Pulse wave signal-driven machine learning for identifying left ventricular enlargement in heart failure patients.

BACKGROUND: Left ventricular enlargement (LVE) is a common manifestation of cardiac remodeling that is closely associated with cardiac dysfunction, heart failure (HF), and arrhythmias. This study aimed to propose a machine learning (ML)-based strategy to identify LVE in HF patients by means of pulse wave signals. METHOD: We constructed two high-quality pulse wave datasets comprising a non-LVE group and an LVE group based on the 264 HF patients. Fourier series calculations were employed to determine if significant frequency differences existed between the two datasets, thereby ensuring their validity. Then, the ML-based identification was undertaken by means of classification and regression models: a weighted random forest model was employed for binary classification of the datasets, and a densely connected convolutional network was utilized to directly estimate the left ventricular diastolic diameter index (LVDdI) through regression. Finally, the accuracy of the two models was validated by comparing their results with clinical measurements, using accuracy and the area under the receiver operating characteristic curve (AUC-ROC) to assess their capability for identifying LVE patients. RESULTS: The classification model exhibited superior performance with an accuracy of 0.91 and an AUC-ROC of 0.93. The regression model achieved an accuracy of 0.88 and an AUC-ROC of 0.89, indicating that both models can quickly and accurately identify LVE in HF patients. CONCLUSION: The proposed ML methods are verified to achieve effective classification and regression with good performance for identifying LVE in HF patients based on pulse wave signals. This study thus demonstrates the feasibility and potential of the ML-based strategy for clinical practice while offering an effective and robust tool for diagnosing and intervening ventricular remodeling.

Impacts of respiratory fluctuations on cerebral circulation: a machine-learning-integrated 0-1D multiscale hemodynamic model.

Objective. This study aims to accurately identify the effects of respiration on the hemodynamics of the human cardiovascular system, especially the cerebral circulation.Approach: we have developed a machine learning (ML)-integrated zero-one-dimensional (0-1D) multiscale hemodynamic model combining a lumped-parameter 0D model for the peripheral vascular bed and a one-dimensional (1D) hemodynamic model for the vascular network.In vivomeasurement data of 21 patients were retrieved and partitioned into 8000 data samples in which respiratory fluctuation (RF) of intrathoracic pressure (ITP) was fitted by the Fourier series. ML-based classification and regression algorithms were used to examine the influencing factors and variation trends of the key parameters in the ITP equations and the mean arterial pressure. These parameters were employed as the initial conditions of the 0-1D model to calculate the radial artery blood pressure and the vertebral artery blood flow volume (VAFV).Main results: during stable spontaneous respiration, the VAFV can be augmented at the inhalation endpoints by approximately 0.1 ml s-1for infants and 0.5 ml s-1for adolescents or adults, compared to those without RF effects. It is verified that deep respiration can further increase the ranges up to 0.25 ml s-1and 1 ml s-1, respectively.Significance. This study reveals that reasonable adjustment of respiratory patterns, i.e. in deep breathing, enhances the VAFV and promotes cerebral circulation.

Machine learning predicts blood lactate levels in children after cardiac surgery in paediatric ICU.

BACKGROUND: Although serum lactate levels are widely accepted markers of haemodynamic instability, an alternative method to evaluate haemodynamic stability/instability continuously and non-invasively may assist in improving the standard of patient care. We hypothesise that blood lactate in paediatric ICU patients can be predicted using machine learning applied to arterial waveforms and perioperative characteristics. METHODS: Forty-eight post-operative children, median age 4 months (2.9-11.8 interquartile range), mean baseline heart rate of 131 beats per minute (range 33-197), mean lactate level at admission of 22.3 mg/dL (range 6.3-71.1), were included. Morphological arterial waveform characteristics were acquired and analysed. Predicting lactate levels was accomplished using regression-based supervised learning algorithms, evaluated with hold-out cross-validation, including, basing prediction on the currently acquired physiological measurements along with those acquired at admission, as well as adding the most recent lactate measurement and the time since that measurement as prediction parameters. Algorithms were assessed with mean absolute error, the average of the absolute differences between actual and predicted lactate concentrations. Low values represent superior model performance. RESULTS: The best performing algorithm was the tuned random forest, which yielded a mean absolute error of 3.38 mg/dL when predicting blood lactate with updated ground truth from the most recent blood draw. CONCLUSIONS: The random forest is capable of predicting serum lactate levels by analysing perioperative variables, including the arterial pressure waveform. Thus, machine learning can predict patient blood lactate levels, a proxy for haemodynamic instability, non-invasively, continuously and with accuracy that may demonstrate clinical utility.

