Acute Coronary SyndromesTRIACS-MBNCT07882849
AI-ECG Detection of Occlusion Myocardial Infarction at Emergency Department Triage (TRIACS)
An observational study of Acute Coronary Syndromes and Chest Pain, run by University Medical Centre Maribor at 1 site in Slovenia.
Trial summary
Chest pain is among the most common reasons people attend an emergency department, and the electrocardiogram (ECG) recorded at triage is the first test used to decide who needs urgent treatment for a blocked coronary artery. Recognising the ECG patterns that indicate an acutely blocked artery is difficult. Read more about this study
- Who is it for?
- Ages 18 years and older
- Conditions the study is for
- Acute Coronary Syndromes, Chest Pain, Myocardial Infarction (MI)
- How many people
- 1,000 participants
- Where
- 1 siteSlovenia
- Study dates
- 19 Oct 2026to 19 Nov 2027
- Phase
- Observational
- Are there results?
- Not yet: the study is under way
Key requirements
Who can take part, as listed by the sponsor. Only the study team can confirm whether you are eligible, so take this list to your doctor.
- Ages 18 years and older
- All sexes
- No healthy volunteers
- Adult, Older adult
Study population: All adults presenting to ED of UKC Maribor, triaged as chest pain that have ECG recorded and agree to be included into study.
Who may be able to join
- Age ≥18 years
- Presentation to ED triage with symptoms suspicious for acute coronary syndrome, including chest discomfort, dyspnoea, epigastric pain, or other anginal equivalents
- Standard 12-lead ECG recorded at triage as part of routine care, available in machine-readable or interpretable image format
Who cannot join
- Traumatic cause of chest pain
- Ventricular arrhythmia or ongoing cardiopulmonary resuscitation at the time of the triage ECG
- ECG of insufficient technical quality for analysis (recorded and reported as an exclusion count)
- Patients with prehospital STEMI diagnosis transported directly to the catheterisation laboratory and bypassing ED triage
- Patients for whom laboratory results are not retrievable by UKC MB
- Declined participation / withdrawal of consent
Trial locations and contacts
1 site in 1 country. Contact the study team or a site near you to ask about taking part.
Slovenia1 site
UKC Maribor
Maribor, 2000
Luka Petravić, MD, +38623210000, luka.petravic@ukc-mb.si
David Šuran, MD, PhD
About this study
Official title: Prospective External Validation of an Artificial Intelligence Electrocardiogram Model for Detection of Occlusion Myocardial Infarction in Patients Presenting to Emergency Department Triage With Suspected Acute Coronary Syndrome: The TRIACS Study
Chest pain is among the most common reasons people attend an emergency department, and the electrocardiogram (ECG) recorded at triage is the first test used to decide who needs urgent treatment for a blocked coronary artery. Recognising the ECG patterns that indicate an acutely blocked artery is difficult. Many such patterns do not produce the classic ST-segment elevation, and identifying them reliably requires a level of ECG expertise that is not always available at the point of triage - particularly at night and among clinicians who are not cardiologists.
Artificial intelligence (AI) models that read the 12-lead ECG have been shown in several retrospective studies, mainly from Western Europe and the United States, to identify acute coronary occlusion more sensitively than conventional interpretation. These models are now in clinical use in a number of health systems. Their performance has not been assessed in a prospectively enrolled, unselected Slovenian emergency department population, and diagnostic accuracy commonly declines when a model is applied outside the setting in which it was developed.
Read more
TRIACS is a prospective, single-centre, observational external validation study in the Emergency Department of University Medical Centre Maribor. All consecutive adults presenting with symptoms suggestive of acute coronary syndrome who have a 12-lead ECG recorded at triage as part of routine care will be enrolled. Each triage ECG is analysed by the AI model. The model's output is withheld from all treating clinicians and plays no part in patient care; every patient receives standard guideline-directed management.
The main aim is to measure how accurately the AI model identifies angiographically confirmed occlusion myocardial infarction in this population, and to compare its accuracy with the triage decision actually made by the clinical team. The study is therefore designed to inform an adoption decision in either direction, rather than to demonstrate a predetermined conclusion.
