1
|
Schwartz N, Oguz I, Wang J, Pouch A, Yushkevich N, Parameshwaran S, Gee J, Yushkevich P, Oguz B. Fully Automated Placental Volume Quantification From 3D Ultrasound for Prediction of Small-for-Gestational-Age Infants. J Ultrasound Med 2022; 41:1509-1524. [PMID: 34553780 PMCID: PMC8940735 DOI: 10.1002/jum.15835] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Received: 03/31/2021] [Revised: 08/08/2021] [Accepted: 08/20/2021] [Indexed: 06/03/2023]
Abstract
OBJECTIVES Early placental volume (PV) has been associated with small-for-gestational-age infants born under the 10th/5th centiles (SGA10/SGA5). Manual or semiautomated PV quantification from 3D ultrasound (3DUS) is time intensive, limiting its incorporation into clinical care. We devised a novel convolutional neural network (CNN) pipeline for fully automated placenta segmentation from 3DUS images, exploring the association between the calculated PV and SGA. METHODS Volumes of 3DUS obtained from singleton pregnancies at 11-14 weeks' gestation were automatically segmented by our CNN pipeline trained and tested on 99/25 images, combining two 2D and one 3D models with downsampling/upsampling architecture. The PVs derived from the automated segmentations (PVCNN ) were used to train multivariable logistic-regression classifiers for SGA10/SGA5. The test performance for predicting SGA was compared to PVs obtained via the semiautomated VOCAL (GE-Healthcare) method (PVVOCAL ). RESULTS We included 442 subjects with 37 (8.4%) and 18 (4.1%) SGA10/SGA5 infants, respectively. Our segmentation pipeline achieved a mean Dice score of 0.88 on an independent test-set. Adjusted models including PVCNN or PVVOCAL were similarly predictive of SGA10 (area under curve [AUC]: PVCNN = 0.780, PVVOCAL = 0.768). The addition of PVCNN to a clinical model without any PV included (AUC = 0.725) yielded statistically significant improvement in AUC (P < .05); whereas PVVOCAL did not (P = .105). Moreover, when predicting SGA5, including the PVCNN (0.897) brought statistically significant improvement over both the clinical model (0.839, P = .015) and the PVVOCAL model (0.870, P = .039). CONCLUSIONS First trimester PV measurements derived from our CNN segmentation pipeline are significantly associated with future SGA. This fully automated tool enables the incorporation of including placental volumetric biometry into the bedside clinical evaluation as part of a multivariable prediction model for risk stratification and patient counseling.
Collapse
Affiliation(s)
- Nadav Schwartz
- Maternal and Child Health Research Program, Department of
OBGYN, University of Pennsylvania, Philadelphia, PA 19104, USA
| | - Ipek Oguz
- Department of EECS, Vanderbilt University, Nashville, TN
37235-1679, USA
| | - Jiancong Wang
- Penn Image Computing and Science Laboratory (PICSL),
Department of Radiology, University of Pennsylvania, Philadelphia, PA 19104-6025,
USA
| | - Alison Pouch
- Penn Image Computing and Science Laboratory (PICSL),
Department of Radiology, University of Pennsylvania, Philadelphia, PA 19104-6025,
USA
| | - Natalie Yushkevich
- Penn Image Computing and Science Laboratory (PICSL),
Department of Radiology, University of Pennsylvania, Philadelphia, PA 19104-6025,
USA
| | - Shobhana Parameshwaran
- Maternal and Child Health Research Program, Department of
OBGYN, University of Pennsylvania, Philadelphia, PA 19104, USA
| | - James Gee
- Penn Image Computing and Science Laboratory (PICSL),
Department of Radiology, University of Pennsylvania, Philadelphia, PA 19104-6025,
USA
| | - Paul Yushkevich
- Penn Image Computing and Science Laboratory (PICSL),
Department of Radiology, University of Pennsylvania, Philadelphia, PA 19104-6025,
USA
| | - Baris Oguz
- Penn Image Computing and Science Laboratory (PICSL),
Department of Radiology, University of Pennsylvania, Philadelphia, PA 19104-6025,
USA
| |
Collapse
|
2
|
