J Clin Monit Comput [JOURNAL]
Sun
200 papers
Quantitative electroencephalogram and machine learning to predict expired sevoflurane concentration in infants.
Kumar R, Skowno J, von Ungern-Sternberg BS … +21 more , Davidson A, Xu T, Zhang J, Song X, Zhang M, Zhao P, Liu H, Jiang Y, Zuo Y, de Graaff JC, Vutskits L, Olbrecht VA, Szmuk P, Simpao AF, Tsui FR, Pratap JN, Padiyath A, Nelson O, Kurth CD, Yuan I, BRAIN Collaborative Investigators
J Clin Monit Comput · 2025 Oct · PMID 40381151 · Full text
Processed electroencephalography (EEG) indices used to guide anesthetic dosing in adults are not validated in young infants. Raw EEG can be processed mathematically, yielding quantitative EEG parameters (qEEG). We hypoth...
Processed electroencephalography (EEG) indices used to guide anesthetic dosing in adults are not validated in young infants. Raw EEG can be processed mathematically, yielding quantitative EEG parameters (qEEG). We hypothesized that machine learning combined with qEEG can accurately classify expired sevoflurane concentrations in young infants. Knowledge from this may contribute to development of future infant-specific EEG algorithms. Frontal EEG collected from infants ≤ 3 months were time-matched as one-minute epochs to expired sevoflurane (eSevo). Fifteen qEEG parameters were extracted from each epoch and eight machine learning models combined the qEEG to classify each epoch into one of four eSevo levels (%): 0.1-1.0, 1.0-2.1, 2.1-2.9, and > 2.9. 64 epochs formed the post hoc SHAP dataset to determine the qEEG that contributed most to the model. The remaining epochs were randomly split 50 times into 80/20 training/testing sets. Accuracy and F1-score determined model performance. 42 infants provided 4574 epochs. The top classifiers K-nearest neighbors, default multi-layer perceptron, and support vector machine achieved 67.5-68.7% accuracy. Burst suppression ratio and entropy β were the top contributors to the models. Post hoc analysis performed without burst suppression ratio yielded similar prediction performance. In young infants, machine learning applied to qEEG predicted eSevo levels with moderate success. Burst suppression ratio, the most important contributor, represented an efficient EEG feature that encapsulated underlying EEG changes seen on other qEEG features. These results provided insight into EEG parameter selection and optimal machine learning models used for future development of infant-specific EEG algorithms.
Editor letter for Sanna Holmskär "Is quantitative pupillometry affected by ambient light? A prospective crossover study. J Clin Monit Comput, 2025".
Couret D, Doucet W, Asmolov R … +2 more , Simeone P, Velly L
J Clin Monit Comput · 2026 Apr · PMID 40381150 · Full text
Abstract loading — click title to view on PubMed.
Improving urinary oxygen monitoring with a transit time algorithm: enhancing AKI detection in cardiac surgery.
Ramezani A, Silverton N, Kuck K
J Clin Monit Comput · 2025 Dec · PMID 40323566 · Publisher ↗
Acute kidney injury (AKI) affects 40-50% of cardiac surgery patients and is closely linked to renal medullary hypoxia. Although urinary oxygen partial pressure (PuO) offers real-time insight into renal oxygenation, varia...
Acute kidney injury (AKI) affects 40-50% of cardiac surgery patients and is closely linked to renal medullary hypoxia. Although urinary oxygen partial pressure (PuO) offers real-time insight into renal oxygenation, variable urine transit times through the urinary catheter can impair measurement accuracy. This study aimed to develop an algorithm that calculates transit time by modeling urine flow as discrete particles and to assess whether it improves PuO estimation. The proposed algorithm models urine flow as discrete particles, tracking transit time through the urinary catheter. The transit time allows correcting oxygen measurements at the catheter exit, mitigating distortions from variable flow rates. Validation used a bench-top system with a flow sensor, a 30-cm glass tube simulating a catheter, and optode-based oxygen sensors positioned inside a flask and at the catheter entry and exit. Flow rates spanned 20-450 mL/h, and flask oxygen 15-120 mmHg, with exit compared to entrance values. Without adjustment, the root mean square error (RMSE) between entrance and exit oxygen measurements was 15.71 mmHg. Incorporating the transit time correction reduced the RMSE to 5.82 mmHg. This marked improvement indicates that the corrected measurements more accurately reflect the true oxygen levels entering the catheter across various flow conditions. By accounting for dynamic urine transit times, the proposed algorithm substantially enhances the accuracy of urinary oxygen monitoring. This improvement in estimating renal oxygenation may facilitate noninvasive detection of renal hypoxia and allow for timely interventions to reduce the incidence and severity of AKI in cardiac surgery patients.
Developing a machine learning-based prediction model for postinduction hypotension.