Read the detailed description
Background. Approximately one quarter to one third of patients ultimately diagnosed with non-ST-elevation myocardial infarction have an acutely occluded culprit artery, and delayed invasive management in this group is associated with substantially higher mortality. The Fifth Universal Definition of Myocardial Infarction (2026) has broadened the formal electrocardiographic criteria for myocardial ischaemia to include ST-segment depression patterns, posterior and right-sided lead criteria, the Aslanger pattern, the South African flag sign and modified Sgarbossa criteria. Applying these criteria reliably requires expert-level ECG interpretation. The practical diagnostic gap at the point of triage is therefore now less a matter of criteria than of the expertise available to apply them.
AI-based ECG interpretation has demonstrated superior sensitivity to conventional millimetre criteria in retrospective multicentre cohorts and is in clinical deployment in several health systems. Prospective validation in unselected emergency department populations remains limited, and reported accuracy commonly declines on external validation.
- Design.** Prospective, single-centre, observational diagnostic accuracy study (external validation of an existing diagnostic model). Reported in accordance with TRIPOD+AI and STARD.
- Setting.** Emergency Department, University Medical Centre Maribor, Slovenia. Approximately 4,500-5,000 chest-pain presentations annually; approximately 910 ED-linked coronary procedures annually.
- Population.** Consecutive adults (≥18 years, no upper age limit) presenting to ED triage with symptoms suspicious for acute coronary syndrome in whom a 12-lead ECG is recorded and digitally captured at triage as part of routine care.
- Index test.** 'PM Cardio, Queen of Hearts', a CE-marked deep-learning model returning a binary verdict accompanied by a confidence tier and, where available, a continuous probability score. Only the first 12-lead ECG recorded at triage is analysed; one ECG per presentation. The decision threshold is fixed a priori at the manufacturer's on-label cut-off for the general chest-pain population. No threshold optimisation is performed for the primary analysis.
- Blinding.** The triage nurse, all treating emergency physicians and all consulting cardiologists are blinded to AI output. AI inference is performed in batch after the clinical episode has concluded. No AI output is entered into the medical record, is accessible from any clinical system, or is communicated to any member of the clinical team at any point. Reference-standard adjudication is performed by an independent adjudicator blinded to AI output.
- Reference standard.** The primary target condition is occlusion myocardial infarction (OMI): symptoms or signs consistent with myocardial ischaemia, AND acute myocardial injury demonstrated by a dynamic rise and/or fall in high-sensitivity cardiac troponin I with at least one value above the sex-specific 99th percentile upper reference limit, meeting the criteria for primary myocardial infarction under the Fifth Universal Definition of Myocardial Infarction (2026), AND angiographic evidence of an acute culprit lesion with either (a) TIMI flow grade 0-1, or (b) TIMI flow grade 2-3 with emergent or urgent percutaneous revascularisation.
Under the Fifth UDMI an isolated troponin elevation without ischaemic symptoms or signs constitutes myocardial injury, not infarction, and does not satisfy this definition. Primary myocardial infarction is no longer restricted to atherothrombosis and encompasses spontaneous coronary artery dissection, coronary embolism and vasospasm; such cases meet the reference standard where the angiographic criterion is satisfied. Myocardial injury with non-obstructive coronary arteries (MINOCA, a working rather than final diagnosis under the Fifth UDMI) is adjudicated as OMI-negative and reported as a named category.
Because coronary angiography is not performed in all patients, a composite reference standard is used with the verification route pre-specified:
- Angiographed during the index episode: adjudicated on angiographic and biomarker data as above.
- Not angiographed: adjudicated OMI-negative where serial hs-cTnI testing (ESC 0/1 h algorithm) rules out myocardial infarction, the discharge diagnosis excludes acute myocardial infarction, and 30-day follow-up shows no myocardial infarction, unplanned revascularisation or cardiac death.
- Biomarker-positive, managed non-invasively: adjudicated individually by the blinded adjudicator using all available clinical, biomarker and imaging data.
Ambiguous or discordant cases are reviewed by a second independent adjudicator, with disagreement resolved by consensus. Inter-adjudicator agreement is reported (Cohen's kappa).
- Comparators.** (i) The standard-of-care triage decision actually made by the clinical team, defined as urgent cardiology referral or catheterisation laboratory activation within 10 minutes of ECG availability, extracted from routine records; (ii) conventional ST-elevation millimetre criteria; (iii) the full electrocardiographic ischaemia criteria of the Fifth UDMI. Comparators (ii) and (iii) are applied to the same triage ECG by a reviewer blinded to AI output, clinical course and outcome.
