1
|
Costanzo A, Ertl-Wagner B, Sussman D. AFNet Algorithm for Automatic Amniotic Fluid Segmentation from Fetal MRI. Bioengineering (Basel) 2023; 10:783. [PMID: 37508809 PMCID: PMC10376488 DOI: 10.3390/bioengineering10070783] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/24/2023] [Revised: 06/25/2023] [Accepted: 06/27/2023] [Indexed: 07/30/2023] Open
Abstract
Amniotic Fluid Volume (AFV) is a crucial fetal biomarker when diagnosing specific fetal abnormalities. This study proposes a novel Convolutional Neural Network (CNN) model, AFNet, for segmenting amniotic fluid (AF) to facilitate clinical AFV evaluation. AFNet was trained and tested on a manually segmented and radiologist-validated AF dataset. AFNet outperforms ResUNet++ by using efficient feature mapping in the attention block and transposing convolutions in the decoder. Our experimental results show that AFNet achieved a mean Intersection over Union (mIoU) of 93.38% on our dataset, thereby outperforming other state-of-the-art models. While AFNet achieves performance scores similar to those of the UNet++ model, it does so while utilizing merely less than half the number of parameters. By creating a detailed AF dataset with an improved CNN architecture, we enable the quantification of AFV in clinical practice, which can aid in diagnosing AF disorders during gestation.
Collapse
Affiliation(s)
- Alejo Costanzo
- Department of Electrical, Computer and Biomedical Engineering, Faculty of Engineering and Architectural Sciences, Toronto Metropolitan University, Toronto, ON M5B 2K3, Canada
- Institute for Biomedical Engineering, Science and Technology (iBEST), Toronto Metropolitan University and St. Michael's Hospital, Toronto, ON M5B 1T8, Canada
| | - Birgit Ertl-Wagner
- Department of Diagnostic Imaging, The Hospital for Sick Children, Toronto, ON M5G 1X8, Canada
- Department of Medical Imaging, Faculty of Medicine, University of Toronto, Toronto, ON M5T 1W7, Canada
| | - Dafna Sussman
- Department of Electrical, Computer and Biomedical Engineering, Faculty of Engineering and Architectural Sciences, Toronto Metropolitan University, Toronto, ON M5B 2K3, Canada
- Institute for Biomedical Engineering, Science and Technology (iBEST), Toronto Metropolitan University and St. Michael's Hospital, Toronto, ON M5B 1T8, Canada
- Department of Obstetrics and Gynecology, Faculty of Medicine, University of Toronto, Toronto, ON M5G 1E2, Canada
| |
Collapse
|
2
|
Jha P, Raghu P, Kennedy AM, Sugi M, Morgan TA, Feldstein V, Pōder L, Penna R. Assessment of Amniotic Fluid Volume in Pregnancy. Radiographics 2023; 43:e220146. [PMID: 37200220 DOI: 10.1148/rg.220146] [Citation(s) in RCA: 1] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 05/20/2023]
Abstract
Amniotic fluid (AF) is an integral part of the fetal environment and is essential for fetal growth and development. Pathways of AF recirculation include the fetal lungs, swallowing, absorption through the fetal gastrointestinal tract, excretion through fetal urine production, and movement. In addition to being a marker for fetal health, adequate AF is necessary for fetal lung development, growth, and movement. The role of diagnostic imaging is to provide a detailed fetal survey, placental evaluation, and clinical correlation with maternal conditions to help identify causes of AF abnormalities and thereby enable specific therapy. Oligohydramnios prompts evaluation for fetal growth restriction as well as genitourinary issues, including renal agenesis, multicystic dysplastic kidneys, ureteropelvic junction obstruction, and bladder outlet obstruction. Premature preterm rupture of membranes should also be clinically excluded as a cause of oligohydramnios. Clinical trials evaluating amnioinfusion are underway as a potential intervention for renal causes of oligohydramnios. Most cases of polyhydramnios are idiopathic, with maternal diabetes being a common cause. Polyhydramnios prompts evaluation for fetal gastrointestinal obstruction and oropharyngeal or thoracic masses, as well as neurologic or musculoskeletal anomalies. Amnioreduction is performed only for maternal indications such as symptomatic polyhydramnios causing maternal respiratory distress. Polyhydramnios with fetal growth restriction is paradoxical and can occur with maternal diabetes and hypertension. When these maternal conditions are absent, this raises concern for aneuploidy. The authors describe the pathways of AF production and circulation, US and MRI assessment of AF, disease-specific disruption of AF pathways, and an algorithmic approach to AF abnormalities. ©RSNA, 2023 Online supplemental material is available for this article. Quiz questions for this article are available through the Online Learning Center.
