Seth Kligerman
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Seth Kligerman has a total of 55 co-authors. Here are the top co-authors by number of shared publications:
- Victor Pretorius
Division of Cardiothoracic Surgery, UCSD, San Diego, California, USA
3 shared publications - Anderson Scott
Department of Bioengineering, UC San Diego , 9500 Gilman Drive, MC0412, La Jolla, CA 92037
2 shared publications - Andrew Yen
From the UC San Diego Health (A.R., L.M.W., A.Y., J.L., D.B., J.B.), San Diego, CA
2 shared publications - Eric D. Adler
Division of Cardiovascular Medicine, UC San Diego Health, University of California (E.D.A.).
2 shared publications - Francisco Contijoch
Shu Chien-Gene Lay Department of Bioengineering, University of California San Diego, La Jolla, USA
2 shared publications - Jenny Z. Yang
Department of Pulmonary & Critical Care Medicine, University of California San Diego Health, 9434 Medical Ctr Dr, San Diego, CA 92037, USA
2 shared publications
Latest Publications
- 15 July 2025
- Published by Elsevier in The Journal of Heart and Lung Transplantation
- Vol. 44 (7) , S1-S7
Diagnosis of chronic thromboembolic pulmonary hypertension
- 15 July 2025
- Published by Elsevier in The Journal of Heart and Lung Transplantation
- Vol. 44 (7) , S1-S7
Diagnosis of chronic thromboembolic pulmonary hypertension
- Open Access
- 18 July 2023
- Published by Elsevier in CHEST Pulmonary
- Vol. 1 (2)
Outcomes Associated With Catheter-Directed Therapies in Chronic Thromboembolic Pulmonary Hypertension
- 18 July 2023
- Published by Elsevier in CHEST Pulmonary
- Vol. 1 (2)
Open AccessOutcomes Associated With Catheter-Directed Therapies in Chronic Thromboembolic Pulmonary Hypertension
- Open Access
- 4 May 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 10, 48577-48588
Deep Learning Radiographic Assessment of Pulmonary Edema: Optimizing Clinical Performance, Training With Serum Biomarkers
A major obstacle when developing convolutional neural networks (CNNs) for medical imaging is the acquisition of training labels: Most current approaches rely on manual class labels from physicians, which may be challenging to obtain. Clinical biomarkers, often measured alongside medical images and used in diagnostic workup, may provide a rich set of data that can be collected retrospectively and utilized to train diagnostic models. In this work, we focused on assessing the potential of blood serum biomarkers, B-type natriuretic peptide (BNP) and NT-pro B-type natriuretic peptide (BNPP), indicative of acute heart failure (HF) and cardiogenic pulmonary edema to be used as continuously valued labels for training a radiographic deep learning algorithm. For this purpose, a CNN was trained using 27748 radiographs to automatically infer BNP and BNPP, and achieved strong performance (AUC =0.903, sensitivity =0.926, specificity =0.857, $r=0.787$ ). Also, the trained models achieved strong performance (AUC =0.801) for pulmonary edema detection when evaluated with radiologist labels. Since relevant radiographic features visible to the CNN may vary greatly based on image resolution, we also assessed the impact of image resolution on model learning and performance, comparing CNNs trained at five image sizes ( $64\times 64$ to $1024\times 1024$ ). Increasing image resolutions had diminishing but positive gains in AUC. Perhaps more importantly, experiments using three activation mapping techniques (saliency, Grad-CAM, XRAI) revealed considerably increased attention in the lungs with larger image sizes. This result emphasizes the need to utilize radiographs near native resolution for optimal CNN performance, which may not be fully captured by summary metrics like AUC.