- Follow-up.** 30 days from index presentation, via hospital records and national registry linkage.
- Conditions
- Acute Coronary Syndromes, Chest Pain, Myocardial Infarction (MI), NSTEMI - Non-ST Segment Elevation MI, ST Elevation Myocardial Infarction, Occlusion Myocardial Infarction
- Keywords
- ECG, AI ECG analysis, ACS, Acute Coronary Syndrome, Chest Pain, Emergency department, PCI, Percutaneous coronary intervention
- Medical subject headings
- Acute Coronary Syndrome, Chest Pain, Myocardial Infarction, Non-ST Elevated Myocardial Infarction, ST Elevation Myocardial Infarction, Emergencies
What participants receive
What participants receive. In a randomised study each person is assigned to one group by chance.
| Study group | Role | What they receive |
|---|---|---|
| Suspected acute coronary syndrome at ED triage | Consecutive adults with a digitally captured triage 12-lead ECG. No intervention assigned; AI analysis performed on routinely acquired ECGs with output withheld from clinicians. |
What is being measured
The primary outcome is the main question the study is designed to answer.
Diagnostic accuracy of the AI-ECG model for angiographically confirmed occlusion myocardial infarction
Sensitivity, specificity, positive predictive value, negative predictive value, overall accuracy and area under the receiver-operating characteristic curve of the AI model's verdict on the first 12-lead ECG recorded at triage, against the pre-specified composite reference standard for OMI. The decision threshold is fixed a priori at the manufacturer's on-label cut-off; no threshold optimisation is performed.
Measured: From triage ECG to 7 and 30 days after index presentation
Secondary outcomes (2)
Comparative diagnostic accuracy versus formal STEMI criteria
Compared with the full ECG criteria for myocardial ischaemia and infarction under the Fifth UDMI (2026): the thresholds in 2a, plus new horizontal or down-sloping ST-segment depression ≥0.5 mm at the J-point in two or more contiguous leads; T-wave inversion >1 mm in two contiguous leads; posterior infarction patterns (ST-depression ≥1 mm in V1-V3, particularly with dominant R wave, confirmed by ST-elevation in V7-V9); right ventricular infarction (ST-elevation in V3R-V6R); the Aslanger pattern; the South African flag sign; and modified Sgarbossa criteria in LBBB and ventricular pacing. Applied by a blinded reviewer.
Measured: From triage ECG to 7 and 30 days after index presentation
Versus conventional ST-elevation millimetre criteria
Compared with ST-elevation at the J-point in two contiguous leads: ≥1 mm in all leads other than V2-V3; in V2-V3, ≥2 mm in men ≥40 years, ≥2.5 mm in men <40 years, ≥1.5 mm in women. Applied by a blinded reviewer. Retained because it corresponds to the criteria used in the prior validation literature and permits direct comparison with those cohorts.
Measured: From triage ECG to 7 and 30 days after index presentation
Timeline
From the start of the study to its completion; the mark is today. Estimated dates are the sponsor’s plan.
- First posted20 Oct 2026
- Study start19 Oct 2026 (estimated)
- Primary completion19 Oct 2027 (estimated)
- Study completion19 Nov 2027 (estimated)
Resources
- Official record on ClinicalTrials.govThe registry entry this page is built from, with its full change history.
- Publications (8)Papers the sponsor has linked to this study.
- Other trials for Acute Coronary SyndromesStudies of the same condition, open and under way.
- What the trial phases meanHow a treatment moves from first tests in people to approval.
More details
For doctors and researchers: how the study is designed, who runs it, and its registry IDs.
