These were texted to me only with "chest pain."
First:
2nd:
What was my response?
Smith: Young thin black male. No OMI.
Texter: Can't fool you. It was indeed.
Queen: #1: NOT OMI, HIGH CONFIDENCE
Queen: #2: NOT OMI, HIGH CONFIDENCE
ECG 1 Interpretation: there is terminal T-wave in V3-V6. Is this Wellens' pattern A? No. this is classic Benign T-wave Inversion. It helps to know that the patient has active chest pain, as Wellen's is a post occlusion (reperfusion) state, with open artery and pain-free.
See 2 dozen examples here:
Understanding this pathognomonic ECG would have greatly benefitted the patient.
ECG 2 Interpretation: Now the T-waves have become upright, so this must be re-occlusion (pseudonormalization), right? No. It is still normal. It is an unfortunate fact that benign T-wave inversion (and ANY T-wave inversion), AND ALSO early repolarization (normal variant STE) can be dynamic.
Outcome:
Patient ruled out for MI by troponins.
Here is a great case of pseudonormalization. Here are multiple cases.
LEARNING POINT:
1. Morphology matters. Pattern recognition is essential. AI does a great job at this but most humans do not.
2. And ECGs can change and evolve even when there is no ischemia. This is where morphology is most critical!
And so:
I sent this ECG to the Queen of Hearts (PMcardio OMI), and here is the verdict:
The Queen is taught be pattern recognition, not by any rules. We just give it ECGs and tell it what they represent. After "seeing" thousands and thousands of them, it can recognize the patterns just like we humans (and AI!) can recognize faces.
- As per Dr. Smith — the already uncanny accuracy of the QoH (Queen Of Hearts) AI application program for identifying acute OMI, is a result of a data base input provided by Drs. Smith and Meyers containing thousands of tracings with documentation of cardiac catheterization results. But the clinical need — is to improve the ECG interpretation ability of the majority of emergency care providers who do not have this extensive ECG-cath-result data base experience to fall back on.
- For clarity in Figure-1 — I've reproduced and put together the 2 serial ECGs that were texted to Dr. Smith in today's case.
Figure-1: Comparison between the 2 ECGs in today's case that were texted to Dr. Smith. (To improve visualization — I've digitized the original ECG using PMcardio). |
- Criterion #1: There is a relatively short QT interval (which I estimate at <400 msec. in ECG #1).
- Criterion #2: The leads with T-wave inversion often have very distinct J-waves (prominent J-point notching being seen in leads V3-thru-V6 in ECG #1).
- Criterion #3: The T-wave inversion is usually in leads V3-V6 — which is in contrast to Wellens' syndrome, in which T-wave inversion is usually in leads V2-V4 (T-wave inversion being seen in leads V3-thru-V6 in ECG #1).
- Criterion #4: The T-wave inversion does not evolve and is generally stable over time — which is in contrast to Wellens' Syndrome, which always evolves. (and although ST-T wave morphology does change in several leads from ECG #1 to #2 — this is not really in a true "evolution" pattern).
- Criterion #5: Chest leads with T-wave inversion often have some ST elevation (This is especially well seen in leads V3,V4,V5 of ECG #1).
- Criterion #6: Right chest leads often have ST elevation typical of classic early repolarization (This is especially noted in lead V2 in ECG #1 — in which the upward concavity = "smiley"-configuration shape of the ST elevation is typical for a repolarization variant).
- Criterion #7: The T-wave inversion in leads V4-V6 is preceded by minimal S-waves (There are no S waves in leads V4-6 of ECG #1).
- Criterion #8: The T-wave inversion in lateral chest leads V4-V6 is preceded by high R-wave amplitude (R waves in leads V4-V6 are ≥20 mm).
- Criterion #9: Leads II, III, and aVF also frequently have T-wave inversion (This is the 1 criterion not satisfied in ECG #1).
- Greenfield and Rembert have noted that up to 1/3 of patients with ER (Early Repolarization) but no cardiac symptoms — may show marked variation in the amount of ST elevation from one tracing to the next. These ER patients do not have ischemia — and the variation in ST elevation has not been shown to be related to either heart rate or QRS amplitudes (Variation in ST-Segment Elevation in Early Repolarization: Electrocardiography 40:10,2007).
- In ECG #2 — 4 leads show a change in ST-T wave appearance. This variation entails: i) Straightening of the ST segment in lead V3, with loss of T wave inversion in this lead; ii) Increased ST elevation in lead V4, albeit with loss of R wave amplitude in this lead; and, iii) Loss of T wave inversion in leads V4-6.
- That said, my overall impression of ECG #2 — was that this variation just did not "look" like a true "dynamic" ST-T wave change indicative of an acutely evolving OMI because: i) Prominent J-point notching persists in the 4 leads with ST-T wave changes in ECG #2; ii) The QTc interval remains relatively short in ECG #2 — with a similar-looking, upward concavity ST segment shape that persists in each of the lateral chest leads; and, iii) There is no significant ST-T wave change in any of the other 8 leads.
- NOTE: If the change in ST-T wave appearance in today's 2 serial tracings represented true "dynamic" variation in an acutely evolving OMI — then I would expect the T wave inversion in leads V3-thru-V6 of ECG #1 to represent reperfusion T waves — with loss of this T wave inversion and the increase in ST elevation seen in ECG #2 to represent acute reocclusion. But in the absence of a clear increase in CP (Chest Pain) severity in association with ECG #2 — I would attribute the change in ST-T wave appearance to the variation that may sometimes be seen with repolarization patterns.