Marcus Victor
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Marcus Victor has a total of 68 co-authors. Here are the top co-authors by number of shared publications:
- Lorenzo Berra
Massachusetts General Hospital
6 shared publications - Maurizio Cereda
Department of Anesthesiology, Mass General Brigham, Boston, MA
6 shared publications - Timothy Glen Gaulton
Department of Anesthesiology, Mass General Brigham, Boston, MA
5 shared publications - Monica M. S. Matsumoto
Electronics Engineering Division, Aeronautics Institute of Technology, Sao Paulo, Brazil
4 shared publications Yi Xin
Mass General Brigham and Harvard Medical School
4 shared publications- Glasiele Cristina Alcala
Division of Respirology, Department of Medicine, University Health Network, Toronto, Canada
3 shared publications
Latest Publications
- Open Access
- 7 July 2026
- Published by American Physiological Society in Journal of Applied Physiology
- Vol. 141 (2) , 423-432
Impact of Baydur Ratio Correction on the Reliability of Esophageal Pressure Measurement
Tidal changes in esophageal pressure (ΔPes) are used as a surrogate for pleural pressure changes (ΔPpl) to assess chest wall mechanics. The Baydur ratio evaluates pressure transmission and guides catheter positioning. The effects of correcting ΔPes using the Baydur ratio on the accuracy of ΔPpl estimation and its dependence on ventilation mode, spontaneous inspiratory effort, and body position, remain unclear. In ten pigs, ΔPpl was measured using intrapleural balloon catheters, while Pes was recorded at three esophageal locations in supine and prone positions. Animals underwent controlled and assisted ventilation. Baydur ratios were derived during assisted ventilation and from external chest compressions during controlled ventilation. ΔPes before and after correction was compared with ΔPpl using mixed-effects models and Bland–Altman analysis. Additionally, we quantified the proportion of values with (ΔPes−ΔPpl) ≤ 1 cmH₂O. During assisted ventilation, ΔPes showed good agreement with posterior ΔPpl. Baydur correction improved scaling and association (slope 0.76 vs 0.50; r 0.81 vs 0.79, p<0.01) and increased values within ±1 cmH₂O from 29% to 45% when the Baydur ratio fell outside the accepted range. During controlled ventilation, ΔPes showed weaker association and wider limits of agreement, without improvement after correction. Similar patterns were observed in prone position. ΔPes more accurately reflected posterior ΔPpl during assisted ventilation. Baydur correction improves scaling and association, particularly when the Baydur ratio falls outside the accepted range, but does not improve and may worsen agreement during controlled ventilation.
- 7 July 2026
- Published by American Physiological Society in Journal of Applied Physiology
- Vol. 141 (2) , 423-432
Open AccessImpact of Baydur Ratio Correction on the Reliability of Esophageal Pressure Measurement
Tidal changes in esophageal pressure (ΔPes) are used as a surrogate for pleural pressure changes (ΔPpl) to assess chest wall mechanics. The Baydur ratio evaluates pressure transmission and guides catheter positioning. The effects of correcting ΔPes using the Baydur ratio on the accuracy of ΔPpl estimation and its dependence on ventilation mode, spontaneous inspiratory effort, and body position, remain unclear. In ten pigs, ΔPpl was measured using intrapleural balloon catheters, while Pes was recorded at three esophageal locations in supine and prone positions. Animals underwent controlled and assisted ventilation. Baydur ratios were derived during assisted ventilation and from external chest compressions during controlled ventilation. ΔPes before and after correction was compared with ΔPpl using mixed-effects models and Bland–Altman analysis. Additionally, we quantified the proportion of values with (ΔPes−ΔPpl) ≤ 1 cmH₂O. During assisted ventilation, ΔPes showed good agreement with posterior ΔPpl. Baydur correction improved scaling and association (slope 0.76 vs 0.50; r 0.81 vs 0.79, p<0.01) and increased values within ±1 cmH₂O from 29% to 45% when the Baydur ratio fell outside the accepted range. During controlled ventilation, ΔPes showed weaker association and wider limits of agreement, without improvement after correction. Similar patterns were observed in prone position. ΔPes more accurately reflected posterior ΔPpl during assisted ventilation. Baydur correction improves scaling and association, particularly when the Baydur ratio falls outside the accepted range, but does not improve and may worsen agreement during controlled ventilation.