Study design
- Study type
- Observational
- Observational model
- Cohort
- Time perspective
- Prospective
- Enrolment
- 1,000 participants (estimated)
Sponsor and investigators
- Lead sponsor
- University Medical Centre Maribor
- Sponsor type
- University, hospital or other
- Collaborators
- 3fs Nordic SI, d.o.o., University of Ljubljana
- Responsible party
- Principal investigator, Luka Petravic, MD, University Medical Centre Maribor
- Investigators
- Luka Petravić, MD, Principal investigator, UKC Maribor
IDs, dates and publications
- ClinicalTrials.gov ID
- NCT07882849
- Sponsor's protocol ID
- RP-2025/02-15
- Other IDs
- UKC-MB-KME-88/25
- Acronym
- TRIACS-MB
- Status last verified
- Oct 2026
- First submitted
- 12 Oct 2026
- Last updated
- 20 Oct 2026
- Results posted
- No
- Data monitoring committee
- No
- FDA-regulated drug
- No
- FDA-regulated device
- No
- Individual data sharing
- Yes: Anonymous study data will be shared via https://www.adp.fdv.uni-lj.si
Publications
Herman R, Mumma BE, Hoyne JD, Cooper BL, Johnson NP, Kisova T, Demolder A, Rafajdus A, Iring A, Palus T, Belmonte M, Barbato E, Baron SJ, Hatala R, Smith SW, Meyers HP, Sharkey SW, Bartunek J, Henry TD. AI-Enabled ECG Analysis Improves Diagnostic Accuracy and Reduces False STEMI Activations: A Multicenter U.S. Registry. JACC Cardiovasc Interv. 2026 Jan 26;19(2):145-156. doi: 10.1016/j.jcin.2025.10.018. Epub 2025 Oct 28.
PubMed 41158088Herman R, Meyers HP, Smith SW, Bertolone DT, Leone A, Bermpeis K, Viscusi MM, Belmonte M, Demolder A, Boza V, Vavrik B, Kresnakova V, Iring A, Martonak M, Bahyl J, Kisova T, Schelfaut D, Vanderheyden M, Perl L, Aslanger EK, Hatala R, Wojakowski W, Bartunek J, Barbato E. International evaluation of an artificial intelligence-powered electrocardiogram model detecting acute coronary occlusion myocardial infarction. Eur Heart J Digit Health. 2023 Nov 28;5(2):123-133. doi: 10.1093/ehjdh/ztad074. eCollection 2024 Mar.
PubMed 38505483Hellerman MB, Wang C, Zhang DT, Kalogeropoulos AP, Skopicki HA. Detecting Occlusion Myocardial Infarction with an AI-Powered ECG Model: A Retrospective Cohort Study. J Pers Med. 2026 Mar 24;16(4):174. doi: 10.3390/jpm16040174.
PubMed 42042541Hassannataj Joloudari J, Mojrian S, Nodehi I, Mashmool A, Kiani Zadegan Z, Khanjani Shirkharkolaie S, Alizadehsani R, Tamadon T, Khosravi S, Akbari Kohnehshari M, Hassannatajjeloudari E, Sharifrazi D, Mosavi A, Loh HW, Tan RS, Acharya UR. Application of artificial intelligence techniques for automated detection of myocardial infarction: a review. Physiol Meas. 2022 Aug 3;43(8). doi: 10.1088/1361-6579/ac7fd9.
PubMed 35803247Bdir S, Jaber M, Tanbouz O, Milhem F, Sarhan I, Bdair M, Alhroob T, Abu Alya W, Qneibi M. Artificial Intelligence for Myocardial Infarction Detection via Electrocardiogram: A Scoping Review. J Clin Med. 2025 Sep 25;14(19):6792. doi: 10.3390/jcm14196792.
PubMed 41095870Oke OA, Cavus N. A systematic review on the impact of artificial intelligence on electrocardiograms in cardiology. Int J Med Inform. 2025 Mar;195:105753. doi: 10.1016/j.ijmedinf.2024.105753. Epub 2024 Dec 9.
PubMed 39674006Bishop AJ, Nehme Z, Nanayakkara S, Anderson D, Stub D, Meadley BN. Artificial neural networks for ECG interpretation in acute coronary syndrome: A scoping review. Am J Emerg Med. 2024 Sep;83:1-8. doi: 10.1016/j.ajem.2024.06.026. Epub 2024 Jun 24.
PubMed 38936320Chan PZ, Ramli MAIB, Chew HSJ. Diagnostic Test Accuracy of artificial intelligence-assisted detection of acute coronary syndrome: A systematic review and meta-analysis. Comput Biol Med. 2023 Dec;167:107636. doi: 10.1016/j.compbiomed.2023.107636. Epub 2023 Oct 25.
