RESUMEN
The Sparsely Annotated Region and Organ Segmentation (SAROS) dataset was created using data from The Cancer Imaging Archive (TCIA) to provide a large open-access CT dataset with high-quality annotations of body landmarks. In-house segmentation models were employed to generate annotation proposals on randomly selected cases from TCIA. The dataset includes 13 semantic body region labels (abdominal/thoracic cavity, bones, brain, breast implant, mediastinum, muscle, parotid/submandibular/thyroid glands, pericardium, spinal cord, subcutaneous tissue) and six body part labels (left/right arm/leg, head, torso). Case selection was based on the DICOM series description, gender, and imaging protocol, resulting in 882 patients (438 female) for a total of 900 CTs. Manual review and correction of proposals were conducted in a continuous quality control cycle. Only every fifth axial slice was annotated, yielding 20150 annotated slices from 28 data collections. For the reproducibility on downstream tasks, five cross-validation folds and a test set were pre-defined. The SAROS dataset serves as an open-access resource for training and evaluating novel segmentation models, covering various scanner vendors and diseases.
Asunto(s)
Tomografía Computarizada por Rayos X , Imagen de Cuerpo Entero , Femenino , Humanos , Masculino , Procesamiento de Imagen Asistido por ComputadorRESUMEN
OBJECTIVES: Accurately acquiring and assigning different contrast-enhanced phases in computed tomography (CT) is relevant for clinicians and for artificial intelligence orchestration to select the most appropriate series for analysis. However, this information is commonly extracted from the CT metadata, which is often wrong. This study aimed at developing an automatic pipeline for classifying intravenous (IV) contrast phases and additionally for identifying contrast media in the gastrointestinal tract (GIT). MATERIALS AND METHODS: This retrospective study used 1200 CT scans collected at the investigating institution between January 4, 2016 and September 12, 2022, and 240 CT scans from multiple centers from The Cancer Imaging Archive for external validation. The open-source segmentation algorithm TotalSegmentator was used to identify regions of interest (pulmonary artery, aorta, stomach, portal/splenic vein, liver, portal vein/hepatic veins, inferior vena cava, duodenum, small bowel, colon, left/right kidney, urinary bladder), and machine learning classifiers were trained with 5-fold cross-validation to classify IV contrast phases (noncontrast, pulmonary arterial, arterial, venous, and urographic) and GIT contrast enhancement. The performance of the ensembles was evaluated using the receiver operating characteristic area under the curve (AUC) and 95% confidence intervals (CIs). RESULTS: For the IV phase classification task, the following AUC scores were obtained for the internal test set: 99.59% [95% CI, 99.58-99.63] for the noncontrast phase, 99.50% [95% CI, 99.49-99.52] for the pulmonary-arterial phase, 99.13% [95% CI, 99.10-99.15] for the arterial phase, 99.8% [95% CI, 99.79-99.81] for the venous phase, and 99.7% [95% CI, 99.68-99.7] for the urographic phase. For the external dataset, a mean AUC of 97.33% [95% CI, 97.27-97.35] and 97.38% [95% CI, 97.34-97.41] was achieved for all contrast phases for the first and second annotators, respectively. Contrast media in the GIT could be identified with an AUC of 99.90% [95% CI, 99.89-99.9] in the internal dataset, whereas in the external dataset, an AUC of 99.73% [95% CI, 99.71-99.73] and 99.31% [95% CI, 99.27-99.33] was achieved with the first and second annotator, respectively. CONCLUSIONS: The integration of open-source segmentation networks and classifiers effectively classified contrast phases and identified GIT contrast enhancement using anatomical landmarks.
