Sunday, April 2, 2023

I was reading EKGs on the system and came across this one......

I was reading EKGs on the system and came across this one......

I am always looking for EKGs that show either subtle OMI or OMI mimics.
Which did I think this was?

I suspected that this was a mimic.  It is hard to say why.  There is "inferior" ST depression, reciprocal to high lateral ST Elevation.  The precordial leads have ST Elevation but also high voltage and look like possible LVH.  I was not certain that it did not represent OMI, but was pretty sure.

So I went to the chart and found that it was not different from previous ECGs and the patient ruled out for MI by serial troponins.

This is a very difficult ECG and so I wanted to know how the PM Cardio Bot would perform.

Here is the transformed ECG (the bot uses a photo of an image, or uses a PNG or JPG or whatever image file, transforms it to a digital file that can be manipulated by AI, then makes an image from that digital file):

Here is the interpretation by the AI bot:

So this AI bot, trained by Pendell and me, recognizes that this is not OMI.


You can get this AI app here:


My Comment by KEN GRAUER, MD (4/2/2023):
I found today's tracing interesting for a number of reasons.

  • As per Dr. Smith — The "Queen of Hearts" AI app predicted with high confidence that the ECG in today's case was not indicative of an acute OMI. As per my recent comment ( = the March 31, 2023 post in Dr. Smith's ECG Blog) — the reason for the amazingly high predictive accuracy of the Queen of Hearts app, is that it has been programmed by Drs. Smith and Meyers (and therefore reflects their uncanny ability to accurately recognize what is — and what is not — an acute OMI).

  • Today's ECG is not an easy tracing to interpret. This ECG is not normal. I like this case as a teaching example — because it illustrates the type of tracing about which definitive decision (ie, acute OMI — or no acute OMI) — is simply not possible from review of this single ECG without the benefit of any history.

  • Today's case also illustrates the importance of remaining a "thinking clinician" when applying predictions from the Queen of Hearts app. Speaking from my experience of participating in one of Dr. Smith's ECG studies, in which I was one of the expert interpreters — Seeing the numerical computer calculation of Dr. Smith's 4-Variable Formula after I formulated my own opinion was always comforting to me when computer predictions agreed with my interpretation. On occasions in which I disagreed with the computer — this difference in interpretation forced me to reevaluate my decision-making process (and in so doing — increased the accuracy of my interpretation!).

  • Finally — I found the combination of ECG findings in today's tracing intriguing and unusual. For clarity in Figure-1 — I highlight some of the findings in today's case.

Figure-1: I've added the mirror-image of leads V2 and V3 from today's tracing.

MY Thoughts on the ECG in Figure-1:
We were not provided with any history about the patient whose ECG is shown in Figure-1. A brief, relevant history is a critical part of the interpretation process — since a story of typical cardiac-sounding chest pain immediately elevates statistical likelihood of an acute event — whereas incidental ordering of an ECG for another reason (ie, as part of routine testing for surgical clearance) — would reduce statistical likelihood of an acute event for an otherwise "non-diagnostic" tracing.
  • The rhythm in Figure-1 is sinus at ~60/minute (ie, There are small-but-present upright P waves in front of each QRS in the long lead II rhythm strip — all manifesting a constant normal PR interval). The QRS is narrow. The QTc is not prolonged given the slow rate. The frontal plane axis is normal (about +30 degrees).
  • Voltage criteria for LVH are satisfied, assuming the patient is not a young adult (ie, Peguero Criteria — given the very deep S wave = 28 mm in lead V3 — See My Comment in the June 20, 2020 post in Dr. Smith's ECG Blog for "My Take" re ECG diagnosis of LVH).