Collapse
Affiliation(s)
- Priyanka Jha
- From the University of California San Francisco, 505 Parnassus Ave, San Francisco, Calif (P.J., P.R., M.S., T.A.M., V.F., L.P.); University of Utah Hospital, Salt Lake City, Utah (A.M.K.); and Inland Imaging, Seattle, Wash (R.P.)
| | - Preethi Raghu
- From the University of California San Francisco, 505 Parnassus Ave, San Francisco, Calif (P.J., P.R., M.S., T.A.M., V.F., L.P.); University of Utah Hospital, Salt Lake City, Utah (A.M.K.); and Inland Imaging, Seattle, Wash (R.P.)
| | - Anne M Kennedy
- From the University of California San Francisco, 505 Parnassus Ave, San Francisco, Calif (P.J., P.R., M.S., T.A.M., V.F., L.P.); University of Utah Hospital, Salt Lake City, Utah (A.M.K.); and Inland Imaging, Seattle, Wash (R.P.)
| | - Mark Sugi
- From the University of California San Francisco, 505 Parnassus Ave, San Francisco, Calif (P.J., P.R., M.S., T.A.M., V.F., L.P.); University of Utah Hospital, Salt Lake City, Utah (A.M.K.); and Inland Imaging, Seattle, Wash (R.P.)
| | - Tara A Morgan
- From the University of California San Francisco, 505 Parnassus Ave, San Francisco, Calif (P.J., P.R., M.S., T.A.M., V.F., L.P.); University of Utah Hospital, Salt Lake City, Utah (A.M.K.); and Inland Imaging, Seattle, Wash (R.P.)
| | - Vickie Feldstein
- From the University of California San Francisco, 505 Parnassus Ave, San Francisco, Calif (P.J., P.R., M.S., T.A.M., V.F., L.P.); University of Utah Hospital, Salt Lake City, Utah (A.M.K.); and Inland Imaging, Seattle, Wash (R.P.)
| | - Liina Pōder
- From the University of California San Francisco, 505 Parnassus Ave, San Francisco, Calif (P.J., P.R., M.S., T.A.M., V.F., L.P.); University of Utah Hospital, Salt Lake City, Utah (A.M.K.); and Inland Imaging, Seattle, Wash (R.P.)
| | - Rubal Penna
- From the University of California San Francisco, 505 Parnassus Ave, San Francisco, Calif (P.J., P.R., M.S., T.A.M., V.F., L.P.); University of Utah Hospital, Salt Lake City, Utah (A.M.K.); and Inland Imaging, Seattle, Wash (R.P.)
| |
Collapse
|
3
|
Souza SCS, Kim K, Dingwall-Harvey ALJ, Fakhraei R, Liao Y, Gaudet LM. Evaluating the correlation between amniotic fluid volume and estimated fetal weight in healthy pregnant women. J Perinat Med 2022; 50:1096-1099. [PMID: 35589556 DOI: 10.1515/jpm-2022-0104] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 02/23/2022] [Accepted: 04/25/2022] [Indexed: 11/15/2022]
Abstract
OBJECTIVES The establishment of cut-offs for normal amniotic fluid volume (AFV) is valuable to predict perinatal outcomes. However, the most common methods to measure AFV are not accurate enough. It is important to understand factors that may be able to increase the accuracy of the calculation of AFV cut-off values. The objective of this study was to verify the correlation between AFV and estimated fetal weight (EFW). METHODS Records from almost 7,000 patients between 2012 and 2017 were accessed through hospital databases. The AFV measurements included in our analysis were obtained using the maximum vertical pocket technique. RESULTS AFV was positively correlated with EFW in the overall, male and female samples; however, the magnitude of the association was small (0.1 CONCLUSIONS The incorporation of EFW together with other factors (e.g., gestational age, fetus sex) may increase the accuracy of the AFV cut-offs calculation and, ultimately, reduce morbidity.
