Positron emission tomography (PET) is currently considered the non-invasive reference standard for lymph node (N-)staging in lung cancer. However, not all patients can undergo this diagnostic ...procedure due to high costs, limited availability, and additional radiation exposure. The purpose of this study was to predict the PET result from traditional contrast-enhanced computed tomography (CT) and to test different feature extraction strategies.
In this study, 100 lung cancer patients underwent a contrast-enhanced
F-fluorodeoxyglucose (FDG) PET/CT scan between August 2012 and December 2019. We trained machine learning models to predict FDG uptake in the subsequent PET scan. Model inputs were composed of (i) traditional "hand-crafted" radiomics features from the segmented lymph nodes, (ii) deep features derived from a pretrained EfficientNet-CNN, and (iii) a hybrid approach combining (i) and (ii).
In total, 2734 lymph nodes 555 (20.3%) PET-positive from 100 patients 49% female; mean age 65, SD: 14 with lung cancer (60% adenocarcinoma, 21% plate epithelial carcinoma, 8% small-cell lung cancer) were included in this study. The area under the receiver operating characteristic curve (AUC) ranged from 0.79 to 0.87, and the scaled Brier score (SBS) ranged from 16 to 36%. The random forest model (iii) yielded the best results AUC 0.871 (0.865-0.878), SBS 35.8 (34.2-37.2) and had significantly higher model performance than both approaches alone (AUC:
< 0.001, z = 8.8 and z = 22.4; SBS:
< 0.001, z = 11.4 and z = 26.6, against (i) and (ii), respectively).
Both traditional radiomics features and transfer-learning deep radiomics features provide relevant and complementary information for non-invasive N-staging in lung cancer.
The bone marrow's iodine uptake in dual-energy CT (DECT) is elevated in malignant disease. We aimed to investigate the physiological range of bone marrow iodine uptake after intravenous contrast ...application, and examine its dependence on vBMD, iodine blood pool, patient age, and sex.
Retrospective analysis of oncological patients without evidence of metastatic disease. DECT examinations were performed on a spectral detector CT scanner in portal venous contrast phase. The thoracic and lumbar spine were segmented by a pre-trained neural network, obtaining volumetric iodine concentration data mg/ml. vBMD was assessed using a phantomless, CE-certified software mg/cm3. The iodine blood pool was estimated by ROI-based measurements in the great abdominal vessels. A multivariate regression model was fit with the dependent variable "median bone marrow iodine uptake". Standardized regression coefficients (β) were calculated to assess the impact of each covariate.
678 consecutive DECT exams of 189 individuals (93 female, age 61.4 ± 16.0 years) were evaluated. AI-based segmentation provided volumetric data of 97.9% of the included vertebrae (n=11,286). The 95
percentile of bone marrow iodine uptake, as a surrogate for the upper margin of the physiological distribution, ranged between 4.7-6.4 mg/ml. vBMD (p <0.001, mean β=0.50) and portal vein iodine blood pool (p <0.001, mean β=0.43) mediated the strongest impact. Based thereon, adjusted reference values were calculated.
The bone marrow iodine uptake demonstrates a distinct profile depending on vBMD, iodine blood pool, patient age, and sex. This study is the first to provide the adjusted reference values.
Background
In the management of cancer patients, determination of TNM status is essential for treatment decision-making and therefore closely linked to clinical outcome and survival. Here, we ...developed a tool for automatic three-dimensional (3D) localization and segmentation of cervical lymph nodes (LNs) on contrast-enhanced computed tomography (CECT) examinations.
Methods
In this IRB-approved retrospective single-center study, 187 CECT examinations of the head and neck region from patients with various primary diseases were collected from our local database, and 3656 LNs (19.5 ± 14.9 LNs/CECT, mean ± standard deviation) with a short-axis diameter (SAD) ≥ 5 mm were segmented manually by expert physicians. With these data, we trained an independent fully convolutional neural network based on 3D foveal patches. Testing was performed on 30 independent CECTs with 925 segmented LNs with an SAD ≥ 5 mm.
Results
In total, 4,581 LNs were segmented in 217 CECTs. The model achieved an average localization rate (LR),
i.e.
, percentage of localized LNs/CECT, of 78.0% in the validation dataset. In the test dataset, average LR was 81.1% with a mean Dice coefficient of 0.71. For enlarged LNs with a SAD ≥ 10 mm, LR was 96.2%. In the test dataset, the false-positive rate was 2.4 LNs/CECT.
Conclusions
Our trained AI model demonstrated a good overall performance in the consistent automatic localization and 3D segmentation of physiological and metastatic cervical LNs with a SAD ≥ 5 mm on CECTs. This could aid clinical localization and automatic 3D segmentation, which can benefit clinical care and radiomics research.
Relevance statement
Our AI model is a time-saving tool for 3D segmentation of cervical lymph nodes on contrast-enhanced CT scans and serves as a solid base for N staging in clinical practice and further radiomics research.
Key points
• Determination of N status in TNM staging is essential for therapy planning in oncology.
• Segmenting cervical lymph nodes manually is highly time-consuming in clinical practice.
• Our model provides a robust, automated 3D segmentation of cervical lymph nodes.
• It achieves a high accuracy for localization especially of enlarged lymph nodes.
• These segmentations should assist clinical care and radiomics research.