- 4 May 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 10, 48577-48588
Open AccessDeep Learning Radiographic Assessment of Pulmonary Edema: Optimizing Clinical Performance, Training With Serum Biomarkers
A major obstacle when developing convolutional neural networks (CNNs) for medical imaging is the acquisition of training labels: Most current approaches rely on manual class labels from physicians, which may be challenging to obtain. Clinical biomarkers, often measured alongside medical images and used in diagnostic workup, may provide a rich set of data that can be collected retrospectively and utilized to train diagnostic models. In this work, we focused on assessing the potential of blood serum biomarkers, B-type natriuretic peptide (BNP) and NT-pro B-type natriuretic peptide (BNPP), indicative of acute heart failure (HF) and cardiogenic pulmonary edema to be used as continuously valued labels for training a radiographic deep learning algorithm. For this purpose, a CNN was trained using 27748 radiographs to automatically infer BNP and BNPP, and achieved strong performance (AUC =0.903, sensitivity =0.926, specificity =0.857, $r=0.787$ ). Also, the trained models achieved strong performance (AUC =0.801) for pulmonary edema detection when evaluated with radiologist labels. Since relevant radiographic features visible to the CNN may vary greatly based on image resolution, we also assessed the impact of image resolution on model learning and performance, comparing CNNs trained at five image sizes ( $64\times 64$ to $1024\times 1024$ ). Increasing image resolutions had diminishing but positive gains in AUC. Perhaps more importantly, experiments using three activation mapping techniques (saliency, Grad-CAM, XRAI) revealed considerably increased attention in the lungs with larger image sizes. This result emphasizes the need to utilize radiographs near native resolution for optimal CNN performance, which may not be fully captured by summary metrics like AUC.
- 15 March 2022
- Published by Wolters Kluwer Health in ASAIO Journal
- Vol. 69 (1) , 69-75
Preoperative Computed Tomography Assessment of Risk of Right Ventricle Failure After Left Ventricular Assist Device Placement
Identification of patients who are at a high risk for right ventricular failure (RVF) after left ventricular assist device (LVAD) implantation is of critical importance. Conventional tools for predicting RVF, including two-dimensional echocardiography, right heart catheterization (RHC), and clinical parameters, generally have limited sensitivity and specificity. We retrospectively examined the ability of computed tomography (CT) ventricular volume measures to identify patients who experienced RVF after LVAD implantation. Between September 2017 and November 2021, 92 patients underwent LVAD surgery at our institution. Preoperative CT-derived ventricular volumes were obtained in 20 patients. Patients who underwent CT evaluation had a similar demographics and rate of RVF after LVAD as patients who did not undergo cardiac CT imaging. In the study cohort, seven of 20 (35%) patients experienced RVF (2 unplanned biventricular assist device, 5 prolonged inotropic support). Computed tomography–derived right ventricular end-diastolic and end-systolic volume indices were the strongest predictors of RVF compared with demographic, echocardiographic, and RHC data with areas under the receiver operating curve of 0.79 and 0.76, respectively. Computed tomography volumetric assessment of RV size can be performed in patients evaluated for LVAD treatment. RV measures of size provide a promising means of pre-LVAD assessment for postoperative RV failure.
- 15 March 2022
- Published by Wolters Kluwer Health in ASAIO Journal
- Vol. 69 (1) , 69-75
Preoperative Computed Tomography Assessment of Risk of Right Ventricle Failure After Left Ventricular Assist Device Placement
Identification of patients who are at a high risk for right ventricular failure (RVF) after left ventricular assist device (LVAD) implantation is of critical importance. Conventional tools for predicting RVF, including two-dimensional echocardiography, right heart catheterization (RHC), and clinical parameters, generally have limited sensitivity and specificity. We retrospectively examined the ability of computed tomography (CT) ventricular volume measures to identify patients who experienced RVF after LVAD implantation. Between September 2017 and November 2021, 92 patients underwent LVAD surgery at our institution. Preoperative CT-derived ventricular volumes were obtained in 20 patients. Patients who underwent CT evaluation had a similar demographics and rate of RVF after LVAD as patients who did not undergo cardiac CT imaging. In the study cohort, seven of 20 (35%) patients experienced RVF (2 unplanned biventricular assist device, 5 prolonged inotropic support). Computed tomography–derived right ventricular end-diastolic and end-systolic volume indices were the strongest predictors of RVF compared with demographic, echocardiographic, and RHC data with areas under the receiver operating curve of 0.79 and 0.76, respectively. Computed tomography volumetric assessment of RV size can be performed in patients evaluated for LVAD treatment. RV measures of size provide a promising means of pre-LVAD assessment for postoperative RV failure.
- Open Access
- 1 April 2020
- Published by Elsevier in The Journal of Heart and Lung Transplantation
- Vol. 39 (4) , S342
Pre-LVAD CT-Derived Measures of RV Size and Function May Be Strong Identifiers of Right Ventricular Failure
- 1 April 2020
- Published by Elsevier in The Journal of Heart and Lung Transplantation
- Vol. 39 (4) , S342
Open AccessPre-LVAD CT-Derived Measures of RV Size and Function May Be Strong Identifiers of Right Ventricular Failure