Paracorporeal Support in Pediatric Patients: The Role of the Patient-Device Interaction.

BACKGROUND: Ventricular assist devices are important in the treatment of pediatric heart failure. Although paracorporeal pulsatile (PP) devices have historically been used, there has been increased use of paracorporeal continuous (PC) devices. We sought to compare the outcomes of children supported with a PP or PC, or combination of devices. METHODS: A retrospective review (2005 to 2019) was made of patients less than 19 years of age from a single center who received a PC, PP, or combination of devices. Patient characteristics were compared between device strategies, and Kaplan-Meier survival analysis was performed. RESULTS: Sixty-six patients were included: 62% male; 62% non-congenital heart disease; median age 0.9 years (interquartile range, 0.2 to 4.9); and median weight 8.5 kg (interquartile range, 4.3 to 17.7 kg). The PC devices were used in 45% of patients, PP in 35%, and a combination in 20%. Patients on PC devices had a lower median weight (P = .02) and a higher proportion of congenital heart disease (P = .02), and more patients required pre-ventricular assist device dialysis (P = .01). There was no difference in pre-ventricular assist device extracorporeal membrane oxygenation use (P = .15). There was a difference in survival among the three device strategies (P = .02). CONCLUSIONS: Differences in survival were evident, with patients on PC support having worse outcomes. Transition from PC to a PP devices was associated with a survival advantage. These findings may be driven by differences in patient characteristics across device strategies. Further studies are required to confirm these findings and to better understand the interaction between patient characteristics and device options.

Impact of Atrial Fibrillation on Fontan Circulation: Fontan Computational Model.

BACKGROUND: Patients with Fontan circulation may have heart failure resulting in atrial fibrillation during the late phase. Inotropic effects to ameliorate hemodynamics on the Fontan circulation are not well understood, especially when in atrial fibrillation. This study was performed to determine whether dobutamine therapy in patients with Fontan circulation has limited effects on improving hemodynamics. METHODS: Lumped computational models (sinus and atrial fibrillation) were used, including biventricular, atriopulmonary connection, and extracardiac total cavopulmonary connection Fontan models. The condition of atrial fibrillation including lack of atrial beat, irregular ventricular contraction, and time-varying elastance for the ventricle was introduced. A different dose of dobutamine was given by varying the elastance of the ventricle, heart rate, and peripheral resistance. RESULTS: In all models, the cardiac output decreased by 22.5% to 25.8% in atrial fibrillation. At 10 µg · kg-1 · min-1 dobutamine in sinus rhythm, the cardiac output increased by 32.3% in the biventricular model but by only 9.2% (P < .001) and 9.1% (P < .001) in the atriopulmonary connection and total cavopulmonary connection Fontan models, respectively. At 10 µg · kg-1 · min-1 dobutamine in atrial fibrillation, the percent increase in the cardiac output in the Fontan circulation (11.8% increase in atriopulmonary connection, P < .001; and 11.9% increase in total cavopulmonary connection, P < .001) was significantly less than that in the biventricular circulation (32.3% increase). CONCLUSIONS: In the Fontan circulation, atrial fibrillation itself reduced the cardiac output by approximately 25%, and dobutamine had a limited effect on increasing the cardiac output, especially when in atrial fibrillation. Maintaining sinus rhythm in patients with Fontan circulation is very important.

Coarctation of the aorta with total anomalous pulmonary venous connection: a case report.

Total anomalous pulmonary venous connection (TAPVC) and coarctation of the aorta (CoA) rarely occur together. In affected patients, blood is supplied to the lower body by saturated ductal flow. Preoperative echocardiography may not show an acceleration of flow at the isthmus (coarctation), and the oxygen saturation (SpO2) at the feet may be satisfactory. Consequently, the severity of CoA is often underestimated before performing surgery. A 6-day-old boy weighing 2.6 kg with a diagnosis of supracardiac TAPVC was referred for surgical correction of his anomaly. The atrial septal defect (ASD) was 6.7 mm in diameter. There was a large patent ductus arteriosus (PDA) without flow acceleration at the preductal entry into the descending aorta. Only the TAPVC repair was planned, but immediately following ligation of the large PDA, the blood pressure in the lower extremity dropped to around 30 mmHg. The ligation was removed. The reason for the blood pressure discrepancy between the upper and the lower body was not clear as there was no arterial line in the upper extremity and a 6.7-mm-diameter ASD can support sufficient blood flow to the lower body without the PDA. A suspected CoA was found and repaired, followed by the TAPVC repair. Caution is necessary when repairing a TAPVC and coexisting large PDA as the severity of the CoA can easily be underestimated due to nonsignificant flow acceleration.