Collapse
Affiliation(s)
- Sara C S Souza
- OMNI Research Group, Clinical Epidemiology Program, Ottawa Hospital Research Institute, Ottawa, Canada
| | - Katherine Kim
- OMNI Research Group, Clinical Epidemiology Program, Ottawa Hospital Research Institute, Ottawa, Canada
| | | | - Romina Fakhraei
- OMNI Research Group, Clinical Epidemiology Program, Ottawa Hospital Research Institute, Ottawa, Canada
- School of Epidemiology and Public Health, University of Ottawa, Ottawa, Canada
| | - Yan Liao
- OMNI Research Group, Clinical Epidemiology Program, Ottawa Hospital Research Institute, Ottawa, Canada
| | - Laura M Gaudet
- OMNI Research Group, Clinical Epidemiology Program, Ottawa Hospital Research Institute, Ottawa, Canada
- School of Epidemiology and Public Health, University of Ottawa, Ottawa, Canada
- Department of Obstetrics and Gynecology, Kingston Health Sciences Center, Kingston, Canada
- Department of Obstetrics and Gynecology, Queen's University, Kingston, Canada
| |
Collapse
|
4
|
Amgalan A, Kapse K, Krishnamurthy D, Andersen NR, Izem R, Baschat A, Quistorff J, Gimovsky AC, Ahmadzia HK, Limperopoulos C, Andescavage NN. Measuring intrauterine growth in healthy pregnancies using quantitative magnetic resonance imaging. J Perinatol 2022; 42:860-865. [PMID: 35194161 PMCID: PMC9380865 DOI: 10.1038/s41372-022-01340-6] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 04/30/2021] [Revised: 11/04/2021] [Accepted: 02/03/2022] [Indexed: 11/09/2022]
Abstract
OBJECTIVE The aim of this study was to determine in utero fetal-placental growth patterns using in vivo three-dimensional (3D) quantitative magnetic resonance imaging (qMRI). STUDY DESIGN Healthy women with singleton pregnancies underwent fetal MRI to measure fetal body, placenta, and amniotic space volumes. The fetal-placental ratio (FPR) was derived using 3D fetal body and placental volumes (PV). Descriptive statistics were used to describe the association of each measurement with increasing gestational age (GA) at MRI. RESULTS Fifty-eight (58) women underwent fetal MRI between 16 and 38 completed weeks gestation (mean = 28.12 ± 6.33). PV and FPR varied linearly with GA at MRI (rPV,GA = 0.83, rFPR,GA = 0.89, p value < 0.001). Fetal volume varied non-linearly with GA (p value < 0.01). CONCLUSIONS We describe in-utero growth trajectories of fetal-placental volumes in healthy pregnancies using qMRI. Understanding healthy in utero development can establish normative benchmarks where departures from normal may identify early in utero placental failure prior to the onset of fetal harm.
Collapse
Affiliation(s)
- Ariunzaya Amgalan
- School of Medicine, Georgetown University, 3900 Reservoir Road, NW, Washington, DC, 20057, USA
| | - Kushal Kapse
- Division of Diagnostic Imaging & Radiology, Children's National Hospital, 111 Michigan Ave. NW, Washington, DC, 20010, USA
| | - Dhineshvikram Krishnamurthy
- Division of Diagnostic Imaging & Radiology, Children's National Hospital, 111 Michigan Ave. NW, Washington, DC, 20010, USA
| | - Nicole R Andersen
- Division of Diagnostic Imaging & Radiology, Children's National Hospital, 111 Michigan Ave. NW, Washington, DC, 20010, USA
| | - Rima Izem
- Division of Biostatistics & Study Methodology, Children's National Hospital, 111 Michigan Ave. NW, Washington, DC, 20010, USA
| | - Ahmet Baschat
- Center for Fetal Therapy, Department of Gynecology and Obstetrics, Johns Hopkins Hospital, Baltimore, MD, 21287, USA
| | - Jessica Quistorff
- Division of Diagnostic Imaging & Radiology, Children's National Hospital, 111 Michigan Ave. NW, Washington, DC, 20010, USA
| | - Alexis C Gimovsky
- Division of Maternal-Fetal Medicine, Department of Obstetrics & Gynecology, George Washington University, Washington, DC, 20037, USA
| | - Homa K Ahmadzia
- Division of Maternal-Fetal Medicine, Department of Obstetrics & Gynecology, George Washington University, Washington, DC, 20037, USA
| | - Catherine Limperopoulos
- Division of Diagnostic Imaging & Radiology, Children's National Hospital, 111 Michigan Ave. NW, Washington, DC, 20010, USA. .,Department of Pediatrics, George Washington University, Washington, DC, 20037, USA.