PubMed 37925910
Study at a glance
From the registry record, last updated by the sponsor on 20 Oct 2026. Recruitment status is as the sponsor reported it: confirm with the study team.
- Official title
- Prospective External Validation of an Artificial Intelligence Electrocardiogram Model for Detection of Occlusion Myocardial Infarction in Patients Presenting to Emergency Department Triage With Suspected Acute Coronary Syndrome: The TRIACS Study
- ClinicalTrials.gov ID
- NCT07882849
- Acronym / sponsor ID
- TRIACS-MB / RP-2025/02-15
- Conditions
- Acute Coronary Syndromes, Chest Pain, Myocardial Infarction (MI), NSTEMI - Non-ST Segment Elevation MI, ST Elevation Myocardial Infarction, Occlusion Myocardial Infarction
- Keywords
- ECG, AI ECG analysis, ACS, Acute Coronary Syndrome, Chest Pain, Emergency department, PCI, Percutaneous coronary intervention
- Study type
- Observational
- Recruitment status
- Not yet recruiting
- Enrolment
- 1,000 participants (estimated)
- Lead sponsor
- University Medical Centre MariborUniversity, hospital or other
- Collaborators
- 3fs Nordic SI, d.o.o., University of Ljubljana
- Study start
- 19 Oct 2026 (estimated)
- Primary completion
- 19 Oct 2027 (estimated)
- Study completion
- 19 Nov 2027 (estimated)
- First posted
- 20 Oct 2026 (actual)
- Results
- Not posted in the source record
Clinical summary
The research question as the sponsor registered it. This describes what the study plans to test, not what it found.
Chest pain is among the most common reasons people attend an emergency department, and the electrocardiogram (ECG) recorded at triage is the first test used to decide who needs urgent treatment for a blocked coronary artery. Recognising the ECG patterns that indicate an acutely blocked artery is difficult. Many such patterns do not produce the classic ST-segment elevation, and identifying them reliably requires a level of ECG expertise that is not always available at the point of triage - particularly at night and among clinicians who are not cardiologists.
Artificial intelligence (AI) models that read the 12-lead ECG have been shown in several retrospective studies, mainly from Western Europe and the United States, to identify acute coronary occlusion more sensitively than conventional interpretation. These models are now in clinical use in a number of health systems. Their performance has not been assessed in a prospectively enrolled, unselected Slovenian emergency department population, and diagnostic accuracy commonly declines when a model is applied outside the setting in which it was developed.
TRIACS is a prospective, single-centre, observational external validation study in the Emergency Department of University Medical Centre Maribor. All consecutive adults presenting with symptoms suggestive of acute coronary syndrome who have a 12-lead ECG recorded at triage as part of routine care will be enrolled. Each triage ECG is analysed by the AI model. The model's output is withheld from all treating clinicians and plays no part in patient care; every patient receives standard guideline-directed management.
The main aim is to measure how accurately the AI model identifies angiographically confirmed occlusion myocardial infarction in this population, and to compare its accuracy with the triage decision actually made by the clinical team. The study is therefore designed to inform an adoption decision in either direction, rather than to demonstrate a predetermined conclusion.
Detailed description
Background. Approximately one quarter to one third of patients ultimately diagnosed with non-ST-elevation myocardial infarction have an acutely occluded culprit artery, and delayed invasive management in this group is associated with substantially higher mortality. The Fifth Universal Definition of Myocardial Infarction (2026) has broadened the formal electrocardiographic criteria for myocardial ischaemia to include ST-segment depression patterns, posterior and right-sided lead criteria, the Aslanger pattern, the South African flag sign and modified Sgarbossa criteria. Applying these criteria reliably requires expert-level ECG interpretation. The practical diagnostic gap at the point of triage is therefore now less a matter of criteria than of the expertise available to apply them.
AI-based ECG interpretation has demonstrated superior sensitivity to conventional millimetre criteria in retrospective multicentre cohorts and is in clinical deployment in several health systems. Prospective validation in unselected emergency department populations remains limited, and reported accuracy commonly declines on external validation.
- Design.** Prospective, single-centre, observational diagnostic accuracy study (external validation of an existing diagnostic model). Reported in accordance with TRIPOD+AI and STARD.
- Setting.** Emergency Department, University Medical Centre Maribor, Slovenia. Approximately 4,500-5,000 chest-pain presentations annually; approximately 910 ED-linked coronary procedures annually.