Pasrija C, Quinn R, Ghoreishi M, Eperjesi T, Lai E, Gorman RC, Gorman JH, Gorman RC, Pouch A, Cortez FV, D'Ambra MN, Gammie JS. A Novel Quantitative Ex Vivo Model of Functional Mitral Regurgitation. Innovations (Phila) 2021; 15:329-337. [PMID: 32830572 DOI: 10.1177/1556984520930336] [Citation(s) in RCA: 5] [Impact Index Per Article: 1.7] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 12/29/2022]
Abstract
OBJECTIVE Durability of mitral valve (MV) repair for functional mitral regurgitation (FMR) remains suboptimal. We sought to create a highly reproducible, quantitative ex vivo model of FMR that functions as a platform to test novel repair techniques. METHODS Fresh swine hearts (n = 10) were pressurized with air to a left ventricular pressure of 120 mmHg. The left atrium was excised and the altered geometry of FMR was created by radially dilating the annulus and displacing the papillary muscle tips apically and radially in a calibrated fashion. This was continued in a graduated fashion until coaptation was exhausted. Imaging of the MV was performed with a 3-dimensional (3D) structured-light scanner, which records 3D structure, texture, and color. The model was validated using transesophageal echocardiography in patients with normal MVs and severe FMR. RESULTS Compared to controls, the anteroposterior diameter in the FMR state increased 32% and the annular area increased 35% (P < 0.001). While the anterior annular circumference remained fixed, the posterior circumference increased by 20% (P = 0.026). The annulus became more planar and the tenting height increased 56% (9 to 14 mm, P < 0.001). The median coaptation depth significantly decreased (anterior leaflet: 5 vs 2 mm; posterior leaflet: 7 vs 3 mm, P < 0.001). The ex vivo normal and FMR models had similar characteristics as clinical controls and patients with severe FMR. CONCLUSIONS This novel quantitative ex vivo model provides a simple, reproducible, and inexpensive benchtop representation of FMR that mimics the systolic valvular changes of patients with FMR.
Collapse
Affiliation(s)
- Chetan Pasrija
- 12264 Division of Cardiac Surgery, University of Maryland School of Medicine, Baltimore, MD, USA
| | - Rachael Quinn
- 12264 Division of Cardiac Surgery, University of Maryland School of Medicine, Baltimore, MD, USA
| | - Mehrdad Ghoreishi
- 12264 Division of Cardiac Surgery, University of Maryland School of Medicine, Baltimore, MD, USA
| | - Thomas Eperjesi
- 6572 Department of Surgery, University of Pennsylvania, PA, USA
| | - Eric Lai
- 6572 Department of Surgery, University of Pennsylvania, PA, USA
| | - Robert C Gorman
- 6572 Department of Surgery, University of Pennsylvania, PA, USA
| | - Joseph H Gorman
- 6572 Department of Surgery, University of Pennsylvania, PA, USA
| | - Robert C Gorman
- 6572 Department of Surgery, University of Pennsylvania, PA, USA
| | - Alison Pouch
- 6572 Department of Surgery, University of Pennsylvania, PA, USA
| | - Felino V Cortez
- 12264 Division of Cardiac Surgery, University of Maryland School of Medicine, Baltimore, MD, USA
| | - Michael N D'Ambra
- 12264 Division of Cardiac Surgery, University of Maryland School of Medicine, Baltimore, MD, USA
| | - James S Gammie
- 12264 Division of Cardiac Surgery, University of Maryland School of Medicine, Baltimore, MD, USA
| |
Collapse
|
3
|
Oguz I, Yushkevich N, Pouch A, Oguz BU, Wang J, Parameshwaran S, Gee J, Yushkevich PA, Schwartz N. Minimally interactive placenta segmentation from three-dimensional ultrasound images. J Med Imaging (Bellingham) 2020; 7:014004. [PMID: 32118089 DOI: 10.1117/1.jmi.7.1.014004] [Citation(s) in RCA: 4] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/15/2019] [Accepted: 01/30/2020] [Indexed: 11/14/2022] Open
Abstract