Markers of peripheral perfusion during high-flow regional cerebral perfusion for aortic arch repair.

OBJECTIVES: High-flow regional cerebral perfusion (HFRCP) provides cerebral and somatic oxygen delivery through collateral vessels during aortic arch repair in small children; however, optimal flow conditions during HFRCP have not been established. We sought to identify markers of peripheral perfusion during HFRCP. METHODS: Between 2009 and 2016, in total 20 consecutive pediatric patients undergoing aortic arch repair with HFRCP were enrolled in this prospective, observational study. Median age was 20 days (range, 6-116 days); median body weight was 2.77 kg (range, 1.8-4.98 kg). Oxygen delivery ratio (Do2R) was calculated as the oxygen delivery during HFRCP divided by the oxygen delivery before HFRCP. Regional oxygen saturations on the forehead and on the thigh (rSo2T) were monitored during HFRCP, and postoperative creatinine kinase and lactate concentrations were measured as postoperative outcomes. Multivariate analyses were performed to clarify the effectiveness of Do2R and rSo2T as markers of peripheral perfusion during HFRCP. RESULTS: No deaths or neurologic impairments occurred. Multivariate analysis showed that the lowest rSo2T (P = .005) and cardiopulmonary bypass time (P = .012) predicted postoperative creatinine kinase concentration. Do2R was the only factor to predict postoperative lactate concentration (P < .001). Receiver operating characteristic analysis showed that Do2R less than 0.66 predicted risk of high postoperative lactate concentration (>5.0 mmol/L), with area under the curve of 0.95. CONCLUSIONS: For aortic arch repair in small children, rSo2T and Do2R during HFRCP are useful markers for predicting peripheral perfusion. Maintaining higher Do2R during HFRCP minimizes postoperative increases in lactate and creatinine kinase concentrations.

Aortic Arch Homograft Reconstruction of Nonconfluent Pulmonary Arteries During Extracardiac Fontan.

Reconstruction of nonconfluent pulmonary arteries during Fontan completion is a challenging technical issue. In this case report, we describe the use of an aortic homograft, including the aortic arch, to complete a Fontan and reconstruct the pulmonary artery confluence in a child with discontinuous pulmonary arteries and bilateral superior caval veins who had undergone bilateral unidirectional Glenn palliation. The configuration of the aortic homograft was ideal to ensure laminar flow from the inferior vena cava to both pulmonary arteries and in maintaining durable elastance posterior to the native aorta.

Mid-term result of atrioventricular valve replacement in patients with a single ventricle.

OBJECTIVES: Atrioventricular valve replacement is the last option to treat the atrioventricular valve regurgitation in single ventricle. This study investigates the mid-term outcomes of the atrioventricular valve replacement based on the Japan Cardiovascular Surgery Database registry. METHODS: From 2008 to 2014, 56 patients [34 males (61%) and 22 females (39%)] with a single ventricular circulation, underwent atrioventricular valve replacement. Questionnaires were collected to review operative data, mid-term mortality, morbidity and redo replacement. Risk factor analysis was performed by the Cox regression model for death and redo replacement. RESULTS: Heterotaxy, a right systemic ventricle and a common atrioventricular valve was present in 46% (26/56), 64% and 57% of patients, respectively. The most common timings for atrioventricular valve replacement were the interstage between the second and third palliations (34%) and after the Fontan operation (34%). Twenty died during the 3.7 ± 2.6-year follow-up. Eleven received redo atrioventricular replacement. The cumulative incidences of redo atrioventricular valve replacement and survival at 3 years were 20% [95% confidence interval (CI) 9-30] and 66% (95% CI 55-80), respectively. Univariable Cox regression analysis revealed that a tricuspid valve was a risk factor for redo valve replacement [hazard ratio (HR) 6.76, 95% CI 1.79-25.6; P = 0.005] and that young age was a risk factor for death (HR 0.77, 95% CI 0.62-0.96; P = 0.019). Fourteen patients required a pacemaker implantation. CONCLUSIONS: Valve replacement for uncontrollable atrioventricular valve regurgitation in single ventricular circulation was associated with a moderately high risk of death, redo replacement and pacemaker implantation, whereas valve replacement at a later period and with a larger prosthetic valve size was associated with low mortality.