| | - Nickie N Andescavage
- Department of Pediatrics, George Washington University, Washington, DC, 20037, USA.,Division of Neonatology, Children's National Hospital, 111 Michigan Ave. NW, Washington, DC, 20010, USA
| |
Collapse
|
5
|
Novel Computerized Analytic Technique for Quantification of Amniotic Fluid Volume in Fetal MRI. AJR Am J Roentgenol 2019; 213:W149-W152. [PMID: 31237453 DOI: 10.2214/ajr.19.21275] [Citation(s) in RCA: 5] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/18/2022]
Abstract
OBJECTIVE. Fetal MRI is increasingly used in the evaluation of suspected congenital anomalies. Assessment of amniotic fluid volume (AFV) is crucial, but no automated quantitative technique is currently available for MRI. The purpose of this study was to develop and evaluate an analytic technique for quantifying AFV in fetal MRI. MATERIALS AND METHODS. Two MRI phantoms containing known quantities of synthetic amniotic fluid were created. A 3D steady-state free precession sequence was used for 1.5-T MRI of the phantoms and as part of a standard clinical fetal MRI protocol. Software was developed and used to retrospectively calculate AFV for the phantom and 20 clinical MRI examinations. Times to completion were recorded. AFV was also calculated by a manual hand-tracing method. To evaluate performance, paired t tests were used to compare computer-generated measurements with known phantom volumes. Intraclass correlation coefficients were calculated to assess agreement between computer-generated and manual measurements. RESULTS. There was no significant difference between computer-generated measurements of known AFV in the MRI phantoms (p > 0.11). When the software program was applied to the clinical MRI examinations, the mean time to complete AFV measurement was 110 seconds. There was excellent reliability between total AFV calculated by the two software users and by means of manual measurements (intraclass correlation coefficient, 0.995; p < 0.01). CONCLUSION. The computerized analysis evaluated in this study rapidly and accurately quantifies AFV in fetal MRI. The results are concordant with known phantom volumes and manual measurements. The technique is promising for objective MRI evaluation of AFV and has the potential to improve prenatal diagnosis and management.
Collapse
|
6
|
Abstract
OBJECTIVE The purpose of this study is to determine which MRI parameters of fetal head and neck masses predict high-morbidity neonatal outcomes, including ex utero intrapartum treatment (EXIT) procedure. MATERIALS AND METHODS This retrospective study (2004-2016) included parameters of polyhydramnios (based on largest vertical pocket), mass effect on the trachea, mass midline extension, and morphologic grade and size of masses. The morbid cohort included those requiring an EXIT procedure, difficult intubation at delivery, or lethal outcome. Predictive modeling with a multivariable logistic regression and ROC analysis was then performed. RESULTS Of 36 fetuses, five were delivered by EXIT procedures, there was one neonatal death within 12 hours after delivery, and another neonate required multiple intubation attempts. The remaining 29 fetuses were delivered at outside institutions with no interventions or neonatal morbidity. The largest vertical pocket and mass effect on the trachea were selected as independent predictors by the logistic regression. The cross-validated ROC AUC was 0.951 (95% CI, 0.8795-1). CONCLUSION The largest vertical pocket measurement and mass effect on the trachea were the most contributory MRI parameters that predicted significant morbidity in fetuses with masses of the face and neck, along with other significant parameters. These parameters predict significant morbid neonatal outcomes, including the need for EXIT procedures.