Katsin M, Glebov M, Berkenstadt H … +5 more , Orkin D, Portnoy Y, Shuchami A, Yaniv-Rosenfeld A, Lazebnik T
J Clin Monit Comput · 2025 Oct · PMID 40323565 · Full text
Arterial hypotension is a common and often unintended event during surgery under general anesthesia, associated with increased postoperative complications, such as kidney injury, myocardial injury, and stroke. Postinduct...
Arterial hypotension is a common and often unintended event during surgery under general anesthesia, associated with increased postoperative complications, such as kidney injury, myocardial injury, and stroke. Postinduction hypotension (PIH) is influenced by patient-specific factors, chronic medication use, and anesthetic induction regimens. Traditional predictive models struggle with this complexity, making machine learning (ML) a promising alternative due to its ability to handle complex datasets and identify hidden patterns. This study aimed to develop and validate an ML-based model for predicting PIH and identifying key clinical predictors. A retrospective cohort study of 20,309 adult patients undergoing non-obstetric surgery under general anesthesia with intravenous induction was conducted. The primary outcome was the occurrence of PIH, defined as mean arterial pressure (MAP) < 55 mmHg within 10 min post-induction. Data were split into training and validation sets using k-fold cross-validation. The model's predictive performance was evaluated using the area under the receiver operating characteristic curve (AUC), and feature importance was assessed using SHapley Additive exPlanations (SHAP) values. PIH occurred in 4,948 patients (24.4%). Key predictors included preinduction systolic and mean arterial pressures, propofol dose, and beta-blocker use. The ML model achieved an AUC of 0.732 in predicting PIH. The ML-based model demonstrated significant predictive capability for PIH, identifying key clinical predictors. This model holds the potential for improving preoperative planning and patient risk stratification. However, further validation through prospective studies is necessary to confirm these findings.
The haemodynamic effects of pneumoperitoneum on pulse pressure variation - a prospective, observational study.
Hovgaard HL, Vistisen ST, Enevoldsen J … +2 more , de Paoli FV, Juhl-Olsen P
J Clin Monit Comput · 2025 Oct · PMID 40323564 · Full text
The effects of pneumoperitoneum on dynamic predictors of fluid responsiveness such as pulse pressure variation (PPV) remain uncertain. This uncertainty arises from potentially opposing physiological mechanisms that affec...
The effects of pneumoperitoneum on dynamic predictors of fluid responsiveness such as pulse pressure variation (PPV) remain uncertain. This uncertainty arises from potentially opposing physiological mechanisms that affect cardiovascular dynamics during conditions with increased intra-abdominal pressure (IAP). Deriving PPV with high precision during induction of pneumoperitoneum may provide new insights into the complex relationship between intra-abdominal pressure changes and PPV. The hypothesis was that PPV derived from a generalised additive model (PPV) would increase with the induction of pneumoperitoneum and the associacted increase in IAP. This was a prospective, observational study in patients undergoing oesophagectomy. Before and after induction of pneumoperitoneum, haemodynamic variables including PPV and stroke volume variation (SVV) were recorded with the Hemosphere monitor. PPV was estimated offline from the arterial blood pressure curve. A total of 34 patients were included in the final analysis. PPV increased by a factor of 1.49 (95% CI: 1.25-1.77) as intra-abdominal pressure increased from baseline to 12 mmHg. SVV and PPV from the HemoSphere monitor increased with a factor of 1.25 (95% CI: 1.13-1.39, p < 0.001) and 1.14 (95% CI: 1.00-1.29, p = 0.048), respectively. PPV derived from a generalised additive model increased approximately 50% from the induction of pneumoperitoneum to an IAP of 12 mmHg. PPV and SVV derived from the Hemosphere monitor also increased signicantly.
Variations of SPI during outpatient laparoscopic cholecystectomy without muscle relaxants using ANI-guided remifentanil.
Boselli E, Allaouchiche B
J Clin Monit Comput · 2025 Aug · PMID 40323563 · Publisher ↗
This study compared ANI and SPI during outpatient laparoscopic cholecystectomy without muscle relaxants. Forty adult patients were included in this prospective observational study. Induction was performed using propofol,...
This study compared ANI and SPI during outpatient laparoscopic cholecystectomy without muscle relaxants. Forty adult patients were included in this prospective observational study. Induction was performed using propofol, ketamine and remifentanil. All patients received bilateral TAP block. Maintenance of anesthesia was performed using remifentanil targeted to ANI 50-80 and desflurane targeted to MAC 0.8-1.2 without muscle relaxants. The ANI and SPI values were collected at different time-points and analyzed using repeated-measures ANOVA. The relationship between ANI and SPI were analyzed by linear regression. All procedures were performed without muscle relaxants. The mean ± SD ANI values significantly decreased from 70 ± 12 at induction to 57 ± 15 at intubation and 56 ± 17 at extubation and were maintained in the 50-80 target range throughout incision to exsufflation. The mean ± SD SPI values significantly decreased from 60 ± 15 at induction to 38 ± 16 at intubation, increased at 73 ± 14 at extubation and were in the 20-50 target range throughout incision to exsufflation. There was a poor but significant negative linear relationship (r = 0.053, p < 0.001) between SPI and ANI values. During laparoscopic cholecystectomy without muscle relaxants, remifentanil titrated to achieve a target ANI range of 50-80 provides SPI values with poor correlation ranging from 20 to 50, corresponding to adequate nociception-antinociception balance. Other studies comparing ANI and SPI guided remifentanil administration are required to determine the effect of each strategy on patient outcomes during laparoscopic cholecystectomy or other types of surgery.