Conventional repair of total anomalous venous drainage without primary sutureless technique: surgical tips to prevent pulmonary vein obstruction.

OBJECTIVES: Although primary sutureless technique for total anomalous pulmonary venous drainage has been introduced to reduce postoperative pulmonary vein obstruction (PVO), controversy still exists about superiority of the procedure between the conventional repair and primary sutureless technique at the initial repair. In our unit, the conventional repair has been consistently used based on four important surgical policies: (1) mark incision lines between 2 chambers to gain anatomically natural alignment, (2) place precise stitches by "intima-to-intima" using monofilament suture, (3) adequate orifice size should be guaranteed in greater than expected mitral valve size, (4) do not hesitate to undertake a redo additional anastomosis by a different approach when an echocardiography shows the velocity more than 1.5 m/s. This study aims to evaluate mid-term outcome of the conventional repair for total anomalous pulmonary venous drainage. METHODS: Between 2004 and 2016, consecutive 15 patients who underwent the conventional repair without the primary sutureless technique were included in this study. Survival, Freedom from reoperation, and PVO were retrospectively reviewed. RESULTS: Mean follow-up period was 4.6 ± 3.7 years. Except for one patient who died of uncontrollable pleural effusion, all other patients survived with 5-year survival rate of 93.3%. For the 14 survivors, there was no PVO, nor reoperation. CONCLUSIONS: Following these policies, the mid-term outcome of the conventional total anomalous pulmonary venous drainage repair was excellent without the primary sutureless technique showing no obstruction. The conventional repair can be safely applied at the initial operation when the morphological condition allows for it.

Partial anomalous pulmonary venous drainage repair concomitant with bilateral semilunar valve replacements and pulmonary artery reconstruction for an adult female 20 years after initial truncus arteriosus repair.

We describe the case of a 21-year-old patient who underwent repairs for multiple lesions including aortic and pulmonary valve replacements, right ventricular outflow tract reconstruction, revision of the right pulmonary artery route, and a repair of partial anomalous pulmonary venous drainage, which was diagnosed during this fourth sternotomy. For these patients with adult CHD, it is most important to address all underlying factors as much as possible at the redo surgery.

Surgical strategy for aortic arch reconstruction after the Norwood procedure based on numerical flow analysis.

OBJECTIVES: Inefficient aortic flow after the Norwood procedure is known to lead to the deterioration of ventricular function due to an increased cardiac workload. To prevent the progression of aortic arch obstruction, arch reconstruction concomitant with second-stage surgery is recommended. The aim of this study was to determine the indications for reconstruction based on numerical simulation and to reveal the morphology that affects the haemodynamic parameters. METHODS: Fifteen patients who underwent the Norwood procedure or arch repair and Damus-Kaye-Stansel anastomosis were enrolled. The pressure gradient in aortic arch was 1.6 ± 3.9 mmHg (ranged from 0 to 12 mmHg) on catheter examination. Six patients who had prominent turbulent flow accompanied with a large flow energy loss index greater than 40 mW/m2 and high wall shear stress greater than 100 Pa underwent arch reconstruction. RESULTS: After arch reconstruction, the energy loss index significantly decreased from 88.5 ± 50.0 mW/m2 to 23.1 ± 10.4 mW/m2 (P = 0.026) and wall shear stress significantly decreased from 194.5 ± 87.4 Pa to 60.3 ± 40.5 Pa (P = 0.0062). There were 3 late deaths due to heart failure caused by progressive atrioventricular valve regurgitation during the follow-up period (60 months). The systemic ventricular function was preserved in the remaining patients without any pressure gradients in the arch. CONCLUSIONS: Determining the surgical strategy for arch reconstruction based on numerical flow analysis may effectively reduce the ventricular load even if no stenosis or pressure gradients are observed on catheter examination or echocardiography.

Mid-term outcomes of congenital mitral valve surgery: Shone's syndrome is a risk factor for death and reintervention.

