RESUMEN
In-utero fetal MRI is emerging as an important tool in the diagnosis and analysis of the developing human brain. Automatic segmentation of the developing fetal brain is a vital step in the quantitative analysis of prenatal neurodevelopment both in the research and clinical context. However, manual segmentation of cerebral structures is time-consuming and prone to error and inter-observer variability. Therefore, we organized the Fetal Tissue Annotation (FeTA) Challenge in 2021 in order to encourage the development of automatic segmentation algorithms on an international level. The challenge utilized FeTA Dataset, an open dataset of fetal brain MRI reconstructions segmented into seven different tissues (external cerebrospinal fluid, gray matter, white matter, ventricles, cerebellum, brainstem, deep gray matter). 20 international teams participated in this challenge, submitting a total of 21 algorithms for evaluation. In this paper, we provide a detailed analysis of the results from both a technical and clinical perspective. All participants relied on deep learning methods, mainly U-Nets, with some variability present in the network architecture, optimization, and image pre- and post-processing. The majority of teams used existing medical imaging deep learning frameworks. The main differences between the submissions were the fine tuning done during training, and the specific pre- and post-processing steps performed. The challenge results showed that almost all submissions performed similarly. Four of the top five teams used ensemble learning methods. However, one team's algorithm performed significantly superior to the other submissions, and consisted of an asymmetrical U-Net network architecture. This paper provides a first of its kind benchmark for future automatic multi-tissue segmentation algorithms for the developing human brain in utero.
Asunto(s)
Procesamiento de Imagen Asistido por Computador , Sustancia Blanca , Embarazo , Femenino , Humanos , Procesamiento de Imagen Asistido por Computador/métodos , Encéfalo/diagnóstico por imagen , Cabeza , Feto/diagnóstico por imagen , Algoritmos , Imagen por Resonancia Magnética/métodosRESUMEN
Objective:To investigate the diagnostic value of prenatal MRI in the detection of abnormal placental cord insertions (APCIs) comparing with prenatal ultrasound and pathological examination.Methods:A retrospective data collection was conducted on 440 patients who underwent both prenatal placental ultrasound and MRI at the Foshan Women and Children Hospital from December 2013 to December 2021. Among them, 37 cases were APCIs confirmed by surgery or pathology. The prenatal placental MRI findings were analyzed and compared with prenatal ultrasound diagnosis. The diagnostic efficacy of prenatal MRI and ultrasound in diagnosing APCIs was calculated.Results:Among the 37 cases of APCIs confirmed by surgery or pathology, 17 cases had marginal cord insertion (MCI), 13 cases had velamentous cord insertion (VCI), 5 cases had vasa previa (VP), and 2 cases had VCI combined with VP. The sensitivity and specificity of ultrasound diagnosis for APCIs were 59.5% (22/37) and 97.8% (394/403), respectively. The sensitivity and specificity of MRI diagnosis for APCIs were 86.5% (32/37) and 98.5% (397/403), respectively. Among the 37 cases of APCIs, prenatal MRI missed diagnosis of 2 cases of MCI, 2 cases of VCI, and misdiagnosed 1 case of VCI as an accessory placenta. MRI identified 10 cases of APCIs missed by ultrasound, including 5 cases of MCI, 2 cases of VP, 2 cases of VCI, and 1 case of combined VCI with VP. Additionally, ultrasound misdiagnosed 4 cases of APCIs, including 2 cases of VCI misdiagnosed as MCI and 2 cases of MCI misdiagnosed as VCI.Conclusions:For APCIs complicated with abnormalities of placental location or morphology, or placental accretion spectrum disease in late pregnancy, MRI has a higher diagnostic efficacy than ultrasound.