EEG features associated with Alzheimer's disease and Frontotemporal dementia are not reflected by processed indices used in anesthesia monitoring.
Schwerin S, Dragovic SZ, Ostertag J … +3 more , Nguyen DM, Schneider G, Kreuzer M
J Clin Monit Comput · 2025 Aug · PMID 40259140 · Full text
Patients with dementia face increased risks after general anesthesia. Improved perioperative electroencephalogram (EEG) monitoring techniques could aid in identifying vulnerable patients. However, current technology reli...
Patients with dementia face increased risks after general anesthesia. Improved perioperative electroencephalogram (EEG) monitoring techniques could aid in identifying vulnerable patients. However, current technology relies on processed indices to measure "depth-of-anesthesia". Analyzing OpenNeuro Dataset ds004504, we compared resting-state, eyes-closed EEG recordings of healthy controls (n = 27) with patients diagnosed with Alzheimer's disease (AD, n = 35) and Frontotemporal dementia (FTD, n = 23). We focused on prefrontal recordings. Analysis included spectral analysis, the "fitting-oscillations&-one-over-f"-algorithm for aperiodic and periodic signal features, as well as calculations of openibis, permutation entropy (PeEn), spectral entropy (SpEn), and spectral edge frequency (SEF). Spectral differences were pronounced, including a higher alpha/theta-ratio of controls (2.62 [95%CI: 1.54-3.62]) compared to both AD (0.55 [95%CI: 0.26-1.92], P < 0.001, AUC: 0.765 [0.642-0.888]) and FTD (0.83 [95%CI: 0.33-1.65], P = 0.007, AUC: 0.779 [0.652-0.907]). Oscillatory peak detection within the alpha frequency band was more robust in control (versus AD: P = 0.003, Cramér's V = 0.374; versus FTD: P = 0.003, Cramér's V = 0.414). Processed index parameters did not show a clear trend. FTD was associated with a higher prefrontal openibis (95.53 [95%CI: 93.43-97.39]) than control (91.98 [95%CI: 89.46-96.27], P = 0.033, AUC: 0.717 [0.572-0.862]) and an elevated SEF (23.68 [95%CI: 14.10-25.57] Hz) compared to AD (16.60 [95%CI: 14.22-22.22] Hz, P = 0.041, AUC: 0.676 [0.532-0.821]). AD and FTD are associated with EEG baseline abnormalities, and a standard prefrontal montage, as used intraoperatively, could present a promising technical screening approach for cognitive vulnerability. However, these EEG features are obscured by processed index parameters currently used in neuroanesthesia monitoring. OpenNeuro Dataset ds004504 "A dataset of EEG recordings from: Alzheimer's disease, Frontotemporal dementia and Healthy subjects" (doi: https://doi.org/10.18112/openneuro.ds004504.v1.0.7 ).
Heart rate variability as a marker of multiple organ dysfunction syndromes: a systematic review.
Wojtanowski A, Hureau M, Jeanne M … +3 more , Bureau C, Recher M, De Jonckheere J
J Clin Monit Comput · 2025 Aug · PMID 40259139 · Full text
Multiple organ dysfunction syndrome (MODS) can be caused by many factors. Assessments of the severity of MODS are currently based on occasional measurements of several clinical variables (laboratory data, vital signs, et...
Multiple organ dysfunction syndrome (MODS) can be caused by many factors. Assessments of the severity of MODS are currently based on occasional measurements of several clinical variables (laboratory data, vital signs, etc.). The analysis of heart rate variability (HRV) as a guide to autonomic nervous system activity might be of value in the continuous assessment of the severity of MODS. We systematically reviewed publications on the value of HRV variables for the diagnosis of MODS in patients of any age admitted to the ICU. Two investigators independently searched the PubMed, Embase, Cochrane and Science Direct databases for articles in English or French published between 2004 and 2024. Ten studies were included and rated for endpoint bias (MODS or mortality), using the revised Quality Assessment of Diagnostic Accuracy Studies. Nine studies assessed MODS, and six assessed mortality. All the studies evidenced low HRV in patients with MODS and in non-survivors. The results of our review show that HRV indices are influenced by the severity of MODS and might serve as a tool for predicting mortality in patients with MODS. However, patient characteristics, and treatments and HRV processing methods must be taken into account when interpreting the results. In order to clarify the impact of MODS on HRV variables, methodologically rigorous studies are now needed.
Impact of cardiopulmonary bypass flow on the lower limit of cerebral autoregulation during cardiac surgery: a randomized cross-over pilot study.
