Reliability of Minimally Trained Operator's Velocity-Time Integral Measurement Guided by Artificial Intelligence VTI

Официальное название

Reliability of Minimally Trained Operator's Velocity-Time Integral Measurement Guided by Artificial Intelligence (MiniTrained-VTI)

Обзор

Stroke volume is a major determinant of tissue perfusion and therefore a key parameter to monitor in patients with hemodynamic instability and hypoperfusion. Left Ventricular Outflow Tract (LVOT) Velocity-Time Integral (VTI) measured using pulsed wave Doppler is widely used as an estimation of stroke volume and should be a competence required for every Intensive Care Unit (ICU) physician. Recently, research in Artificial Intelligence (AI) applied to medical imaging constituted a breakthrough in the acquisition of images. The goal of the present study is to characterize and quantify the reliability and reproducibility of LVOT VTI measurements by comparing the measures obtained by minimally-trained operators and expert physicians, guided by UltraSight AI software.

Подробное описание

The main goal of Intensive Care Unit (ICU) physicians is to ensure cellular oxygenation by maintaining adequate organ perfusion in their patients. Stroke volume is a major determinant of tissue perfusion and therefore a key parameter to monitor in patients with hemodynamic instability. Left Ventricular Outflow Tract (LVOT) Velocity-Time Integral (VTI) measured using pulsed wave Doppler is widely used as an estimation of stroke volume to assess hemodynamic modifications. This value reflects the stroke distance, which varies proportionately to stroke volume in case of hemodynamic variations resulting from therapeutic interventions (fluid administration, vasoactive drugs…) or disease processes. An increase in stroke volume (or LVOT VTI) is expected in response to fluid administration and attests for its efficacy. A lack of increase indicates that the cardiovascular system is no longer fluid-responsive, and that fluid administration is not improving tissue perfusion and creates congestion. Therefore, measuring aortic VTI should be a competence required for every ICU physician. However, international ICU guidelines on echocardiography do not consider LVOT VTI measurement as a basic skill but rather as a competence of advanced operators. More recently, the European Society of Intensive Care Medicine published expert recommendations on echocardiography, setting the evaluation of LVOT VTI as basic skill but with a weak recommendation, lacking published evidence to support this statement.

The main difficulty in measuring LVOT VTI is obtaining an adequate apical 5-chamber view.

Recently, research in artificial intelligence (AI) applied to medical imaging constituted a breakthrough in the acquisition of images. UltraSight is a company specialized in AI applied to echocardiography. Their software is based on neural network using machine learning to analyse extremely precisely the image obtained by an operator. The software indicates to the operator in real time on-screen how to optimize the image by mobilizing the probe until the desired view is correctly obtained, with the best quality.

The main objective of the present study is to characterize and to quantify the reliability and reproducibility of LVOT VTI measurements by comparing the measures obtained by minimally trained operators and experts, using an ultrasound platform equipped with real-time AI-based guidance (UltraSight). If interchangeability of minimally trained operators and expert measurements can be demonstrated, this will constitute a strong basis to upgrade the measurement of LVOT VTI as a basic competence in critical care ultrasound. The secondary objectives are to assess the concordance of therapeutic decisions made by the ICU clinician in charge of the patient (i.e.: continue or interrupt fluid administration) based on the VTI variation obtained by the minimally-trained operator, and that based on the VTI variation obtained by the expert, the agreement of the absolute value of the measure of LVOT VTI obtained by the minimally trained operators and the experts, the correlation between the measures of the VTI variation (% change following a fluid challenge of 250 mL or a passive leg-raising test) between the minimally-trained operators and those obtained by experts.

Вмешательства

• Другое Fluid challenge (cristalloids) OR passive leg raising
  Patients in whom fluid administration is considered necessary, based on hypoperfusion criteria will be included in the trial. One member of group A and one of group B will proceed independently to evaluate LVOT VTI, guided by the UltraSight AI software to obtain the best 5-chamber view. The measure of LVOT VTI will be calculated as the average of three consecutive cardiac cycles. The order of acquisition between group A and B will be randomized. Each operator will be blinded to the values obta

Первичные конечные точки

• Relative difference expressed in percentage of LVOT VTI measurement evaluated by the minimally trained operators as compared with LVOT VTI measurement by experts. [Срок оценки: Less than 60 minutes]

Вторичные конечные точки (7)

• The relative difference expressed in percentage of LVOT VTI measurement, pre-fluid challenge, evaluated by the minimally trained operators as compared with those obtained by experts. [Срок оценки: Less than 60 minutes]
• The relative difference expressed in percentage of LVOT VTI measurement, post-fluid challenge, evaluated by the minimally trained operators as compared with those obtained by experts. [Срок оценки: Less than 60 minutes]
• Difference in the VTI variation [i.e.: % change after a fluid challenge of 250 mL, or a passive leg-raising test] obtained by the minimally-trained operators and that obtained by experts. [Срок оценки: Less than 60 minutes]
• Correlation of the measure of the VTI variation before and after a fluid challenge of 250 mL or after a passive leg-raising test between the minimally trained operators and the experts. [Срок оценки: Less than 60 minutes]
• Difference of the absolute value of the measure of LVOT VTI obtained by the minimally trained operators and the experts. [Срок оценки: Less than 60 minutes]
• Concordance in therapeutic decision by the blinded attending physician based on the measure of the VTI variation by the minimally trained operators compared to experts. [Срок оценки: Less than 60 minutes]
• Percentage of pairs (minimally-trained/expert) with difference of measurement: o < -14% o [-14%; 14%] o > 14% [Срок оценки: Less than 60 minutes]