Collapse
|
7
|
Zhao SX, Xiao YH, Lv FR, Zhang ZW, Sheng B, Ma HL. Lateral ventricular volume measurement by 3D MR hydrography in fetal ventriculomegaly and normal lateral ventricles. J Magn Reson Imaging 2017; 48:266-273. [PMID: 29251798 DOI: 10.1002/jmri.25927] [Citation(s) in RCA: 6] [Impact Index Per Article: 0.9] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/01/2017] [Accepted: 11/30/2017] [Indexed: 11/06/2022] Open
Abstract
BACKGROUND In fetuses with prenatal ventriculomegaly (VM), ventricular volume on MRI has been shown to correlate with poor postnatal outcomes and in utero death. 3D magnetic resonance hydrography (MRH) has been widely used for MR cholangiopancreatography. PURPOSE To investigate the reliability of 3D MRH for lateral ventricular volume measurement in fetuses with VM and normal lateral ventricles, using manual multisection planimetry (MSP) as a reference standard. STUDY TYPE Prospective study. POPULATION Thirty-five fetuses with VM at 24-37 gestational weeks (GA) and 35 fetuses with normal lateral ventricles at 24-38 GA. FIELD STRENGTH/SEQUENCE 1.5T MRI with 3D MRH and T2 -weighted single-shot fast-spin echo sequence. ASSESSMENT Left, right, and total lateral ventricle volumes in fetuses were acquired from 3D MRH and manual MSP. All image analysis was performed by a radiologist twice and another radiologist once, blindly. STATISTICAL TESTS Analysis of linear regression analysis, Pearson's correlation coefficient, Bland-Altman plots, intraclass correlation coefficient (ICC), and independent samples t-test were used for statistical analyses. RESULTS There were highly significant relationships between all 3D MRH and manual MSP measurements of lateral ventricular volumes (rVM = 0.92-0.98; rN = 0.95-0.98; all P < 0.0001; VM: VM group, N: normal group), although left, right, and total lateral ventricular volumes measured by 3D MRH tended to be slightly larger than MSP (biasVM 0.1 ± 0.95, 0.26 ± 0.63, and 0.3 ± 0.68 mL, respectively; biasN 0.1 ± 0.95, 0.26 ± 0.63, and 0.3 ± 0.68 mL, respectively). Interrater agreement and intrarater repeatability were also excellent for 3D MRH (ICCVM = 0.994-0.99, ICCN = 0.989-0.992; ICCVM = 0.975-0.987, ICCN = 0.958-0.971, respectively). 3D MRH showed significantly reduced measurement time (VM: 3.55 ± 0.42 vs. 11.81 ± 0.13 min; N: 3.08 ± 0.39 vs. 12.12 ± 0.11 min; all P < 0.0001). DATA CONCLUSION Lateral ventricular volume measurement by 3D MRH was comparable to manual MSP. LEVEL OF EVIDENCE 1 Technical Efficacy Stage 1 J. Magn. Reson. Imaging 2017.
Collapse
Affiliation(s)
- Si-Xiu Zhao
- Department of Radiology, First Affiliated Hospital of Chongqing Medical University, Yuanjiagang, Yuzhong District, Chongqing, China
| | - Yun-Hua Xiao
- Department of Radiology, People's Hospital of Chongqing Yubei District, Yubei District, Chongqing, China
| | - Fu-Rong Lv
- Department of Radiology, First Affiliated Hospital of Chongqing Medical University, Yuanjiagang, Yuzhong District, Chongqing, China
| | - Zhi-Wei Zhang
- Department of Radiology, First Affiliated Hospital of Chongqing Medical University, Yuanjiagang, Yuzhong District, Chongqing, China
| | - Bo Sheng
- Department of Radiology, First Affiliated Hospital of Chongqing Medical University, Yuanjiagang, Yuzhong District, Chongqing, China
| | - Hong-Li Ma
- Department of Radiology, First Affiliated Hospital of Chongqing Medical University, Yuanjiagang, Yuzhong District, Chongqing, China
| |
Collapse
|