Desebbe O, Berna A, Joosten A … +5 more , Raphael D, Malapert G, Rolo D, Taccone FS, Gergele L
J Clin Monit Comput · 2025 Jun · PMID 40220213 · Publisher ↗
Assessment of cerebral autoregulation is challenging under different hemodynamic conditions during cardiac surgery and must be rapidly calculated in order to optimize mean arterial pressure (MAP). Whether systemic flow d...
Assessment of cerebral autoregulation is challenging under different hemodynamic conditions during cardiac surgery and must be rapidly calculated in order to optimize mean arterial pressure (MAP). Whether systemic flow during cardiopulmonary bypass impacts the lower limit of cerebral autoregulation (LLA) remains unclear. Forty patients requiring cardiac surgery were included in this randomized crossover study. Patients assigned to the conventional/high blood flow arm received 20 min of conventional cardiopulmonary bypass (CPB) blood flow (2.2 L/min.m²) followed by 20 min of high blood flow (2.8 L/min.m²), both during aortic cross clamping. Patients assigned to the high/conventional arm received the same flows but in reverse order. During each 20-minute period, MAP was gradually increased from 40 to 90 mmHg, while PaCO, hematocrit, depth of anesthesia, central temperature and arterial oxygen tension were kept constant. Continuous cerebral blood flow velocities of the middle cerebral artery (Fv) were monitored using transcranial doppler. Cerebral autoregulation was calculated using a Pearson's correlation coefficient (Mean flow index, Mxa) between the MAP and Fv. Mxa values were then plotted across MAP ranges. The LLA was defined as the corresponding MAP value when Mxa initially decreased and crossed the threshold value of 0.4. A mixed model, including the LLA as the dependent variable, the CPB flow and period as fixed effects and patients as a random effect was used to compare conventional and high CPB flows. Thirty-seven patients were analyzed. The LLA mean difference between groups, adjusted on the period, was - 2.8 (SE 2.4) mmHg with 95% CI [-7.8, + 2.1 mmHg], p = 0.2538). 24% of patients presented an LLA < 65 mmHg during the conventional CPB flow phase versus 35% during the high CPB flow phase. Increasing the cardiopulmonary pump flow did not decrease the LLA during cardiac surgery.
A smartphone application for semi-automated QT interval analysis based on a snapshot of an electrocardiogram trace displayed on a patient monitor.
Beckmann D, Flick M, Kouz K … +1 more , Saugel B
J Clin Monit Comput · 2025 Aug · PMID 40208501 · Full text
We developed a smartphone application (SMART-QT application) that can semi-automatically measure QT and QTc intervals based on a snapshot of the electrocardiogram (ECG) trace and the heart rate displayed on a patient mon...
We developed a smartphone application (SMART-QT application) that can semi-automatically measure QT and QTc intervals based on a snapshot of the electrocardiogram (ECG) trace and the heart rate displayed on a patient monitor. In this study, we aimed to validate the SMART-QT application. In this prospective single-center method comparison study, we measured QT and QTc intervals with the SMART-QT application (QT and QTc; test method) and simultaneously manually measured QT and QTc intervals from a 12-lead ECG (QT and QTc; reference method) in 57 adult volunteers and patients who had sinus rhythm and no acute or chronic cardiac comorbidities. To investigate the agreement between QT and QT and between QTc and QTc, we performed Bland-Altman analyses and calculated the mean of the differences, the standard deviation, and the 95%-limits of agreement (95%-LOA). We defined clinically acceptable agreement as maximum mean of the differences ± standard deviation of 20 ± 20 ms. The mean of the differences between QT and QT was 14 ± 20 ms (95%-LOA -26 to 54 ms). The mean of the differences between QTc and QTc was 13 ± 15 ms (95%-LOA -16 to 42 ms). The agreement between QT and QT and between QTc and QTc was clinically acceptable in adult volunteers and patients without cardiac comorbidities.
Is quantitative pupillometry affected by ambient light? A prospective crossover study.
Holmskär S, Öhrn M, Furudahl M … +2 more , Kesti J, Pansell J
J Clin Monit Comput · 2025 Oct · PMID 40208500 · Full text
PURPOSE: Pupillary examination is a central part of the neurological assessment. While quantitative pupillometry (QP) improves reliability, the impact of ambient light, particularly on the Neurological Pupil index (NPi),...
PURPOSE: Pupillary examination is a central part of the neurological assessment. While quantitative pupillometry (QP) improves reliability, the impact of ambient light, particularly on the Neurological Pupil index (NPi), remains unclear. This study aimed to clarify the effects of ambient light on QP parameters in a critical care setting. METHODS: We performed a prospective crossover study, including 20 adult patients requiring invasive ventilation. Pupillometry was performed during bright condition (BC1), then dark condition (DC), then bright condition again (BC2). In our primary analysis we compared NPi values across conditions (DC1 vs. BC, BC vs. DC2, DC1 vs. DC2). In the secondary analysis, we compared all other QP parameters. RESULTS: All QP values except constriction velocity and dilation velocity were non-normal. The median NPi was significantly lower in BC compared to dark conditions DC1 in both eyes. In 25% of participants the NPi decreased by 0.6 or more. Conversely, a significant increase in median NPi of both eyes was observed when switching from bright conditions back to dark (BC vs. DC2). No significant difference was found between the two dark condition measurements (DC1 and DC2). The secondary analysis showed that the differences in NPi were driven by differences in most, but not all, QP parameters included in NPi. CONCLUSIONS: We corroborate previous findings that the level of ambient light affects QP parameters in critically ill patients. This needs to be considered for accurate interpretation of QP parameters. Future studies may explore potential automated light correction methods for wider clinical applicability.