Критерии участия

Критерии включения

All patients aged 18 and more

Hospitalized in ICU, in whom fluid administration is considered necessary by the clinician in charge, based on the presence of hypoperfusion criterion:

• >10% decrease in mean arterial pressure with respect to baseline value
• Skin mottling, oliguria (<0,5 ml/kg/h)
• change in the level of consciousness
• hyperlactatemia
• decrease in central venous oxygen saturation Affiliation to a French social security system (beneficiary or legal) Participant's or next of kin non-opposition or emergency procedure

Критерии исключения

Patients with atrial fibrillation, due to the higher variability in LVOT VTI; Patient on Emergency Medical Assistance; Patient under guardianship, curatorship, deprived of liberty.

Критерии приведены из реестра в оригинале (на английском). Окончательную оценку соответствия проводит исследовательский центр.

Здоровые добровольцы: Нет

Дизайн исследования

Распределение

Рандомизированное

Модель

Одна группа

Маскирование

Тройное слепое

Основная цель

Другое

Центры проведения

Франция · 3 центра

• CHU de Limoges — Limoges
• Hôpital Lariboisière - APHP — Paris
• Hôpital européen Georges Pompidou - APHP — Paris

Публикации

• Cecconi M, De Backer D, Antonelli M, Beale R, Bakker J, Hofer C, Jaeschke R, Mebazaa A, Pinsky MR, Teboul JL, Vincent JL, Rhodes A. Consensus on circulatory shock and hemodynamic monitoring. Task force of the European Society of Intensive Care Medicine. Intensive Care Med. 2014 Dec;40(12):1795-815. doi: 10.1007/s00134-014-3525-z. Epub 2014 Nov 13. PMID 25392034
• Mercado P, Maizel J, Beyls C, Titeca-Beauport D, Joris M, Kontar L, Riviere A, Bonef O, Soupison T, Tribouilloy C, de Cagny B, Slama M. Transthoracic echocardiography: an accurate and precise method for estimating cardiac output in the critically ill patient. Crit Care. 2017 Jun 9;21(1):136. doi: 10.1186/s13054-017-1737-7. PMID 28595621
• Expert Round Table on Echocardiography in ICU. International consensus statement on training standards for advanced critical care echocardiography. Intensive Care Med. 2014 May;40(5):654-66. doi: 10.1007/s00134-014-3228-5. Epub 2014 Mar 11. No abstract available. PMID 24615559
• Robba C, Wong A, Poole D, Al Tayar A, Arntfield RT, Chew MS, Corradi F, Doufle G, Goffi A, Lamperti M, Mayo P, Messina A, Mongodi S, Narasimhan M, Puppo C, Sarwal A, Slama M, Taccone FS, Vignon P, Vieillard-Baron A; European Society of Intensive Care Medicine task force for critical care ultrasonography*. Basic ultrasound head-to-toe skills for intensivists in the general and neuro intensive care PMID 34787687
• Mor-Avi V, Khandheria B, Klempfner R, Cotella JI, Moreno M, Ignatowski D, Guile B, Hayes HJ, Hipke K, Kaminski A, Spiegelstein D, Avisar N, Kezurer I, Mazursky A, Handel R, Peleg Y, Avraham S, Ludomirsky A, Lang RM. Real-Time Artificial Intelligence-Based Guidance of Echocardiographic Imaging by Novices: Image Quality and Suitability for Diagnostic Interpretation and Quantitative Analysis. Circ Ca PMID 37955139
• Jozwiak M, Mercado P, Teboul JL, Benmalek A, Gimenez J, Depret F, Richard C, Monnet X. What is the lowest change in cardiac output that transthoracic echocardiography can detect? Crit Care. 2019 Apr 11;23(1):116. doi: 10.1186/s13054-019-2413-x. PMID 30971307
• Levy N, Meslin S, Barthelemy R, Benremily F, Bourgeois C, Bourzeix P, Chousterman B, Djadi-Prat J, Ep A, Kezar A, Laidet C, Lanoy E, Leopold V, Pereira H, Plateker O, Rivoalen AS, de Roquetaillade C, Vignon P, Bruno J, Cholley B. Reliability of minimally trained operator's left ventricular outflow tract velocity-time integral measurement guided by artificial intelligence: protocol for a multicentr PMID 41125264

Идентификаторы

NCT: NCT06486467 · APHP230510 · 2022-A02820-43

Первоисточники (государственные реестры)

• ClinicalTrials.gov (NIH, США)

Открыть это исследование на ClinicalTrials.gov ↗