Asunto(s)
Medios de Contraste , Aprendizaje Automático , Tomografía Computarizada por Rayos X , Humanos , Estudios Retrospectivos , Tomografía Computarizada por Rayos X/métodos , Masculino , Femenino , Tracto Gastrointestinal/diagnóstico por imagen , Interpretación de Imagen Radiográfica Asistida por Computador/métodos , Persona de Mediana Edad , AlgoritmosRESUMEN
Background: This study aimed to evaluate the impact of an AI-assisted fracture detection program on radiology residents' performance in pediatric and adult trauma patients and assess its implications for residency training. Methods: This study, conducted retrospectively, included 200 radiographs from participants aged 1 to 95 years (mean age: 40.7 ± 24.5 years), encompassing various body regions. Among these, 50% (100/200) displayed at least one fracture, totaling one hundred thirty-five fractures, assessed by four radiology residents with different experience levels. A machine learning algorithm was employed for fracture detection, and the ground truth was established by consensus among two experienced senior radiologists. Fracture detection accuracy, reporting time, and confidence were evaluated with and without AI support. Results: Radiology residents' sensitivity for fracture detection improved significantly with AI support (58% without AI vs. 77% with AI, p < 0.001), while specificity showed minor improvements (77% without AI vs. 79% with AI, p = 0.0653). AI stand-alone performance achieved a sensitivity of 93% with a specificity of 77%. AI support for fracture detection significantly reduced interpretation time for radiology residents by an average of approximately 2.6 s (p = 0.0156) and increased resident confidence in the findings (p = 0.0013). Conclusion: AI support significantly enhanced fracture detection sensitivity among radiology residents, particularly benefiting less experienced radiologists. It does not compromise specificity and reduces interpretation time, contributing to improved efficiency. This study underscores AI's potential in radiology, emphasizing its role in training and interpretation improvement.
RESUMEN
PURPOSE: The study aimed to develop the open-source body and organ analysis (BOA), a comprehensive computed tomography (CT) image segmentation algorithm with a focus on workflow integration. METHODS: The BOA combines 2 segmentation algorithms: body composition analysis (BCA) and TotalSegmentator. The BCA was trained with the nnU-Net framework using a dataset including 300 CT examinations. The CTs were manually annotated with 11 semantic body regions: subcutaneous tissue, muscle, bone, abdominal cavity, thoracic cavity, glands, mediastinum, pericardium, breast implant, brain, and spinal cord. The models were trained using 5-fold cross-validation, and at inference time, an ensemble was used. Afterward, the segmentation efficiency was evaluated on a separate test set comprising 60 CT scans. In a postprocessing step, a tissue segmentation (muscle, subcutaneous adipose tissue, visceral adipose tissue, intermuscular adipose tissue, epicardial adipose tissue, and paracardial adipose tissue) is created by subclassifying the body regions. The BOA combines this algorithm and the open-source segmentation software TotalSegmentator to have an all-in-one comprehensive selection of segmentations. In addition, it integrates into clinical workflows as a DICOM node-triggered service using the open-source Orthanc research PACS (Picture Archiving and Communication System) server to make the automated segmentation algorithms available to clinicians. The BCA model's performance was evaluated using the Sørensen-Dice score. Finally, the segmentations from the 3 different tools (BCA, TotalSegmentator, and BOA) were compared by assessing the overall percentage of the segmented human body on a separate cohort of 150 whole-body CT scans. RESULTS: The results showed that the BCA outperformed the previous publication, achieving a higher Sørensen-Dice score for the previously existing classes, including subcutaneous tissue (0.971 vs 0.962), muscle (0.959 vs 0.933), abdominal cavity (0.983 vs 0.973), thoracic cavity (0.982 vs 0.965), bone (0.961 vs 0.942), and an overall good segmentation efficiency for newly introduced classes: brain (0.985), breast implant (0.943), glands (0.766), mediastinum (0.880), pericardium (0.964), and spinal cord (0.896). All in all, it achieved a 0.935 average Sørensen-Dice score, which is comparable to the one of the TotalSegmentator (0.94). The TotalSegmentator had a mean voxel body coverage of 31% ± 6%, whereas BCA had a coverage of 75% ± 6% and BOA achieved 93% ± 2%. CONCLUSIONS: The open-source BOA merges different segmentation algorithms with a focus on workflow integration through DICOM node integration, offering a comprehensive body segmentation in CT images with a high coverage of the body volume.
