When doing research many unexpected important discoveries are made while pursuing a totally different goal. Recording mitral annular motion with M-mode echocardiography is one such discovery. When I saw my first ultrasound instrument in 1963, my immediate idea was to use this instrument to detect pericardial effusion. With the use of a Smith-Kline Instruments (Sunnyvale, CA) ultrasonoscope owned by the neurologists at Indiana University for the detection of the midline of the brain, I was able to demonstrate that diagnostic ultrasound could detect pericardial fluid. At this point I wanted to know what else was known about the use of ultrasound to examine the heart. I obtained a copy of Dr. Edler’s 1961 review article in Acta Medica Scandinavica. I also learned that Dr. Joyner in Philadelphia had published an article in Circulation in 1963 duplicating Edler’s observation of mitral valve motion in patients with mitral stenosis. I visited Dr. Joyner and also attended a 1965 meeting in Pittsburg on general diagnostic ultrasound. Dr. Edler was one of the speakers. From my visit to Joyner’s lab and hearing Edler’s lecture, they felt that the only useful clinical application of cardiac ultrasound was using the mitral valve diastolic E to F slope for the detection of mitral stenosis.
The ultrasound instrument that I had access to was an A-mode ultrasonoscope. My visit to Philadelphia had a dual purpose. One was to visit Dr. Joyner. The other was to obtain a “free” “Slow Sweep” module from Smith-Kline Instruments so that I could obtain M-mode recordings. One objective was to confirm that the E to F slope was useful to detect mitral stenosis, which was a very common problem in those days.
It didn’t take long to document that patients with mitral stenosis did indeed have a flat E to F slope. The problem was that I could also record a flat E to F slope in some patients without mitral stenosis. Thus, that finding alone was not specific for mitral stenosis. I was not the first to recognize this lack of specificity. Unfortunately this known problem hindered the credibility of diagnostic ultrasound. Some early investigators apparently dabbled with cardiac ultrasound, ran into this problem and abandoned the technique. This credibility problem also contributed to the skepticism in the early use of diagnostic ultrasound for the detection of pericardial effusion, which had nothing to do with the EF slope.
Dr. Adib Zaky, who had immigrated to the United States from Egypt, was working with me and we became curious as to why the E to F slope was reduced in patients without mitral stenosis. Our theory was that with mitral stenosis the flat E to F slope reflected the slow rate of left ventricular filling due to the narrowed mitral orifice. In patients without mitral stenosis a flat E to F slope was likely the result of abnormal mitral ring or annular motion to which the mitral leaflet was attached. At that time we used the term “mitral ring” rather than “mitral annulus”.
To test our theory, we began exploring the possibility of recording echoes from the mitral ring. With a trial and error approach, we found that if one placed the transducer at the cardiac apex, one could record the mitral annulus as a distinctly different echo from the valve leaflet ( Figure 1 ). Much to our delight, the early diastolic slope of both the annulus and mitral valve were similar ( Figure 2 ). Dr. Zaky was quite excited in finding a possible explanation for the nonspecificity of the mitral E to F slope for the diagnosis of mitral stenosis. My own reaction to this finding was different. My interest in examining the left ventricle was the main reason I went to see my first ultrasound instrument in the first place. Thus, when I saw the mitral annulus recordings, I was struck by the similarity of the recordings and left ventricular volume curves ( Figure 3 ). We noted that the amplitude of the mitral ring motion was reduced in patients with a bad left ventricle. We then proceeded to see if the amplitude of mitral ring (annulus) motion was related to left ventricular (LV) function in some way. Ejection fraction was not popular in those days. It had already been discarded as not being a useful measure of LV function because of its being so preload and afterload dependent. As a result we used the amplitude together with a rough measurement of the size of the LV to show a relationship with LV stroke volume. It turned out to have a fairly good relationship and that work was published in Circulation in 1967. We presented this material at the world’s first meeting devoted to cardiovascular ultrasound in Indianapolis in January 1968. By this time Richard Popp was a fellow working in our laboratory. At that meeting Dr. Popp presented our work in measuring the thickness of the posterior LV wall.
