intensive Care Medicine.
 

Abstracts  

2008

Allerød, C. ; Kjærgaard, S. ; Karbing, Dan Stieper ; Espersen, K. ; Rees, Stephen Edward. Retrospective evaluation of a decision support system for advising on ventilator settings in patients with ARDS/ALI. 2008. p. S75, No. 0283 Konferencen: European Society of Intensive Care Medicine, ESICM Annual Congress, nr. 21, Lisbon, Portugal, 21. september 2008 - 24. september 2008. Intensive Care Medicine. Suppl. 1. Springer.

Allerød, C. ; Karbing, Dan Stieper; Kjærgaard, Søren Christensen . Retrospective evaluation of a decision support system for advising on ventilator settings in patients with ARDS/ALI. 2008. p. 48 Konferencen: Årsmøde i Dansk Selskab for Anæstesiologi og Intensiv Medicin, Dasaim, København, danmark, 13th. november 2008 - 15th. november 2008. Dasinfo. 4. Dansk Selskab for Anæstesiologi og Intensiv Medicin.

Karbing, Dan Stieper; Kjærgaard, Søren ; Smith, Bram Wallace; Espersen, Kurt ; Andreassen, Steen ; Rees, Stephen Edward. Variation in the PaO₂/FiO₂ ratio with FiO₂ : is it relevant?  2008. s. 250-251 Konferencen: Joint SCATA ESCTAIC Scientific Meeting, London, Storbritannien, 21. november 2007 - 23. november 2007. Journal of Clinical Monitoring and Computing. 22. 3. Springer New York LLC.

2006

Dan S. Karbing, Bram W. Smith, Søren Kjærgaard, Charlotte Allerød, Steen Andreassen, Stephen E. Rees (2006) Examining the effect of positive end-expiratory pressure on gas exchange and lung mechanics. 17th annual meeting of the European Society for Computing and Technology in Anaesthesia and Intensive Care (ESCTAIC) Milan, September 2006. Abstract.

Dan S. Karbing, Bram W. Smith, Stephen E. Rees and Steen Andreassen (2006) Modelling the relationship between gas exchange and lung mechanics. 17th annual meeting of the European Society for Computing and Technology in Anaesthesia and Intensive Care (ESCTAIC) Milan, September 2006. Abstract.

2005

Smith, B. W., Rees, S. E., Christensen, T. F., Karbing, D. S., Andreassen. S. (2005) Getting the Most from Clinical Data through Physiological Modeling and Medical Decision Support. 16th annual meeting of the European Society for Computing and Technology in Anaesthesia and Intensive Care (ESCTAIC) Denmark, September 2005. Abstract.