Time-Aligned Heart Strain Curves Detect Severe Coronary Disease but Miss Mild Blockages

Resting echocardiography has long been an attractive candidate for screening patients with chronic chest pain: it involves no radiation, can be repeated freely, and costs far less than computed tomography coronary angiography. Yet its ability to pick up coronary artery disease before vessels are severely narrowed has remained uncertain. A new study from the University Hospital of North Norway in Tromsø now reports that a refined way of analyzing heart muscle deformation—time-aligned segmental strain and strain rate curves—reliably flags patients whose disease is advanced enough to require bypass surgery, but cannot distinguish patients with disease in a single coronary artery from those with non-obstructive changes.

The work, published in the journal Cardiovascular Ultrasound, forms part of the RECCAD study, which asks whether resting echocardiography can identify patients with chronic coronary artery disease. The researchers enrolled 510 consecutive chest pain patients referred for coronary computed tomography angiography between 2016 and 2021, along with a separate group of 102 patients who had suffered a myocardial infarction and were examined one to three days after undergoing percutaneous coronary intervention. Of the 510 chest pain patients, 485 entered the final analysis; 25 were excluded because of pre-existing cardiac conditions such as valvular heart disease, bundle branch block, previous infarction, cardiomyopathy, myocarditis, or atrial fibrillation.

The technical core of the study lies in how strain curves were standardized. Because every heartbeat has a different duration and the aortic valve closes at a different moment in each patient, raw curves are hard to compare across individuals. The team therefore resampled every curve to a fixed length of 100 points per cardiac cycle, aligning aortic valve closure to exactly 40 percent of the cycle by interpolating the systolic and diastolic phases separately. This produced a common temporal framework in which deformation patterns could be compared segment by segment.

From this standardized framework, the investigators built reference curves for each of 18 myocardial segments. These references came from 314 of the chest pain patients who had no coronary artery disease or only stenoses below 30 percent, and no valvular pathology, myocardial disease, or bundle branch block. Each patient’s segmental curves were then compared against the corresponding reference in two ways: by the difference between the patient’s peak systolic and diastolic strain and strain rate values and the reference peaks, and by the average difference between the curves over systole for strain and over the first third of diastole for strain rate. Thresholds for calling a segment pathological were set at the fifth percentile of deviations observed in the no-CAD group.

Patients were classified by the degree of coronary obstruction into no CAD, non-obstructive CAD, and one-, two-, or three-vessel disease, and separately by treatment into no-CAD, non-obstructive, PCI, and CABG groups. Significant stenosis was defined as a 70 percent or greater lumen reduction in the right coronary, left anterior descending, or circumflex artery, or at least 50 percent in the left main artery. Patients with main stem stenosis plus at least one-vessel disease were counted as three-vessel disease. Invasive coronary angiography was performed in all patients with positive or inconclusive computed tomography results and in all myocardial infarction patients.

The results showed a clear gradient with disease severity. Patients in the revascularized myocardial infarction group showed substantial deviations across all parameters tested, with similar patterns in the CABG and three-vessel disease groups, and these differences were statistically significant compared with the no-CAD group despite the small size of the CABG group. Early diastolic strain rate curves showed the most significant deviations between groups defined by CAD severity. Notably, only segmental peak early diastolic strain rate and global longitudinal strain rate in early diastole showed significant differences between the PCI or one-vessel disease groups and the no-CAD group.

Receiver operating characteristic analysis quantified how well the parameters could identify patients needing revascularization, comparing them against chest pain patients without an indication for intervention. For detecting patients who underwent percutaneous coronary intervention, performance was generally poor, with area under the curve values below 0.7. For detecting patients who required coronary artery bypass grafting, performance was excellent, with AUC values reaching 0.933. In both analyses, peak early diastolic strain rate deviation and global longitudinal strain rate in early diastole emerged as the best indicators among all tested parameters. When the same parameters were used to distinguish patients without or with non-obstructive CAD from revascularized infarction patients, AUC values dropped below 0.8, which the authors interpret as evidence that some patients regain near-normal global and segmental function after revascularization.

The segmental reference approach itself proved consequential. When the researchers contrasted segment-specific reference curves with a single global average curve applied across all segments, they found that apical pathologies were frequently missed without segmental normalization: 28.3 percent of myocardial infarction patients showed pathological apical segments when segmental references were used, versus only 19.5 percent with global normal values. Mid-ventricular segments were least dependent on segmental references, while basolateral and basoseptal segments with reduced strain were disproportionately flagged as abnormal under global cut-offs. Segmental-based definitions of pathological deviation improved diagnostic accuracy particularly for apical segments, and the need for segmental reference values was most evident for early diastolic strain rate parameters.

Across all stages of coronary disease, the deviating strain curves displayed consistent patterns: delayed contraction, reduced systolic shortening slopes, and post-systolic shortening, concentrated predominantly in basal and mid-ventricular segments. Post-systolic shortening also appeared in basal septal segments of the no-CAD group as a normal phenomenon. In infarcted regions, segments showed initial systolic stretching, delayed contraction, and reduced shortening slopes followed by post-systolic strain. Deviating apical curves rarely showed post-systolic shortening, which the authors attribute to weaker contraction in neighboring segments, since apical segments normally generate higher strains than mid-ventricular ones.

The study has important limitations that shape how the findings should be read. The two-vessel disease group contained only 12 patients, the three-vessel group only 7, and only 8 patients were referred for bypass surgery; these small groups produced significant results with high effect sizes but were underpowered for reliable negative conclusions, whereas the one-vessel and PCI groups were large enough to support negative findings. The chest pain population was also not a healthy reference group—hypertension, obesity, and diabetes are known to reduce strain, and references derived from a risk-factor-free population might have yielded different results. No SYNTAX score was calculated, which might have allowed a more comprehensive grouping that accounted for coronary occlusions. The authors conclude that one-vessel coronary artery disease cannot be reliably distinguished from non-obstructive CAD using resting speckle tracking echocardiography, but that global longitudinal strain rate in early diastole effectively identifies high-grade disease requiring bypass grafting, and that resting strain analysis is therefore best suited to identifying high-risk patients with advanced coronary disease rather than those with mild or single-vessel disease.

Subject of Research: Diagnostic value of time-aligned segmental strain and strain rate curves from resting echocardiography for detecting obstructive coronary artery disease in chronic chest pain patients

Article Title: Time-aligned segmental strain as a diagnostic marker for obstructive coronary artery disease in patients with chronic chest pain

Article References: Rösner, A., Caglayan, H. A., Jazbani, M., Sari, S. Ø., Castellote, P. M., Kjønås, D., & Queirós, S. (2026). Time-aligned segmental strain as a diagnostic marker for obstructive coronary artery disease in patients with chronic chest pain. Cardiovascular Ultrasound, 24(1), Article 26. https://doi.org/10.1186/s12947-026-00388-0

Image Credits: AI Generated

DOI: 10.1186/s12947-026-00388-0

Keywords: coronary artery disease, echocardiography, speckle tracking, myocardial strain, strain rate, chest pain, coronary computed tomography angiography, coronary artery bypass grafting, percutaneous coronary intervention, diastolic function, RECCAD study, diagnostic accuracy