Date of Award

Spring 1970

Document Type


Degree Name

Doctor of Engineering Science in Mechanical Engineering


Mechanical Engineering

First Advisor

Martin J. Levy

Second Advisor

Robert M. Jacobs

Third Advisor

Russell W. Brancato

Fourth Advisor

Jui Sheng Hsieh

Fifth Advisor

Eugene H. Smithberg


This research is directed toward the development and application of a new approach to the evaluation of cardiac work to diagnose cardiac state. A procedure has been developed for calculating the work performed on the fluid by the left ventricle during the heartbeat. The procedure involves the continuous direct measurement of ventricular fluid mixture temperature during and following the controlled injection, through a catheter, of a known volume of cold saline into the left ventricle. The measured mixture temperatures are used to calculate continuous ventricular volumes during the systolic and diastolic functions of the heartbeat. Plotting measured ventricular pressure versus the volume of the ventricle results in the work diagram for the left ventricle.

The method described above to evaluate cardiac work is based upon the assumption of instantaneous and uniform mixing of the injected saline and the ventricular fluid. This assumption is typically made in studies which employ indicator dilution methods to measure cardiac output and ventricular end-volumes. The effect of ventricular non-mixing of indicators as a source of error in ventricular volume calculations and cardiac output measurement was studied. From references and the author's original invitro and invivo experimental work a description of indicator mixing in the left ventricle and of its effects on indicator dilution studies is given.

A theoretical deterministic analysis of nonuniform ventricular mixing is presented to derive expressions for stroke volume as a function of indicator concentration measured at the aorta. It was found that a purely deterministic analysis when supplemented with a probabilistic analysis results in an analysis of nonuniform ventricular mixing. A mathematical model is developed which explains the shape of indicator concentration curves and allows for the evaluation of ventricular mixing and cardiac state.

A derivation of the classical Stewart-Hamilton relationship for the calculation of cardiac output from dye indicator studies is presented. With this derivation conclusions are formulated which show the validity of the Stewart-Hamilton equation for the case of nonuniform ventricular mixing of the indicator and the limitations of this relationship in the presence of certain heart defects.

The deterministic and probabilistic analyses are applied to the thermodilution technique' to derive expressions relating ventricular volume and ejection fraction to measured fluid temperatures for the case of nonuniform mixing of the injected cold saline. An originally designed "Thermocatheter", which employs a single catheter to inject and measure invivo ventricular fluid temperatures, was used to evaluate cardiac state applying the mathematical analyses presented.

Experimental studies performed in a heart model, mongrel dogs, and in human subjects are presented in verification of the analytical approaches used.