OBJECTIVES: To study the recent trends and outcomes of congenital mitral valve surgery in children. METHODS: From 2008 to 2014, 84 procedures in 66 consecutive patients (41 procedures in 31 patients with mitral stenosis and 43 procedures in 35 patients with mitral regurgitation) were retrospectively evaluated. The mean age at surgery was 4.3 ± 5.4 years, and 27 patients (41%) were neonates or infants. RESULTS: Seven (11%) patients died during the follow-up period of 3.2 ± 2.3 years and 5 (71%) were <1 year. Ten mitral valve replacements were performed in 8 patients, including 1 pulmonary valve homograft, 3 Contegra conduits of 12 mm thickness in the intra-annular position and 6 mechanical valves. Shone's syndrome, dysplastic valve, a need for valve replacement and age <1 year were the risk factors for death or reoperation in a univariable analysis, while in a multivariable analysis of all patients, valve replacement and age <1 year remained as risk factors. In a multivariable analysis of 27 patients aged <1 year, mitral valve dysplasia was a significant risk factor for reoperation or death. The 5-year rate of freedom from death or reoperation in neonates or infants was 55% and that in patients aged >1 year was 88% (P = 0.003). CONCLUSIONS: An age of <1 year, mitral valve dysplasia and a need for mitral valve replacement were associated with a higher incidence of death or reoperation. Primary mitral valve replacement or univentricular strategy may have to be considered for symptomatic neonates with Shone's syndrome.

[Perioperative Management of Fontan Operation].

Surgical results of Fontan operation has been improved over the decades due to the introduction of the staged operations and some modifications of Fontan route from the classical atrio-pulmonary connection to total cavo-pulmonary connection. However, issues remain because of the single ventricular physiology of Fontan circulation. This article explains about the preoperative checklist for Fontan operation and tips on the postoperative management including an early extubation, use of inhaled nitric oxide after extubation, anticoagulation therapy, and efficacy of angiotensin converting enzyme inhibitor, supported by the cutting-edge evidence. Some patients who underwent Fontan operation, however, suffer from protein-losing enteropathy, heart failure, and thus are classified as failing Fontan. Treatment for these patients with failed Fontan is an unsolved problem in the state where heart transplantation is inadequately available in Japan.

Daily transient discontinuation of extracorporeal LVAD to prevent thromboembolism of mechanical aortic valve prosthesis.

Patients with mechanical aortic valves are generally contraindicated for left ventricular assist device (LVAD) insertion because the prosthetic valve often becomes fixed in closed position. A 41-year-old woman with mechanical aortic valve prosthesis experienced sudden chest pain and developed cardiogenic shock. A paracorporeal pulsatile LVAD and a monopivot centrifugal pump as a right VAD (RVAD) were implanted. The mechanical aortic valve was intentionally left in place. Soon after the operation, LVAD support was discontinued daily for few seconds to allow the mechanical aortic valve to open and to avoid thrombus formation. The patient was successfully weaned off RVAD and received anticoagulation therapy with warfarin. On postoperative day 141, she was transferred to a university hospital where a HeartMate II LVAD was implanted, and the aortic valve was successfully replaced with a bioprosthetic valve. The patient is currently awaiting heart transplantation.

Transareolar Video-Assisted Approach to the Atrial Septal Defect and Tricuspid Valve.

Minimally invasive atrial septal defect closure and tricuspid annuloplasty in female patients are normally performed through a right submammary anterior minithoracotomy approach. However, when the aortic root is located higher, the direction of aortic cannulation becomes not ideal through the submammary incision. In such cases, transareolar approach is useful. Through this approach, aortic cannulation and tricuspid operation can be performed with endoscopic assistance, and ASD closure can be performed under direct vision.

Impact of the location of the fenestration on Fontan circulation haemodynamics: a three-dimensional, computational model study.

OBJECTIVES: There is no consensus or theoretical explanation regarding the optimal location for the fenestration during the Fontan operation. We investigated the impact of the location of the fenestration on Fontan haemodynamics using a three-dimensional Fontan model in various physiological conditions. METHODS: A three-dimensional Fontan model was constructed on the basis of CT images, and a 4-mm-diameter fenestration was located between the extracardiac Fontan conduit and the right atrium at three positions: superior, middle, and inferior part of the conduit. Haemodynamics in the Fontan route were analysed using a three-dimensional computational fluid dynamic model in realistic physiological conditions, which were predicted using a lumped parameter model of the cardiovascular system. The respiratory effect of the caval flow was taken into account. The flow rate through the fenestration, the effect of lowering the central venous pressure, and wall shear stress in the Fontan circuit were evaluated under central venous pressures of 10, 15, and 20 mmHg. The pulse power index and pulsatile energy loss index were calculated as energy loss indices. RESULTS: Under all central venous pressures, the middle-part fenestration demonstrated the most significant effect on enhancing the flow rate through the fenestration while lowering the central venous pressure. The middle-part fenestration produced the highest time-averaged wall shear stress, pressure pulse index, and pulsatile energy loss index. CONCLUSIONS: Despite slightly elevated energy loss, the middle-part fenestration most significantly increased cardiac output and lowered central venous pressure under respiration in the Fontan circulation.

