RESUMO
A 44-year-old male patient presented to our clinic with nocturnal palpitations. The smartwatch electrocardiogram (ECG) demonstrated ST-segment deviation and non-sustained ventricular tachycardias. The patient suffered from coronary one-vessel disease. The coronary angiography revealed de novo proximal left anterior descending stenosis and in-stent restenosis, which necessitated coronary stenting. However, the recurrent palpitations reappeared 2â¯weeks later. The smartwatch ECGs again demonstrated ST-segment deviation and wide QRS tachycardia, reassessments with coronary angiography and magnetic resonance imaging revealed no relevant ischemia. The patient was diagnosed with vasospastic angina. His symptoms disappeared after percutaneous administration of nitrates.This case highlights the utility of smartwatch ECGs to support clinical diagnosis, decisions, and follow-up in the case of ischemic attacks. Learning objective: This case report highlights the importance of an adequate anamnesis and the meaningful use of wearable devices. The smartwatch enables the patient to record electrocardiograms (ECGs) easily and immediately during symptoms. That is one of the major advantages of this new technology, because we can even detect events that are too subtle to be detected in the normal 24-h Holter ECG. This case highlights the utility of smartwatch ECGs to support clinical diagnosis and decisions by establishing a valid symptom-ECG correlation.
RESUMO
Artificial intelligence (AI) is developing rapidly in the medical technology field, particularly in image analysis. ECG-diagnosis is an image analysis in the sense that cardiologists assess the waveforms presented in a 2-dimensional image. We hypothesized that an AI using a convolutional neural network (CNN) may also recognize ECG images and patterns accurately. We used the PTB ECG database consisting of 289 ECGs including 148 myocardial infarction (MI) cases to develop a CNN to recognize MI in ECG. Our CNN model, equipped with 6-layer architecture, was trained with training-set ECGs. After that, our CNN and 10 physicians are tested with test-set ECGs and compared their MI recognition capability in metrics F1 (harmonic mean of precision and recall) and accuracy. The F1 and accuracy by our CNN were significantly higher (83 ± 4%, 81 ± 4%) as compared to physicians (70 ± 7%, 67 ± 7%, P < 0.0001, respectively). Furthermore, elimination of Goldberger-leads or ECG image compression up to quarter resolution did not significantly decrease the recognition capability. Deep learning with a simple CNN for image analysis may achieve a comparable capability to physicians in recognizing MI on ECG. Further investigation is warranted for the use of AI in ECG image assessment.
Assuntos
Algoritmos , Inteligência Artificial , Bases de Dados Factuais , Aprendizado Profundo , Eletrocardiografia/métodos , Infarto do Miocárdio/diagnóstico , Redes Neurais de Computação , Humanos , Infarto do Miocárdio/diagnóstico por imagemRESUMO
Poor survival has been demonstrated after ventricular assist device (VAD) implantation for Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) profile 1 and 2 patients compared with more stable levels. However, risk factors within this high-risk cohort have not been determined so far. The aim of the present study was to identify risk factors associated with this very high mortality rate. Between February 1993 and January 2013, 298 patients underwent VAD implantation in our institution. One hundred nine patients were in INTERMACS level 1 and 49 patients were in INTERMACS level 2 and were therefore defined as hemodynamically critical (overall 158 patients). Assist devices implanted were: HVAD HeartWare n = 18; Incorâ n = 11; VentrAssistâ n = 2; DeBakeyâ n = 22; and pulsatile systems n = 105. After cumulative support duration of 815.35 months, Kaplan-Meier analysis revealed a survival of 63.9, 48.8, and 40.3% at 1, 6, and 12 months, respectively. Cox regression analyses identified age > 50 (P = 0.001, odds ratio [OR] 2.48), white blood cell count > 13.000/µL (P = 0.01, OR 2.06), preoperative renal replacement therapy (P = 0.001, OR 2.63), and postcardiotomy failure (P < 0.001, OR 2.79) as independent predictors of mortality. Of note, last generation VADs were not associated with significantly better 6-month survival (P = 0.59). Patients without the aforementioned risk factors could yield a survival of 79.2% at 6 months. This single-center experience shows that VAD implantation in hemodynamically unstable patients generally results in poor early outcome, even in third-generation pumps. However, avoiding the aforementioned risk factors could result in improved outcome.