- Open Access
- 5 June 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 87111-87123
AI-Based Design, Optimization, and Validation of a CMOS 65-nm Current Conveyor IC for Chest Electrical Impedance Tomography
Electrical Impedance Tomography (EIT) is a non-invasive medical imaging technique that generates detailed functional images based on electrical impedance measurements. Achieving optimal electrical performance in EIT systems requires designing a robust current source with minimal harmonic distortion and a high output impedance. Traditional manual design methodologies for analog integrated circuits are often time-consuming and labor-intensive. This paper presents an approach using heuristic algorithms, specifically the Genetic Algorithm, Competitive Imperialist Algorithm, and Simulated Annealing, to accelerate the design and optimization of a current source for EIT applications. The proposed current source was designed in approximately 10 hours on a personal computer and manufactured in 65 nm CMOS technology, achieving a harmonic distortion of less than 0.95%, an output impedance over 600 kΩ, and an output current amplitude of 1.2 mA, occupying an area of 0.0123mm2 thereby validating the effectiveness of the design and optimization process.
- 5 June 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 87111-87123
Open AccessAI-Based Design, Optimization, and Validation of a CMOS 65-nm Current Conveyor IC for Chest Electrical Impedance Tomography
Electrical Impedance Tomography (EIT) is a non-invasive medical imaging technique that generates detailed functional images based on electrical impedance measurements. Achieving optimal electrical performance in EIT systems requires designing a robust current source with minimal harmonic distortion and a high output impedance. Traditional manual design methodologies for analog integrated circuits are often time-consuming and labor-intensive. This paper presents an approach using heuristic algorithms, specifically the Genetic Algorithm, Competitive Imperialist Algorithm, and Simulated Annealing, to accelerate the design and optimization of a current source for EIT applications. The proposed current source was designed in approximately 10 hours on a personal computer and manufactured in 65 nm CMOS technology, achieving a harmonic distortion of less than 0.95%, an output impedance over 600 kΩ, and an output current amplitude of 1.2 mA, occupying an area of 0.0123mm2 thereby validating the effectiveness of the design and optimization process.
- Top Cited Papers
- 1 September 2025
- Published by American Physiological Society in Journal of Applied Physiology
- Vol. 139 (3) , 849-862
Biomechanical implications of mass loading in a swine model of acute hypoxemic respiratory failure
In a ventilated swine model, we isolated the mechanical effects of mass loading from those of lung injury. Despite optimization, positive end-expiratory pressure (PEEP) could not restore baseline pulmonary compliance or lung volumes in either healthy or injured lungs. Partitioned respiratory mechanics and imaging reveal that geometric constraints and load-induced airway closure, demonstrated here for the first time in injured swine, may limit the effectiveness of recruitment maneuvers and optimized PEEP in mass-loaded lungs.
- 1 September 2025
- Published by American Physiological Society in Journal of Applied Physiology
- Vol. 139 (3) , 849-862
Top Cited PapersBiomechanical implications of mass loading in a swine model of acute hypoxemic respiratory failure
In a ventilated swine model, we isolated the mechanical effects of mass loading from those of lung injury. Despite optimization, positive end-expiratory pressure (PEEP) could not restore baseline pulmonary compliance or lung volumes in either healthy or injured lungs. Partitioned respiratory mechanics and imaging reveal that geometric constraints and load-induced airway closure, demonstrated here for the first time in injured swine, may limit the effectiveness of recruitment maneuvers and optimized PEEP in mass-loaded lungs.