Outcomes of patients who declined surgery for acute Stanford type A aortic dissection with patent false lumen of the ascending aorta.

OBJECTIVES: This study aimed to evaluate the outcomes of patients who did not undergo initial aortic surgery for acute Stanford type A aortic dissection with a patent false lumen of the ascending aorta. METHODS: Inpatient and outpatient records were retrospectively reviewed. RESULTS: We identified 195 patients with acute type A aortic dissection with a patent ascending false lumen between January 1998 and March 2016. Of these, 137 underwent aortic surgery, 16 died before surgery and 42 declined aortic surgery. The ages of the patients who underwent and those who declined aortic surgery were 60.0 ± 10.6 years and 72.3 ± 12.4 years, respectively. The mortality rate of those who underwent and those who declined aortic surgery was 15 and 62% at 30 days and 19% and 67 at 90 days, respectively ( P < 0.0001). In the 58 patients who did not undergo initial aortic surgery, the maximum aortic diameter was correlated with survival ( P = 0.0037). At follow-up (3.7 ± 4.5 years; range 0-16.4 years), survival at 1, 5 and 10 years in those who underwent and those who declined initial aortic surgery was 78, 68 and 49%, and 29, 24 and 12%, respectively ( P < 0.0001). CONCLUSIONS: In this study of patients with acute Stanford type A aortic dissection with a patent false lumen of the ascending aorta, the mortality of those who declined initial aortic surgery was 62% at 30 days and 67% at 90 days, respectively, and a smaller aortic diameter was significantly associated with better survival.

Epicardial Implantable Cardioverter-Defibrillator in a 2-Month-Old Infant.

We describe the implantation of an implantable cardioverter defibrillator (ICD) in a 2-month-old infant with frequent sustained ventricular tachycardia (VT) refractory to antiarrhythmic agents. An epicardial ICD shock coil lead and pacing leads were placed, as was a cumbersome device console that was stored in a pocket between the left external and internal oblique muscles. These methods were safe and feasible even for such a small infant, and possible adverse events were avoided.

Factors associated with a low initial cerebral oxygen saturation value in patients undergoing cardiac surgery.

Regional cerebral oximetry using near-infrared spectroscopy device, an INVOS 5100 C (Medtronic, Minneapolis, MN, USA), during cardiac surgery aims to avoid perioperative neurological impairment, especially during cardiopulmonary bypass. However, it is not uncommon to encounter critically low initial cerebral regional oxygen saturation or a low value unresponsive to intervention. Therefore, it is important to identify factors associated with low saturation value other than true cerebral hypoxia. We investigated the relationship between preoperative regional cerebral oxygen saturation and clinical variables during cardiac surgery. From January 2013 to May 2016, 462 patients underwent elective cardiac surgery. Patient's ≤12 years of age, with acute cerebral infarction, with previous intracranial hemorrhage or neurosurgery, with concomitant aortic surgery, and having off-pump coronary artery bypass surgery were excluded. The remaining 223 patients were monitored by intraoperative regional cerebral oximetry. Univariate analysis found that scalp-cortex distance, cerebrospinal fluid thickness, left ventricular ejection fraction, hemoglobin concentration, estimated glomerular filtration rate, and hemodialysis were significantly correlated with the initial regional oxygen saturation value. Multiple regression analysis revealed that scalp-cortex distance, left ventricular ejection fraction, hemoglobin, and hemodialysis remained as significant variables. A receiver operating characteristic analysis found that for a low initial regional oxygen saturation value of 40%, the thresholds of scalp-cortex distance, left ventricular ejection fraction, and hemoglobin concentration were 17.6 mm, 45.2%, and 7.5 g/dl, respectively. In conclusion, brain atrophy, poor left ventricular function, anemia, and hemodialysis were associated with low initial cerebral regional oxygen saturation values in adult cardiac surgery patients.