RESUMEN
The complex process of manual biomarker extraction from body composition analysis (BCA) has far restricted the analysis of SARS-CoV-2 outcomes to small patient cohorts and a limited number of tissue types. We investigate the association of two BCA-based biomarkers with the development of severe SARS-CoV-2 infections for 918 patients (354 female, 564 male) regarding disease severity and mortality (186 deceased). Multiple tissues, such as muscle, bone, or adipose tissue are used and acquired with a deep-learning-based, fully-automated BCA from computed tomography images of the chest. The BCA features and markers were univariately analyzed with a Shapiro-Wilk and two-sided Mann-Whitney-U test. In a multivariate approach, obtained markers were adjusted by a defined set of laboratory parameters promoted by other studies. Subsequently, the relationship between the markers and two endpoints, namely severity and mortality, was investigated with regard to statistical significance. The univariate approach showed that the muscle volume was significant for female (pseverity ≤ 0.001, pmortality ≤ 0.0001) and male patients (pseverity = 0.018, pmortality ≤ 0.0001) regarding the severity and mortality endpoints. For male patients, the intra- and intermuscular adipose tissue (IMAT) (p ≤ 0.0001), epicardial adipose tissue (EAT) (p ≤ 0.001) and pericardial adipose tissue (PAT) (p ≤ 0.0001) were significant regarding the severity outcome. With the mortality outcome, muscle (p ≤ 0.0001), IMAT (p ≤ 0.001), EAT (p = 0.011) and PAT (p = 0.003) remained significant. For female patients, bone (p ≤ 0.001), IMAT (p = 0.032) and PAT (p = 0.047) were significant in univariate analyses regarding the severity and bone (p = 0.005) regarding the mortality. Furthermore, the defined sarcopenia marker (p ≤ 0.0001, for female and male) was significant for both endpoints. The cardiac marker was significant for severity (pfemale = 0.014, pmale ≤ 0.0001) and for mortality (pfemale ≤ 0.0001, pmale ≤ 0.0001) endpoint for both genders. The multivariate logistic regression showed that the sarcopenia marker was significant (pseverity = 0.006, pmortality = 0.002) for both endpoints (ORseverity = 0.42, 95% CIseverity: 0.23-0.78, ORmortality = 0.34, 95% CImortality: 0.17-0.67). The cardiac marker showed significance (p = 0.018) only for the severity endpoint (OR = 1.42, 95% CI 1.06-1.90). The association between BCA-based sarcopenia and cardiac biomarkers and disease severity and mortality suggests that these biomarkers can contribute to the risk stratification of SARS-CoV-2 patients. Patients with a higher cardiac marker and a lower sarcopenia marker are at risk for a severe course or death. Whether those biomarkers hold similar importance for other pneumonia-related diseases requires further investigation.
Asunto(s)
COVID-19 , Sarcopenia , Tejido Adiposo/diagnóstico por imagen , Biomarcadores , Composición Corporal , Femenino , Humanos , Masculino , Estudios Retrospectivos , SARS-CoV-2 , Sarcopenia/diagnóstico por imagen , Tomografía Computarizada por Rayos X/métodosRESUMEN
Patients with neuroendocrine tumors of gastro-entero-pancreatic origin (GEP-NET) experience changes in fat and muscle composition. Dual-energy X-ray absorptiometry (DXA) and bioelectrical impedance analysis (BIA) are currently used to analyze body composition. Changes thereof could indicate cancer progression or response to treatment. This study examines the correlation between CT-based (computed tomography) body composition analysis (BCA) and DXA or BIA measurement. 74 GEP-NET-patients received whole-body [68Ga]-DOTATOC-PET/CT, BIA, and DXA-scans. BCA was performed based on the non-contrast-enhanced, 5 mm, whole-body-CT images. BCA from CT shows a strong correlation between body fat ratio with DXA (r = 0.95, ρC = 0.83) and BIA (r = 0.92, ρC = 0.76) and between skeletal muscle ratio with BIA: r = 0.81, ρC = 0.49. The deep learning-network achieves highly accurate results (mean Sørensen-Dice-score 0.93). Using BCA on routine Positron emission tomography/CT-scans to monitor patients' body composition in the diagnostic workflow can reduce additional exams whilst substantially amplifying measurement in slower progressing cancers such as GEP-NET.
Asunto(s)
Composición Corporal , Tomografía Computarizada por Tomografía de Emisión de Positrones , Absorciometría de Fotón/métodos , Composición Corporal/fisiología , Índice de Masa Corporal , Impedancia Eléctrica , Humanos , Tomografía Computarizada por Rayos XRESUMEN
(1) Background: Epi- and Paracardial Adipose Tissue (EAT, PAT) have been spotlighted as important biomarkers in cardiological assessment in recent years. Since biomarker quantification is an increasingly important method for clinical use, we wanted to examine fully automated EAT and PAT quantification for possible use in cardiovascular risk stratification. (2) Methods: 966 patients with intermediate Framingham risk scores for Coronary Artery Disease referred for coronary calcium scans were included in clinical routine retrospectively. The Coronary Artery Calcium Score (CACS) was extracted and tissue quantification was performed by a deep learning network. (3) Results: The Computed Tomography (CT) segmentations predicted by the network indicated no significant correlation between EAT volume and EAT radiodensity when compared to Agatston score (r = 0.18, r = -0.09). CACS 0 category patients showed significantly lower levels of total EAT and PAT volumes and higher EAT and PAT densities than CACS 1-99 category patients (p < 0.01). Notably, this difference did not reach significance regarding EAT attenuation in male patients. Women older than 50 years, thus more likely to be postmenopausal, were shown to be at higher risk of coronary calcification (p < 0.01, OR = 4.59). CACS 1-99 vs. CACS ≥100 category patients remained below significance level (EAT volume: p = 0.087, EAT attenuation: p = 0.98). (4) Conclusions: Our study proves the feasibility of a fully automated adipose tissue analysis in clinical cardiac CT and confirms in a large clinical cohort that volume and attenuation of EAT and PAT are not correlated with CACS. Broadly available deep learning based rapid and reliable tissue quantification should thus be discussed as a method to assess this biomarker as a supplementary risk predictor in cardiac CT.