- Population.** Consecutive adults (≥18 years, no upper age limit) presenting to ED triage with symptoms suspicious for acute coronary syndrome in whom a 12-lead ECG is recorded and digitally captured at triage as part of routine care.
- Index test.** 'PM Cardio, Queen of Hearts', a CE-marked deep-learning model returning a binary verdict accompanied by a confidence tier and, where available, a continuous probability score. Only the first 12-lead ECG recorded at triage is analysed; one ECG per presentation. The decision threshold is fixed a priori at the manufacturer's on-label cut-off for the general chest-pain population. No threshold optimisation is performed for the primary analysis.
- Blinding.** The triage nurse, all treating emergency physicians and all consulting cardiologists are blinded to AI output. AI inference is performed in batch after the clinical episode has concluded. No AI output is entered into the medical record, is accessible from any clinical system, or is communicated to any member of the clinical team at any point. Reference-standard adjudication is performed by an independent adjudicator blinded to AI output.
- Reference standard.** The primary target condition is occlusion myocardial infarction (OMI): symptoms or signs consistent with myocardial ischaemia, AND acute myocardial injury demonstrated by a dynamic rise and/or fall in high-sensitivity cardiac troponin I with at least one value above the sex-specific 99th percentile upper reference limit, meeting the criteria for primary myocardial infarction under the Fifth Universal Definition of Myocardial Infarction (2026), AND angiographic evidence of an acute culprit lesion with either (a) TIMI flow grade 0-1, or (b) TIMI flow grade 2-3 with emergent or urgent percutaneous revascularisation.
Under the Fifth UDMI an isolated troponin elevation without ischaemic symptoms or signs constitutes myocardial injury, not infarction, and does not satisfy this definition. Primary myocardial infarction is no longer restricted to atherothrombosis and encompasses spontaneous coronary artery dissection, coronary embolism and vasospasm; such cases meet the reference standard where the angiographic criterion is satisfied. Myocardial injury with non-obstructive coronary arteries (MINOCA, a working rather than final diagnosis under the Fifth UDMI) is adjudicated as OMI-negative and reported as a named category.
Because coronary angiography is not performed in all patients, a composite reference standard is used with the verification route pre-specified:
- Angiographed during the index episode: adjudicated on angiographic and biomarker data as above.
- Not angiographed: adjudicated OMI-negative where serial hs-cTnI testing (ESC 0/1 h algorithm) rules out myocardial infarction, the discharge diagnosis excludes acute myocardial infarction, and 30-day follow-up shows no myocardial infarction, unplanned revascularisation or cardiac death.
- Biomarker-positive, managed non-invasively: adjudicated individually by the blinded adjudicator using all available clinical, biomarker and imaging data.
Ambiguous or discordant cases are reviewed by a second independent adjudicator, with disagreement resolved by consensus. Inter-adjudicator agreement is reported (Cohen's kappa).
- Comparators.** (i) The standard-of-care triage decision actually made by the clinical team, defined as urgent cardiology referral or catheterisation laboratory activation within 10 minutes of ECG availability, extracted from routine records; (ii) conventional ST-elevation millimetre criteria; (iii) the full electrocardiographic ischaemia criteria of the Fifth UDMI. Comparators (ii) and (iii) are applied to the same triage ECG by a reviewer blinded to AI output, clinical course and outcome.
- Follow-up.** 30 days from index presentation, via hospital records and national registry linkage.
Study design
- Observational model
- Cohort
- Time perspective
- Prospective
- Number of arms
- 1
| Arm | Type | Receives |
|---|---|---|
| Suspected acute coronary syndrome at ED triage | – | –Consecutive adults with a digitally captured triage 12-lead ECG. No intervention assigned; AI analysis performed on routinely acquired ECGs with output withheld from clinicians. |
Who may qualify
- Ages
- 18 years and older
- Sex
- All
- Healthy volunteers
- Not accepted
- Age groups
- Adult, Older adult
- Study population
- All adults presenting to ED of UKC Maribor, triaged as chest pain that have ECG recorded and agree to be included into study.