Purpose: Placental size in early pregnancy has been associated with important clinical outcomes, including fetal growth. However, extraction of placental size from three-dimensional ultrasound (3DUS) requires time-consuming interactive segmentation methods and is prone to user variability. We propose a semiautomated segmentation technique that requires minimal user input to robustly measure placental volume from 3DUS images. Approach: For semiautomated segmentation, a single, central 2D slice was manually annotated to initialize an automated multi-atlas label fusion (MALF) algorithm. The dataset consisted of 47 3DUS volumes obtained at 11 to 14 weeks in singleton pregnancies (28 anterior and 19 posterior). Twenty-six of these subjects were imaged twice within the same session. Dice overlap and surface distance were used to quantify the automated segmentation accuracy compared to expert manual segmentations. The mean placental volume measurements obtained by our method and VOCAL (virtual organ computer-aided analysis), a leading commercial semiautomated method, were compared to the manual reference set. The test-retest reliability was also assessed. Results: The overlap between our automated segmentation and manual (mean Dice: 0.824 ± 0.061 , median: 0.831) was within the range reported by other methods requiring extensive manual input. The average surface distance was 1.66 ± 0.96 mm . The correlation coefficient between test-retest volumes was r = 0.88 , and the intraclass correlation was ICC ( 1 ) = 0.86 . Conclusions: MALF is a promising method that can allow accurate and reliable segmentation of the placenta with minimal user interaction. Further refinement of this technique may allow for placental biometry to be incorporated into clinical pregnancy surveillance.
Collapse
Affiliation(s)
- Ipek Oguz
- Vanderbilt University, Department of Electrical Engineering and Computer Science, Nashville, Tennessee, United States.,University of Pennsylvania, Penn Image Computing and Science Laboratory, Department of Radiology, Philadelphia, Pennsylvania, United States
| | - Natalie Yushkevich
- University of Pennsylvania, Penn Image Computing and Science Laboratory, Department of Radiology, Philadelphia, Pennsylvania, United States
| | - Alison Pouch
- University of Pennsylvania, Penn Image Computing and Science Laboratory, Department of Radiology, Philadelphia, Pennsylvania, United States
| | - Baris U Oguz
- University of Pennsylvania, Penn Image Computing and Science Laboratory, Department of Radiology, Philadelphia, Pennsylvania, United States
| | - Jiancong Wang
- University of Pennsylvania, Penn Image Computing and Science Laboratory, Department of Radiology, Philadelphia, Pennsylvania, United States
| | - Shobhana Parameshwaran
- University of Pennsylvania, Department of Obstetrics and Gynecology, Maternal and Child Health Research Program, Philadelphia, Pennsylvania, United States
| | - James Gee
- University of Pennsylvania, Penn Image Computing and Science Laboratory, Department of Radiology, Philadelphia, Pennsylvania, United States
| | - Paul A Yushkevich
- University of Pennsylvania, Penn Image Computing and Science Laboratory, Department of Radiology, Philadelphia, Pennsylvania, United States
| | - Nadav Schwartz
- University of Pennsylvania, Department of Obstetrics and Gynecology, Maternal and Child Health Research Program, Philadelphia, Pennsylvania, United States
| |
Collapse
|
4
|
Vrudhula A, Aly A, Yushkevich P, Pouch A, Shang E, Gorman R, Gorman J, Jackson BM. PC234. Automated Image Analysis to Determine Local Wall Strain of Abdominal Aortic Aneurysms. J Vasc Surg 2019. [DOI: 10.1016/j.jvs.2019.04.410] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/17/2022]
|
5
|
Pouch A, Zaborska A, Pazdro K. The history of hexachlorobenzene accumulation in Svalbard fjords. Environ Monit Assess 2018; 190:360. [PMID: 29799069 PMCID: PMC5968051 DOI: 10.1007/s10661-018-6722-3] [Citation(s) in RCA: 4] [Impact Index Per Article: 0.7] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Received: 12/11/2017] [Accepted: 05/11/2018] [Indexed: 05/18/2023]
Abstract