Detection of fluid responsiveness by changes of perfusion index and pleth-variability index during passive leg raising in spontaneously breathing post-surgical patients: a prospective interventional study.
Rauch S, Seraglio PM, Dal Cappello T … +3 more , Roveri G, Falk M, Bock M
J Clin Monit Comput · 2025 Oct · PMID 40192908 · Full text
Predicting fluid responsiveness is crucial in treating circulatory failure, as only about half of patients benefit from volume expansion by increasing cardiac output (CO). Dynamic tests like passive leg raising (PLR) are...
Predicting fluid responsiveness is crucial in treating circulatory failure, as only about half of patients benefit from volume expansion by increasing cardiac output (CO). Dynamic tests like passive leg raising (PLR) are preferred over static parameters. While PLR reliably predicts fluid responsiveness, it typically requires invasive measurement of stroke volume (SV) or CO. The perfusion index (PI) and pleth variability index (PVI) are non-invasive metrics derived from oxygen saturation signals. PI has been shown to correlate with SV, and PVI has predicted fluid responsiveness in mechanically ventilated patients, but their role in spontaneously breathing patients remains unclear. This study aimed to assess whether PI and PVI could predict fluid responsiveness in post-surgical, spontaneously breathing patients during a PLR test. The hypothesis was that PI would increase and PVI would decrease in fluid responders. The prospective study included spontaneously breathing patients after major abdominal surgery in the ICU of Merano Hospital, Italy. SV and CO were measured using the FloTrac™ system, and PI and PVI were assessed using the Radical 7 monitor. Patients were considered responders if SV increased by ≥ 10% during the PLR test. Of the 47 patients enrolled, 23 (48.9%) were fluid responders. The percentage change in PI from baseline to 60 s into the PLR test was + 41.2% in responders and + 11.3% in non-responders. A PI increase of ≥ 23% predicted responders with 70% sensitivity and 75% specificity, with an area under the ROC curve of 0.74. Twenty-two patients (47%) were inside the grey zone. PVI did not differ significantly between groups. In conclusion, PI could be a helpful non-invasive tool for predicting fluid responsiveness during a PLR test in spontaneously breathing patients, though its diagnostic accuracy appears to be moderate.
Spontaneous recovery from rocuronium measured by mechanomyography during 100- or 200-Hz tetanic stimulations compared to normalized train-of-four with acceleromyography.
Dubois PE, Moreillon F, Bihin B … +5 more , De Dorlodot C, Meyer S, Maseri A, Passeraub PA, d'Hollander AA
J Clin Monit Comput · 2025 Oct · PMID 40167977 · Full text
Neuromuscular block recovery was evaluated using high-frequency tetanic ulnar nerve simulations compared to normalized train-of-four (NTOF) in anesthetized patients. Under intravenous general anesthesia, we compared rocu...
Neuromuscular block recovery was evaluated using high-frequency tetanic ulnar nerve simulations compared to normalized train-of-four (NTOF) in anesthetized patients. Under intravenous general anesthesia, we compared rocuronium-induced neuromuscular recovery using 5 s 100- and 200-Hz tetanic stimulations via isometric mechanomyography to acceleromyographic NTOF in 20 consenting patients. The primary outcome was the comparison by Student's t-tests of 100- and 200-Hz tetanic fade ratios (residual force at the end of the contraction / maximal force reached during the 5 s) before rocuronium administration and at different recovery levels. The secondary outcome was the quantification of any significant fade occurring with 100- and 200-Hz stimulations after reaching the acceleromyographic NTOF ratio of 0.9 during subsequent stages of spontaneous recovery until their fade ratios exceeded 0.9. During early (TOF count ≥ 1) and intermediate (NTOF ratio ≥ 0.5) stages of recovery, both 100- and 200-Hz tetanic fade ratios were similarly low. However, during late recovery when NTOF ratio ≥ 0.9, 200-Hz stimulation induced a significantly deeper muscular fade than 100-Hz (tetanic fade ratio 0.20 ± 0.23 vs. 0.64 ± 0.29, P < 0.001). The delays between the recovery of NTOF ratio 0.9 and 100- or 200-Hz tetanic fade ratio 0.9 were 7.7 ± 7.1 and 43.6 ± 14.6 min, respectively. In anesthetized humans, mechanomyographic 200-Hz tetanic stimulation detects lighter levels of residual paralysis than NTOF and 100-Hz tetanic stimulation during a valuable additional period. Registered in the ClinicalTrials.gov Registry NCT05474638 on July 15th 2022.