RESUMEN
OBJECTIVES: Body tissue composition is a long-known biomarker with high diagnostic and prognostic value not only in cardiovascular, oncological, and orthopedic diseases but also in rehabilitation medicine or drug dosage. In this study, the aim was to develop a fully automated, reproducible, and quantitative 3D volumetry of body tissue composition from standard CT examinations of the abdomen in order to be able to offer such valuable biomarkers as part of routine clinical imaging. METHODS: Therefore, an in-house dataset of 40 CTs for training and 10 CTs for testing were fully annotated on every fifth axial slice with five different semantic body regions: abdominal cavity, bones, muscle, subcutaneous tissue, and thoracic cavity. Multi-resolution U-Net 3D neural networks were employed for segmenting these body regions, followed by subclassifying adipose tissue and muscle using known Hounsfield unit limits. RESULTS: The Sørensen Dice scores averaged over all semantic regions was 0.9553 and the intra-class correlation coefficients for subclassified tissues were above 0.99. CONCLUSIONS: Our results show that fully automated body composition analysis on routine CT imaging can provide stable biomarkers across the whole abdomen and not just on L3 slices, which is historically the reference location for analyzing body composition in the clinical routine. KEY POINTS: ⢠Our study enables fully automated body composition analysis on routine abdomen CT scans. ⢠The best segmentation models for semantic body region segmentation achieved an averaged Sørensen Dice score of 0.9553. ⢠Subclassified tissue volumes achieved intra-class correlation coefficients over 0.99.
Asunto(s)
Redes Neurales de la Computación , Semántica , Abdomen , Composición Corporal , Humanos , Procesamiento de Imagen Asistido por Computador , Tomografía Computarizada por Rayos XRESUMEN
PURPOSE: Detection and validation of the chest X-ray view position with use of convolutional neural networks to improve meta-information for data cleaning within a hospital data infrastructure. MATERIAL AND METHODS: Within this paper we developed a convolutional neural network which automatically detects the anteroposterior and posteroanterior view position of a chest radiograph. We trained two different network architectures (VGG variant and ResNet-34) with data published by the RSNA (26â684 radiographs, class distribution 46â% AP, 54â% PA) and validated these on a self-compiled dataset with data from the University Hospital Essen (4507, radiographs, class distribution 55â% PA, 45â% AP) labeled by a human reader. For visualization and better understanding of the network predictions, a Grad-CAM was generated for each network decision. The network results were evaluated based on the accuracy, the area under the curve (AUC), and the F1-score against the human reader labels. Also a final performance comparison between model predictions and DICOM labels was performed. RESULTS: The ensemble models reached accuracy and F1-scores greater than 95â%. The AUC reaches more than 0.99 for the ensemble models. The Grad-CAMs provide insight as to which anatomical structures contributed to a decision by the networks which are comparable with the ones a radiologist would use. Furthermore, the trained models were able to generalize over mislabeled examples, which was found by comparing the human reader labels to the predicted labels as well as the DICOM labels. CONCLUSION: The results show that certain incorrectly entered meta-information of radiological images can be effectively corrected by deep learning in order to increase data quality in clinical application as well as in research. KEY POINTS: · The predictions for both view positions are accurate with respect to external validation data.. · The networks based their decisions on anatomical structures and key points that were in-line with prior knowledge and human understanding.. · Final models were able to detect labeling errors within the test dataset.. CITATION FORMAT: · Hosch R, Kroll L, Nensa F etâal. Differentiation Between Anteroposterior and Posteroanterior Chest X-Ray View Position With Convolutional Neural Networks. Fortschr Röntgenstr 2021; 193: 168â-â176.