In the present study, we investigated the spatial and historical trends of hexachlorobenzene (HCB) contamination in dated sediments of three Svalbard fjords (Kongsfjorden, Hornsund, Adventfjorden) differing in environmental conditions and human impact. HCB concentrations ranging from below limit of quantification (6.86 pg/g d.w.) to 143.99 pg/g d.w. were measured. The highest concentrations were measured in two surface sediment layers of the core collected in Hornsund near the melting glacier. The lowest concentrations of HCB were measured in Adventfjorden, suggesting that local source of HCB is not significant and global transport processes are the major transport pathways. The history of HCB deposition did not fully reflect the history of HCB emission (largest in 1950s and 1960s). In case of several sediment cores, the HCB enrichment in surface (recent) sediments was noticed. This can indicate importance of secondary sources of HCB, e.g., the influx of HCB accumulated over decades on the surface of glaciers. Detected levels of HCB were generally low and did not exceed background concentration levels; thus, a negative effect on benthic organisms is not expected.
Collapse
Affiliation(s)
- A Pouch
- Institute of Oceanology Polish Academy of Sciences, Powstańców Warszawy 55, 81-712, Sopot, Poland.
| | - A Zaborska
- Institute of Oceanology Polish Academy of Sciences, Powstańców Warszawy 55, 81-712, Sopot, Poland
| | - K Pazdro
- Institute of Oceanology Polish Academy of Sciences, Powstańców Warszawy 55, 81-712, Sopot, Poland
| |
Collapse
|
6
|
Lynch JE, Pouch A, Sanders R, Hinders M, Rudd K, Sevick J. Gaseous microemboli sizing in extracorporeal circuits using ultrasound backscatter. Ultrasound Med Biol 2007; 33:1661-75. [PMID: 17570578 DOI: 10.1016/j.ultrasmedbio.2007.04.008] [Citation(s) in RCA: 29] [Impact Index Per Article: 1.7] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Received: 10/18/2006] [Revised: 04/11/2007] [Accepted: 04/19/2007] [Indexed: 05/15/2023]
Abstract
This paper describes efforts to estimate the size of gaseous microemboli (GME) in extracorporeal blood circuits based on the amplitude of backscattered ultrasound, starting with analytic modeling of the scattering behavior of GME in blood. After neglecting resonance effects, this model predicts a linear relationship between the amplitude of backscattered echoes and the diameter of GME. Computer simulations based on the cylindrical acoustic finite integration technique were performed to test some of the simplifying assumptions of the analytical model, with the simulations predicting small deviations from the linear approximation that could be treated as random scatter. Ultrasonic and microscopic measurements of injected GME were then performed on a test circuit to determine the linear correlation coefficient between echo amplitude and GME diameter in conditions like those employed in real cardiopulmonary bypass (CPB) circuits. The correlation coefficient determined through this study was further validated in a closed-loop CPB circuit using canine blood. This study shows that the amplitude of ultrasonic backscattered echoes can be used to accurately estimate the size distribution of a population of detected GME when the spacing of emboli is great enough to minimize interference and other multi-path scattering effects. With the high flow rates found in CPB circuits, typically ranging from 2 to 6 L per minute (equivalent to a flow velocity of 0.3 to 1 m/s through the circuit tubing), this assumption will be valid even when hundreds of emboli per second pass through the circuit. Therefore, sizing of GME using the ultrasonic backscatter models described in this paper is a viable method for estimating embolic load delivered to a patient during a CPB procedure.
Collapse
Affiliation(s)
- John E Lynch
- Luna Innovations Incorporated, Hampton, VA 23185, USA.
| | | | | | | | | | | |
Collapse
|