Automated mitochondrial oxygen consumption (mitoVO) analysis via a bi-directional long short-term memory neural network.
de Wijs CJ, Behr JR, Streng LWJM … +3 more , van der Graaf ME, Harms FA, Mik EG
J Clin Monit Comput · 2025 Oct · PMID 40159587 · Full text
Monitoring in vivo mitochondrial oxygen tension (mitoPO) enables the measurement of mitochondrial oxygen consumption (mitoVO), providing deeper insights into the skin's mitochondrial environment. However, current mitoVO...
Monitoring in vivo mitochondrial oxygen tension (mitoPO) enables the measurement of mitochondrial oxygen consumption (mitoVO), providing deeper insights into the skin's mitochondrial environment. However, current mitoVO analysis often relies on manual identification of start and end points, which introduces substantial inter-user variability. Addressing this limitation is crucial for broader adoption, comparability, and reproducibility across research groups. Therefore, the aim of this study was to develop a neural network-based software that automatically analyzes mitoVO. A Bi-directional Long Short-Term Memory neural network was trained on 125 mitoPO measurement sequences and optimized through Bayesian optimization. It identifies start points and measurement periods, then applies a modified Michaelis-Menten fit to calculate mitoVO. This framework, embedded in automated software, was validated against the consensus of 3 raters. Bayesian optimization yielded an overall network performance of 94.2% on the test set. The neural network identified 91% of mitoVO start points within a ± 5-sample range of the manual consensus. Mean mitoVO values for the consensus and software were 6.56 and 6.63 mmHg s, respectively, corresponding to a bias of -0.057 mmHg s. Multiple runs of the network on the same dataset produced identical results, confirming consistency and eliminating inter-user variability. The developed neural network-based software automatically and consistently analyzes mitoVO measurements, substantially reducing reliance on subjective judgments. By enabling a standardized approach to mitoVO analysis, this tool improves data comparability and reproducibility across research settings. Future work will focus on further refining precision and extending functionality through multi-center collaborations.
Pulse rate variability as a predictor for length of stay for patients with bronchiolitis in the pediatric intensive care unit.
Kwon SB, Weinerman B, Nametz D … +8 more , Alalqum T, Lee IS, Megjhani M, McLaren SH, Ranard B, Ku Y, Geneslaw A, Park S
J Clin Monit Comput · 2025 Jun · PMID 40131664 · Full text
Patients admitted to pediatric Intensive Care Unit (PICU) due to bronchiolitis have unpredictable length of stay (LOS). The aim of this study is to observe the difference in the relationship between pulse rate variabilit...
Patients admitted to pediatric Intensive Care Unit (PICU) due to bronchiolitis have unpredictable length of stay (LOS). The aim of this study is to observe the difference in the relationship between pulse rate variability (PRV) and heart rate variability (HRV) for patients with bronchiolitis admitted to the PICU and its association with LOS. The first 12 h of physiologic data after PICU admission were used for analysis. Electrocardiogram (ECG) and photoplethysmogram (PPG) were divided into non-overlapping 5-minute segments, and R-peak and PPG-peak were obtained to calculate PRV and HRV. Correlation was calculated between HRV and PRV for both PICU short-stay and long-stay groups. A total of 119 patients were included in this study, where 66 are short-stay and 53 are long-stay group. For both LOS groups, PRV and HRV parameters were significantly higher HRV parameters compared to PRV. SDSD, SDNN, RMSSD, pNN50, SD1, and SD2 were 13.72, 10.24, 13.72, 0.77, 9.7, 10.6, and 0.85 for HRV. For PRV it was 5.88, 4.83, 5.88, 0.75, 4.16, 5.28, and 0.85. However, in the comparison of the correlations between PRV and HRV parameters, the short-stay group had significantly higher correlation compared to the long-stay group. The correlations in the short-stay group are above 0.72-0.82, whereas for the long-stay group the correlation ranged from 0.29 to 0.67. This study demonstrates that the correlation between the PRV and HRV is lower in patients with longer length of stay, suggesting this can be a potential metric for LOS in PICU.
Non-invasive estimation of beat-by-beat aortic blood pressures from electrical impedance tomography data processed by machine learning.
Müller-Graf F, Thönes JP, Krukewitt L … +7 more , Frenkel P, Richter H, Spors S, Kühn V, Zitzmann AR, Boehm SH, Reuter DA
J Clin Monit Comput · 2025 Oct · PMID 40131663 · Full text
Hypotension in perioperative and intensive care settings is a significant risk factor associated with complications such as myocardial infarction and kidney injury thereby increasing perioperative complications and morta...