Key inclusion criteria (3 listed)
- Age ≥18 years
- Presentation to ED triage with symptoms suspicious for acute coronary syndrome, including chest discomfort, dyspnoea, epigastric pain, or other anginal equivalents
- Standard 12-lead ECG recorded at triage as part of routine care, available in machine-readable or interpretable image format
Key exclusion criteria (6 listed)
- Traumatic cause of chest pain
- Ventricular arrhythmia or ongoing cardiopulmonary resuscitation at the time of the triage ECG
- ECG of insufficient technical quality for analysis (recorded and reported as an exclusion count)
- Patients with prehospital STEMI diagnosis transported directly to the catheterisation laboratory and bypassing ED triage
- Patients for whom laboratory results are not retrievable by UKC MB
Eligibility is decided by the study team after screening; this overview does not establish that a patient qualifies. Full inclusion and exclusion criteria.
Primary outcome
Diagnostic accuracy of the AI-ECG model for angiographically confirmed occlusion myocardial infarction
Time frame: From triage ECG to 7 and 30 days after index presentation
Sensitivity, specificity, positive predictive value, negative predictive value, overall accuracy and area under the receiver-operating characteristic curve of the AI model's verdict on the first 12-lead ECG recorded at triage, against the pre-specified composite reference standard for OMI. The decision threshold is fixed a priori at the manufacturer's on-label cut-off; no threshold optimisation is performed.
2 secondary outcomes
Comparative diagnostic accuracy versus formal STEMI criteria
Time frame: From triage ECG to 7 and 30 days after index presentation
Versus conventional ST-elevation millimetre criteria
Time frame: From triage ECG to 7 and 30 days after index presentation
Results and safety
This study is still under way. Results are not available in the source record yet.
Research links
- Official record on ClinicalTrials.govLast updated by the sponsor 20 Oct 2026.
- Background: PubMed 41158088Herman R, Mumma BE, Hoyne JD, Cooper BL, Johnson NP, Kisova T, Demolder A, Rafajdus A, Iring A, Palus T, Belmonte M, Barbato E, Baron SJ, Hatala R, Smith SW, Meyers HP, Sharkey SW, Bartunek J, Henry TD. AI-Enabled ECG Analysis Improves Diagnostic Accuracy and Reduces False STEMI Activations: A Multicenter U.S. Registry. JACC Cardiovasc Interv. 2026 Jan 26;19(2):145-156. doi: 10.1016/j.jcin.2025.10.018. Epub 2025 Oct 28.
- Background: PubMed 38505483Herman R, Meyers HP, Smith SW, Bertolone DT, Leone A, Bermpeis K, Viscusi MM, Belmonte M, Demolder A, Boza V, Vavrik B, Kresnakova V, Iring A, Martonak M, Bahyl J, Kisova T, Schelfaut D, Vanderheyden M, Perl L, Aslanger EK, Hatala R, Wojakowski W, Bartunek J, Barbato E. International evaluation of an artificial intelligence-powered electrocardiogram model detecting acute coronary occlusion myocardial infarction. Eur Heart J Digit Health. 2023 Nov 28;5(2):123-133. doi: 10.1093/ehjdh/ztad074. eCollection 2024 Mar.
- Background: PubMed 42042541Hellerman MB, Wang C, Zhang DT, Kalogeropoulos AP, Skopicki HA. Detecting Occlusion Myocardial Infarction with an AI-Powered ECG Model: A Retrospective Cohort Study. J Pers Med. 2026 Mar 24;16(4):174. doi: 10.3390/jpm16040174.
- Background: PubMed 35803247Hassannataj Joloudari J, Mojrian S, Nodehi I, Mashmool A, Kiani Zadegan Z, Khanjani Shirkharkolaie S, Alizadehsani R, Tamadon T, Khosravi S, Akbari Kohnehshari M, Hassannatajjeloudari E, Sharifrazi D, Mosavi A, Loh HW, Tan RS, Acharya UR. Application of artificial intelligence techniques for automated detection of myocardial infarction: a review. Physiol Meas. 2022 Aug 3;43(8). doi: 10.1088/1361-6579/ac7fd9.
- Background: PubMed 41095870Bdir S, Jaber M, Tanbouz O, Milhem F, Sarhan I, Bdair M, Alhroob T, Abu Alya W, Qneibi M. Artificial Intelligence for Myocardial Infarction Detection via Electrocardiogram: A Scoping Review. J Clin Med. 2025 Sep 25;14(19):6792. doi: 10.3390/jcm14196792.