- 13 June 2025
- Published by IOP Publishing in Biomedical Physics & Engineering Express
- Vol. 11 (4) , 045013
Supervised and unsupervised learning for lung perfusion data segmentation in electrical impedance tomography
Objective: Effective lung gas exchange relies on the balance between alveolar ventilation and perfusion, which can be disrupted in mechanically ventilated patients. Lung perfusion assessment using electrical impedance tomography (EIT) typically involves a sudden injection of a hypertonic saline solution. The large field of view provided by EIT often results in ambivalent behavior of many voxel waveforms following an indicator injection, where some exhibit indicator kinetics solely through the lungs (pulmonary), while others show passage through both the heart and lungs (hybrid). Consequently, a segmentation algorithm is essential for accurate perfusion evaluation. Approach: Sixteen pigs (29–35 kg) were mechanically ventilated and received a 10 ml bolus of 7.5% NaCl solution to assess lung perfusion during a healthy stage and, later, in an injured stage after receiving 3.5 ml kg−1 of HCl to induce acute lung injury. Supervised (Bagged Trees, Neural Networks, and Support Vector Machine) and unsupervised (K-means, Hierarchical, and Principal Component Analysis) learning methods were employed using 115 saline injections comprising voxel waveforms to label voxels as either hybrid or pulmonary. All segmentation methods were compared to a ground-truth mask manually drawn. A training dataset (81 injections) was used to train and cross-validate (five-fold) the supervised methods using previously extracted features. The test dataset (34 injections) was used to test both supervised and unsupervised learning algorithms. Main Results: A Principal Component Analysis (unsupervised learning) method exhibited the best overall performance, achieving 83% sensitivity, 92% specificity, 89% accuracy, and 84% dice similarity coefficient. No significant difference in performance was observed between healthy and injured subsets. Unsupervised methods consistently yielded more physiologically plausible and less scattered regions of interest. Significance: Accurate voxel labeling is crucial for lung perfusion assessment, as it enables discrimination of the indicator passage through the heart and lungs, thereby improving the estimation of regional pulmonary blood flow.
- 13 June 2025
- Published by IOP Publishing in Biomedical Physics & Engineering Express
- Vol. 11 (4) , 045013
Supervised and unsupervised learning for lung perfusion data segmentation in electrical impedance tomography
Objective: Effective lung gas exchange relies on the balance between alveolar ventilation and perfusion, which can be disrupted in mechanically ventilated patients. Lung perfusion assessment using electrical impedance tomography (EIT) typically involves a sudden injection of a hypertonic saline solution. The large field of view provided by EIT often results in ambivalent behavior of many voxel waveforms following an indicator injection, where some exhibit indicator kinetics solely through the lungs (pulmonary), while others show passage through both the heart and lungs (hybrid). Consequently, a segmentation algorithm is essential for accurate perfusion evaluation. Approach: Sixteen pigs (29–35 kg) were mechanically ventilated and received a 10 ml bolus of 7.5% NaCl solution to assess lung perfusion during a healthy stage and, later, in an injured stage after receiving 3.5 ml kg−1 of HCl to induce acute lung injury. Supervised (Bagged Trees, Neural Networks, and Support Vector Machine) and unsupervised (K-means, Hierarchical, and Principal Component Analysis) learning methods were employed using 115 saline injections comprising voxel waveforms to label voxels as either hybrid or pulmonary. All segmentation methods were compared to a ground-truth mask manually drawn. A training dataset (81 injections) was used to train and cross-validate (five-fold) the supervised methods using previously extracted features. The test dataset (34 injections) was used to test both supervised and unsupervised learning algorithms. Main Results: A Principal Component Analysis (unsupervised learning) method exhibited the best overall performance, achieving 83% sensitivity, 92% specificity, 89% accuracy, and 84% dice similarity coefficient. No significant difference in performance was observed between healthy and injured subsets. Unsupervised methods consistently yielded more physiologically plausible and less scattered regions of interest. Significance: Accurate voxel labeling is crucial for lung perfusion assessment, as it enables discrimination of the indicator passage through the heart and lungs, thereby improving the estimation of regional pulmonary blood flow.