Hypotension in perioperative and intensive care settings is a significant risk factor associated with complications such as myocardial infarction and kidney injury thereby increasing perioperative complications and mortality. Continuous blood pressure monitoring is essential, yet challenging due to the invasive nature of current methods. Non-invasive techniques like Electrical Impedance Tomography (EIT) have been explored but face challenges in accurate and consistent blood pressure estimation. A machine learning (ML) approach was used to predict aortic blood pressures from EIT voltage measurements in landrace pigs. A convolutional neural network (CNN) was trained on a dataset of 75 298 heartbeats, to predict systolic (SAP), mean (MAP), and diastolic arterial pressures (DAP) of individuals whose arterial pressures were unknown to the algorithm. The Intraclass Correlation Coefficient (3,1) with absolute agreement (ICC) was calculated and the concordance was estimated, comparing reference blood pressure measurements and ML-derived estimates. A risk classification was estimated for the calculated blood pressure as suggested by Saugel et al. 2018. The ML-model demonstrated moderate correlations with invasive blood pressure measurements (ICC for SAP of 0.530, for MAP of 0.563, and for DAP of 0.521.) with a low risk score for 75.8% of the SAP and 64.2% of MAP estimated blood pressures. ML-techniques using EIT-voltages showed promising preliminary results in non-invasive aortic blood pressure estimation. Despite limitations in the amount of available training data and the experimental setup, this study illustrates the potential of integrating ML in EIT signal processing for real-time, non-invasive blood pressure monitoring.
Empirical pharmacodynamic model of phenylephrine and intrathecal bupivacaine for mean arterial pressure prediction in obstetric patients presenting for elective cesarean delivery under spinal anesthesia.
Davoud SC, Ozaslan B, Aiello EM … +5 more , Kleinlein R, Eberhard B, Hassan H, Doyle FJ, Kovacheva VP
J Clin Monit Comput · 2025 Aug · PMID 40120014 · Publisher ↗
Cesarean delivery under spinal anesthesia may be complicated by hypotension in up to 80% of the patients. The response to standard-of-care prophylactic phenylephrine infusion varies, and there is little guidance on achie...
Cesarean delivery under spinal anesthesia may be complicated by hypotension in up to 80% of the patients. The response to standard-of-care prophylactic phenylephrine infusion varies, and there is little guidance on achieving optimal blood pressure control. In this work, we developed a data-driven pharmacodynamic relationship between intravenous phenylephrine, intrathecal bupivacaine, and maternal mean arterial pressure (MAP) in patients presenting for cesarean delivery. In this single-center cohort study, secondary use data were available for normotensive patients presenting for cesarean delivery. Intraoperative MAP, intrathecal bupivacaine, and intravenous phenylephrine doses were recorded prospectively every minute. The recorded data were used to identify and confirm a time series (Autoregressive with Exogenous Input (ARX)) model for predicting the MAP using MATLAB 2021a System Identification Toolbox and the Prediction Error Method. An independent model validation was conducted using a second dataset collected after the model fitting stage. Model identification was performed on 172 patients, using 70% for model fitting and 30% for testing. The final ARX model, which takes the past three data points to make predictions, performed 48.9% better than a mean constant model for one-minute ahead MAP predictions with a root mean square error (RMSE) of 3.6 ± 1.3 mmHg. Similar performance was observed on independent validation using a second dataset (N = 84), yielding an RMSE of 4.2 ± 1.6 mmHg for one-minute ahead MAP predictions. Our ARX model showed good performance at up to a three-minute prediction horizon and could be used for future decision support applications to guide phenylephrine dose titration.
Evaluation of non-invasive sensors for monitoring core temperature.
Thomas SS, Flickinger KL, Elmer J … +1 more , Callaway CW
J Clin Monit Comput · 2025 Dec · PMID 40120013 · Publisher ↗
We evaluated the accuracy and precision of zero-heat flux (ZHF) and dual sensor (DS) non-invasive temperature probes in intensive care unit (ICU) patients undergoing hypothermic temperature control, hypothesizing that bo...
We evaluated the accuracy and precision of zero-heat flux (ZHF) and dual sensor (DS) non-invasive temperature probes in intensive care unit (ICU) patients undergoing hypothermic temperature control, hypothesizing that both devices would accurately estimate core temperature. In a single-center prospective cohort study, we enrolled 35 ICU patients and applied continuous, non-invasive ZHF and/or DS probes to the lateral forehead or anterior chest to collect 358 observations. Conditions potentially influencing temperature estimation were recorded. Using Bland-Altman analysis with multiple paired observations per individual, we compared the bias between non-invasive probes and direct core temperature measurements. Lin's concordance coefficient (LCC) was computed to quantify precision. The mean bias between the ZHF probe and invasive temperature was + 0.98 °C; for the DS probe, it was - 2.19 °C. In hypothermic patients, the ZHF probe's accuracy improved (bias + 0.28 °C, LCC 0.86), while the DS probe remained inaccurate (bias - 2.52 °C, LCC 0.07). Clinical confounders like vasoactive agents or temperature control devices did not consistently affect bias, accuracy, or precision. Neither the ZHF nor DS non-invasive probes provided sufficient accuracy or precision to guide clinical decisions in the ICU. These results contrast with previous studies reporting biases within ± 0.5 °C. However, the ZHF probe showed promising limited deviation, especially in hypothermic patients.