- Background: PubMed 39674006Oke OA, Cavus N. A systematic review on the impact of artificial intelligence on electrocardiograms in cardiology. Int J Med Inform. 2025 Mar;195:105753. doi: 10.1016/j.ijmedinf.2024.105753. Epub 2024 Dec 9.
- Background: PubMed 38936320Bishop AJ, Nehme Z, Nanayakkara S, Anderson D, Stub D, Meadley BN. Artificial neural networks for ECG interpretation in acute coronary syndrome: A scoping review. Am J Emerg Med. 2024 Sep;83:1-8. doi: 10.1016/j.ajem.2024.06.026. Epub 2024 Jun 24.
- Background: PubMed 37925910Chan PZ, Ramli MAIB, Chew HSJ. Diagnostic Test Accuracy of artificial intelligence-assisted detection of acute coronary syndrome: A systematic review and meta-analysis. Comput Biol Med. 2023 Dec;167:107636. doi: 10.1016/j.compbiomed.2023.107636. Epub 2023 Oct 25.
- Individual participant dataSharing planned: Yes. Anonymous study data will be shared via https://www.adp.fdv.uni-lj.si
Related trials for Acute Coronary Syndromes
Recruiting
CT Stress Myocardial Perfusion, Fractional Flow Reserve and Angiography in Patients With Stable Chest Pain Syndromes
Coronary Artery Disease, Chest Syndrome
Recruiting
WIN Ratio Analysis to Determine a Strategy of Non- Invasive SUpport for Respiratory Failure in the EmeRgency Department
Acute Hypoxic Respiratory Failure, Acute Respiratory Distress Syndrome (ARDS)
RecruitingPhase 4
Early Single Antiplatelet Therapy After IVUS-Guided PCI in Acute Coronary Syndrome
Acute Coronary Syndromes, STEMI
Recruiting
AI-Assisted Interpretation of Cardiac CT in the Emergency Department
Chest Pain, Acute Coronary Syndrome
Enrolling by invitation
Optical Coherence Tomography in Acute Vessel Evaluation
Ischemic Heart Disease, Ischemic Coronary Artery Disease
Referral and contact
AzTrials does not enrol participants or take referrals. Contact the study team or a site listed below; they decide who can take part.
Luka Petravić, MD principal investigator
UKC Maribor
1 site in 1 country: see every site and its status.
Questions about this trial
What is this clinical trial studying?
An observational study of Acute Coronary Syndromes and Chest Pain, run by University Medical Centre Maribor at 1 site in Slovenia.
The full registered description is in the About this study section and the official record.
Who may be eligible to take part?
The sponsor lists 3 inclusion and 6 exclusion criteria. Eligibility is decided by the study team after screening, so read the full criteria and contact a listed site.
Is this clinical trial currently recruiting?
The record says: opening soon, as of the sponsor's last update on 20 Oct 2026. Status can change between updates, so confirm with the study contact before making plans.
Where is the study taking place?
The record lists 1 site in Slovenia. Whether a site is taking part right now should be confirmed with that site. See every site.
What are the possible benefits and risks?
They depend on the study and the treatment, and no study can promise a benefit. The study team explains the known risks during informed consent, before anyone joins; the record itself does not list them.
Are results available for this study?
This study is still under way. Results are not available in the source record yet.
How can a doctor refer a patient or contact the research team?
Contact Luka Petravić, MD, MD (+38623210000, luka.petravic@ukc-mb.si), or a site near the patient. AzTrials does not enrol participants or take referrals.
What does the primary outcome measure?
Diagnostic accuracy of the AI-ECG model for angiographically confirmed occlusion myocardial infarction (time frame: From triage ECG to 7 and 30 days after index presentation). It is what the study was designed to measure, not a finding.
Where can I read the original study information?
On the ClinicalTrials.gov record for NCT07882849, which also lists 8 linked publications.
Does a listing mean the treatment is approved or proven to work?
No. A listing means the study is registered. It does not show that the treatment is safe, works, or is approved by a regulator.
How often is this information updated?
AzTrials copies changed records from ClinicalTrials.gov every night and reloads the whole registry every week. The sponsor last updated this record on 20 Oct 2026.