The most remarkable findings on this tracing are the abnormal ST-T waves in multiple leads. 
  • In high-lateral leads I and aVL — there are wide Q waves (considering the small QRS amplitude) — and slight ST elevation with hyperacute-appearing T waves. 
  • In the inferior leads — there are mirror-image reciprocal ST-T wave changes to the slight ST elevation and peaked T wave in lead aVL.
  • In leads V4 and V5 — the ST segments are completely flat, with disproportionately peaked T waves. 
  • In lead V6 — there is unexpected ST depression with terminal T wave positivity (unexpected given the contrast in appearance from the ST-T waves seen in neighboring leads V4,V5).
  • In anterior leads V1,V2,V3 — R wave progression is delayed but an initial r wave is present in each of these 3 anterior leads. The remarkable finding is the overly peaked T waves, especially in leads V2,V3. Even with the extremely deep S wave in lead V3 — I thought the T waves in leads V2,V3 were disproportionately tall.

Putting It All Together:
As stated earlier — this is not a normal tracing. That said — I found it difficult to know what to make of these findings without the benefit of a brief, relevant history and a prior tracing for comparison (that we were not initiallly provided with).
  • "Young" adults often manifest increased QRS amplitude without true chamber enlargement. While no strict "cut-off" point for what constitutes a "younger adult" has been established — I've found voltage criteria for LVH to be a less reliable indicator of true LV chamber enlargement in adults under 35.
  • Assuming today's patient is ≥35 years old — Peguero Criteria for LVH voltage would be easily exceeded (ie, Sum of deepest S in any chest lead + S in V4 ≥23mm [female] or ≥28mm [male]).
  • ST-T wave changes of LV “strain” most often manifest in one or more of the lateral leads. But instead of seeing ST-T wave changes of LV “strain” in lateral leads — some patients manifest a “mirror-image” of strain in anterior leads. The unusual (unexpectedly peaked) shape of the elevated ST-T waves in leads V2,V3 in Figure-1 could reflect LV “strain” in this patient with significantly increased voltage. Doesn’t the mirror-image of these ST-T waves in the RED inserts for leads V2,V3 in Figure-1 look consistent with how LV “strain” might look in the lateral leads of a patient with marked hypertension?
  • BOTTOM Line: If today's patient presented to the ED for new-onset cardiac-sounding chest pain — I would not be able to rule out an acute cardiac event from this single tracing.

What I Still Want to Know:
As impressed as I am that the Queen of Hearts app correctly identified with high confidence that today's tracing was not indicative of an acute OMI — this does not negate the fact that today's ECG is not a normal tracing.
  • As per the follow-up provided by Dr. Smith — Negative troponins and the lack of any significant change in today's ECG from the appearance of prior tracings ruled out an acute coronary event.

  • That said — I remain intellectually curious as to WHY in ECG #1 there appears to be significant LVH — and WHY there are 11/12 leads that manifest abnormal ST-T wave changes potentially consistent with some form of underlying heart disease? 


Editorial Comment by Ken Grauer: I have frequently used PMcardio digitization over recent months for educational purposes. For this function — I love this application, as it dramatically improves the appearance of many artifact-laden and often distorted smart phone ECG photographs that are sent my way, which I otherwise would not be able to use for teaching. Readers of this ECG Blog have doubtless seen many of the tracings I post with notation, "visualization improved by PMcardio".
  • To the above — I add a cautionary note that in my opinion — No computerized application should substitute for clinician oversight and judgment. While the digitized reproductions I have used are overall excellent (!) — they are not perfect. My suggestion is therefore to verify accuracy of digitization by comparison with the original tracing, because on occasion some elements of scanned tracings may be to some extent "off". 
  • That said — the performance I've seen thus far of the Queen of Hearts function is truly impressive for its accuracy in recognizing acute OMI (again, primarily because this app has been taught by the expertise of Drs. Smith and Meyers!).
  • Putting this all together — optimal use of ECG digitization, with additional use of the Queen of Hearts application in patients with new chest pain — is in my opinion, best achieved by the thinking clinician who still closely oversees the case, with full appreciation for the valuable input provided by Queen of Hearts


No comments:

Post a Comment

DEAR READER: I have loved receiving your comments, but I am no longer able to moderate them. Since the vast majority are SPAM, I need to moderate them all. Therefore, comments will rarely be published any more. So Sorry.

Recommended Resources