Haemodynamic monitoring and management during non-cardiac surgery: a survey among German anaesthesiologists.
Vojnar B, Achenbach P, Flick M … +5 more , Reuter D, Sander M, Saugel B, Schubert AK, Gaik C
J Clin Monit Comput · 2025 Oct · PMID 40120012 · Full text
In 2023, the first German guideline on intraoperative haemodynamic monitoring and management for adults having non-cardiac surgery was published. The aim of this survey was to identify how anaesthetists in Germany manage...
In 2023, the first German guideline on intraoperative haemodynamic monitoring and management for adults having non-cardiac surgery was published. The aim of this survey was to identify how anaesthetists in Germany managed intraoperative haemodynamics and blood pressure before its publication. In September to October 2023, members of the German Society of Anaesthesiology and Intensive Care Medicine (DGAI) were invited via email to participate in this anonymous online survey. Thirty-one questions covered demographics, clinical experience, approaches to perioperative blood pressure measurement and common thresholds, as well as the use of advanced haemodynamic monitoring and its potential therapeutic implications. 1,079 fully completed questionnaires were included in the analysis. When intermittent oscillometry was used to measure blood pressure, a 3-minute interval was usually applied during induction of anaesthesia (42%; 451/1,079). For invasive blood pressure monitoring, more than half (53%; 574/1,079) inserted an arterial line after induction of anaesthesia. Nearly all (94%; 1,012/1,079) focused on the mean arterial pressure for blood pressure monitoring, with a large majority (77%; 779/1012) considering values below 60-65 mmHg to be critically low. Intraoperative hypotension was managed based on an internal protocol by only 21% (223/1,079). Regarding advanced haemodynamic monitoring, 43% (459/1,079) frequently used pulse contour analysis, while 67% (721/1,079) reported that monitors with finger-cuff technology were not available in their department. 47% (504/1,079) cited a lack of experience as one of the main reasons for the infrequent use of cardiac output monitoring. This survey among DGAI members provides important insights into current practices of haemodynamic monitoring and management prior to the publication of the recent German guideline on 'Intraoperative haemodynamic monitoring and management of adults having non-cardiac surgery'.
Continuous monitoring of intracranial pressure and end tidal carbon dioxide variations in traumatic brain injury: introducing the carbon dioxide reactivity index (CO2Rx).
Gritti P, Bonfanti M, Zangari R … +10 more , Bonanomi E, Di Matteo M, Corbella D, Farina A, Lecchi L, Togni T, Mandelli P, Lanterna LA, Biroli F, Lorini FL
J Clin Monit Comput · 2025 Jun · PMID 40120011 · Publisher ↗
PURPOSE: The continuous monitoring of cerebral metabolic autoregulation in patients with severe traumatic brain injury (TBI) is poorly documented in the literature and largely absent from clinical practice. This study ai...
PURPOSE: The continuous monitoring of cerebral metabolic autoregulation in patients with severe traumatic brain injury (TBI) is poorly documented in the literature and largely absent from clinical practice. This study aimed to assess whether variations in intracranial pressure (ICP) and end-tidal carbon dioxide (ETCO2) can form the basis of an index for cerebrovascular autoregulation reactivity, and whether this index can improve the prediction of clinical outcomes in both adult and pediatric TBI patients. METHODS: Data from adult and pediatric patients with severe TBI were retrospectively analyzed. The Carbon Dioxide Reactivity Index (CO2Rx) was introduced as a novel tool to assess cerebrovascular reactivity in response to variations in CO2 and ICP. CO2Rx was calculated by analyzing the relationship between ICP and ETCO2, sampled at approximately 5-minute intervals, using linear correlation within moving time windows ranging from 40 to 180 min in 10-minute increments. The discriminatory power of CO2Rx in predicting clinical outcomes was evaluated through Receiver Operating Characteristic (ROC) curve analysis. The primary outcome measures included in-hospital mortality and the 12-month Glasgow Outcome Scale-Extended (GOSE) score. RESULTS: The study included 218 TBI patients (40 pediatric and 178 adult). CO2Rx values showed a significant correlation with outcomes, with a CO2Rx threshold of 0.28 effectively distinguishing between favorable and unfavorable outcomes. For the fatal/non-fatal outcome, the CO2Rx crude model alone had an Area Under the Curve (AUC) of 0.737. When combined with other predictors (Impact Core + ICP + CO2Rx), this model achieved the highest AUC of 0.929. CONCLUSION: CO2Rx demonstrated significant predictive value for mortality and unfavorable outcomes in TBI patients, serving as a continuous index of cerebrovascular reactivity to CO2. It holds potential to improve severe TBI management by optimizing the interaction between ventilation and metabolic autoregulation. TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT NCT05043545.
About
- Frequency
- Sun
- Papers found
- 200
- RSS feed
- Subscribe