- 15 May 2024
- Published by Oxford University Press (OUP) in American Journal of Respiratory and Critical Care Medicine
- Vol. 209 (10) , 1263-1265
First-Pass Kinetics Model to Estimate Pulmonary Perfusion by Electrical Impedance Tomography During Uninterrupted Breathing
- 15 May 2024
- Published by Oxford University Press (OUP) in American Journal of Respiratory and Critical Care Medicine
- Vol. 209 (10) , 1263-1265
First-Pass Kinetics Model to Estimate Pulmonary Perfusion by Electrical Impedance Tomography During Uninterrupted Breathing
- 14 May 2024
- Published by Wolters Kluwer Health in Anesthesiology
- Vol. 141 (2) , 377-378
Electrical Impedance Tomography for Assessing the Impact of Inhaled Nitric Oxide on Pulmonary Artery Pressure
ANESTHESIOLOGY is the highest-impact, peer-reviewed medical journal that publishes trusted evidence that transforms the practice of perioperative,
- 14 May 2024
- Published by Wolters Kluwer Health in Anesthesiology
- Vol. 141 (2) , 377-378
Electrical Impedance Tomography for Assessing the Impact of Inhaled Nitric Oxide on Pulmonary Artery Pressure
ANESTHESIOLOGY is the highest-impact, peer-reviewed medical journal that publishes trusted evidence that transforms the practice of perioperative,
- 4 January 2024
- Published by Springer Nature
- p. 313-323
Evaluation of Explainable AI Methods in CNN Classifiers of COVID-19 CT Images
- 4 January 2024
- Published by Springer Nature
- p. 313-323
Evaluation of Explainable AI Methods in CNN Classifiers of COVID-19 CT Images
- 4 January 2024
- Published by Springer Nature
- p. 167-176
Unsupervised Clustering Methods for Lung Perfusion Data Segmentation in Electrical Impedance Tomography
- 4 January 2024
- Published by Springer Nature
- p. 167-176
Unsupervised Clustering Methods for Lung Perfusion Data Segmentation in Electrical Impedance Tomography
- 1 January 2024
- Published by IOP Publishing in Physiological Measurement
- Vol. 45 (1) , 015002
Esophageal balloon catheter system identification to improve respiratory effort time features and amplitude determination
Objective . Understanding a patient’s respiratory effort and mechanics is essential for the provision of individualized care during mechanical ventilation. However, measurement of transpulmonary pressure (the difference between airway and pleural pressures) is not easily performed in practice. While airway pressures are available on most mechanical ventilators, pleural pressures are measured indirectly by an esophageal balloon catheter. In many cases, esophageal pressure readings take other phenomena into account and are not a reliable measure of pleural pressure. Approach. A system identification approach was applied to provide accurate pleural measures from esophageal pressure readings. First, we used a closed pressurized chamber to stimulate an esophageal balloon and model its dynamics. Second, we created a simplified version of an artificial lung and tried the model with different ventilation configurations. For validation, data from 11 patients (five male and six female) were used to estimate respiratory effort profile and patient mechanics. Main results. After correcting the dynamic response of the balloon catheter, the estimates of resistance and compliance and the corresponding respiratory effort waveform were improved when compared with the adjusted quantities in the test bench. The performance of the estimated model was evaluated using the respiratory pause/occlusion maneuver, demonstrating improved agreement between the airway and esophageal pressure waveforms when using the normalized mean squared error metric. Using the corrected muscle pressure waveform, we detected start and peak times 130 ± 50 ms earlier and a peak amplitude 2.04 ± 1.46 cmH 2 O higher than the corresponding estimates from esophageal catheter readings. Significance. Compensating the acquired measurements with system identification techniques makes the readings more accurate, possibly better portraying the patient’s situation for individualization of ventilation therapy.
- 1 January 2024
- Published by IOP Publishing in Physiological Measurement
- Vol. 45 (1) , 015002
Esophageal balloon catheter system identification to improve respiratory effort time features and amplitude determination
Objective . Understanding a patient’s respiratory effort and mechanics is essential for the provision of individualized care during mechanical ventilation. However, measurement of transpulmonary pressure (the difference between airway and pleural pressures) is not easily performed in practice. While airway pressures are available on most mechanical ventilators, pleural pressures are measured indirectly by an esophageal balloon catheter. In many cases, esophageal pressure readings take other phenomena into account and are not a reliable measure of pleural pressure. Approach. A system identification approach was applied to provide accurate pleural measures from esophageal pressure readings. First, we used a closed pressurized chamber to stimulate an esophageal balloon and model its dynamics. Second, we created a simplified version of an artificial lung and tried the model with different ventilation configurations. For validation, data from 11 patients (five male and six female) were used to estimate respiratory effort profile and patient mechanics. Main results. After correcting the dynamic response of the balloon catheter, the estimates of resistance and compliance and the corresponding respiratory effort waveform were improved when compared with the adjusted quantities in the test bench. The performance of the estimated model was evaluated using the respiratory pause/occlusion maneuver, demonstrating improved agreement between the airway and esophageal pressure waveforms when using the normalized mean squared error metric. Using the corrected muscle pressure waveform, we detected start and peak times 130 ± 50 ms earlier and a peak amplitude 2.04 ± 1.46 cmH 2 O higher than the corresponding estimates from esophageal catheter readings. Significance. Compensating the acquired measurements with system identification techniques makes the readings more accurate, possibly better portraying the patient’s situation for individualization of ventilation therapy.