Graphical Abstract
Virtual non-calcium (VNCa) images from dual-energy computed tomography (DECT) have shown high potential to diagnose bone marrow disease of the spine, which is frequently disguised by dense trabecular ...bone on conventional CT. In this study, we aimed to define reference values for VNCa bone marrow images of the spine in a large-scale cohort of healthy individuals. DECT was performed after resection of a malignant skin tumor without evidence of metastatic disease. Image analysis was fully automated and did not require specific user interaction. The thoracolumbar spine was segmented by a pretrained convolutional neuronal network. Volumetric VNCa data of the spine's bone marrow space were processed using the maximum, medium, and low calcium suppression indices. Histograms of VNCa attenuation were created for each exam and suppression setting. We included 500 exams of 168 individuals (88 female, patient age 61.0 ± 15.9). A total of 8298 vertebrae were segmented. The attenuation histograms' overlap of two consecutive exams, as a measure for intraindividual consistency, yielded a median of 0.93 (IQR: 0.88-0.96). As our main result, we provide the age- and sex-specific bone marrow attenuation profiles of a large-scale cohort of individuals with healthy trabecular bone structure as a reference for future studies. We conclude that artificial-intelligence-supported, fully automated volumetric assessment is an intraindividually robust method to image the spine's bone marrow using VNCa data from DECT.
Purpose
Malignant intracranial germ cell tumors (GCTs) are rare diseases in Western countries. They arise in midline structures and diagnosis is often delayed. We evaluated imaging characteristics ...and early tumor signs of suprasellar and bifocal GCT on MRI.
Methods
Patients with the diagnosis of a germinoma or non-germinomatous GCT (NGGCT) who received non-contrast sagittal T1WI on MRI pre-therapy were included. Loss of the posterior pituitary bright spot (PPBS), the expansion and size of the tumor, and the expansion and infiltration of surrounding structures were evaluated. Group comparison for histologies and localizations was performed.
Results
A total of 102 GCT patients (median age at diagnosis 12.3 years, range 4.4–33.8; 57 males; 67 in suprasellar localization) were enrolled in the study. In the suprasellar cohort, NGGCTs (
n
= 20) were noticeably larger than germinomas (
n
= 47;
p
< .001). Each tumor showed involvement of the posterior lobe or pituitary stalk. A PPBS loss (total
n
= 98) was observed for each localization and entity in more than 90% and was related to diabetes insipidus. Osseous infiltration was observed exclusively in suprasellar GCT (significantly more frequent in NGGCT;
p
= .004). Time between the first MRI and therapy start was significantly longer in the suprasellar cohort (
p
= .005), with an even greater delay in germinoma compared to NGGCT (
p
= .002). The longest interval to treatment had circumscribed suprasellar germinomas (median 312 days).
Conclusion
A loss of the PPBS is a hint of tumor origin revealing small tumors in the neurohypophysis. Using this sign in children with diabetes insipidus avoids a delay in diagnosis.
Objectives
Differentiation between COVID-19 and community-acquired pneumonia (CAP) in computed tomography (CT) is a task that can be performed by human radiologists and artificial intelligence (AI). ...The present study aims to (1) develop an AI algorithm for differentiating COVID-19 from CAP and (2) evaluate its performance. (3) Evaluate the benefit of using the AI result as assistance for radiological diagnosis and the impact on relevant parameters such as accuracy of the diagnosis, diagnostic time, and confidence.
Methods
We included
n
= 1591 multicenter, multivendor chest CT scans and divided them into AI training and validation datasets to develop an AI algorithm (
n
= 991 CT scans;
n
= 462 COVID-19, and
n
= 529 CAP) from three centers in China. An independent Chinese and German test dataset of
n
= 600 CT scans from six centers (COVID-19 / CAP;
n
= 300 each) was used to test the performance of eight blinded radiologists and the AI algorithm. A subtest dataset (180 CT scans;
n
= 90 each) was used to evaluate the radiologists’ performance without and with AI assistance to quantify changes in diagnostic accuracy, reporting time, and diagnostic confidence.
Results
The diagnostic accuracy of the AI algorithm in the Chinese-German test dataset was 76.5%. Without AI assistance, the eight radiologists’ diagnostic accuracy was 79.1% and increased with AI assistance to 81.5%, going along with significantly shorter decision times and higher confidence scores.
Conclusion
This large multicenter study demonstrates that AI assistance in CT-based differentiation of COVID-19 and CAP increases radiological performance with higher accuracy and specificity, faster diagnostic time, and improved diagnostic confidence.
Key Points
• AI can help radiologists to get higher diagnostic accuracy, make faster decisions, and improve diagnostic confidence.
• The China-German multicenter study demonstrates the advantages of a human-machine interaction using AI in clinical radiology for diagnostic differentiation between COVID-19 and CAP in CT scans.
The extent of lung involvement in coronavirus disease 2019 (COVID-19) pneumonia, quantified on computed tomography (CT), is an established biomarker for prognosis and guides clinical decision-making. ...The clinical standard is semi-quantitative scoring of lung involvement by an experienced reader. We aim to compare the performance of automated deep-learning- and threshold-based methods to the manual semi-quantitative lung scoring. Further, we aim to investigate an optimal threshold for quantification of involved lung in COVID pneumonia chest CT, using a multi-center dataset.
In total 250 patients were included, 50 consecutive patients with RT-PCR confirmed COVID-19 from our local institutional database, and another 200 patients from four international datasets (n=50 each). Lung involvement was scored semi-quantitatively by three experienced radiologists according to the established chest CT score (CCS) ranging from 0-25. Inter-rater reliability was reported by the intraclass correlation coefficient (ICC). Deep-learning-based segmentation of ground-glass and consolidation was obtained by CT Pulmo Auto Results prototype plugin on IntelliSpace Discovery (Philips Healthcare, The Netherlands). Threshold-based segmentation of involved lung was implemented using an open-source tool for whole-lung segmentation under the presence of severe pathologies (R231CovidWeb, Hofmanninger
, 2020) and consecutive quantitative assessment of lung attenuation. The best threshold was investigated by training and testing a linear regression of deep-learning and threshold-based results in a five-fold cross validation strategy.
Median CCS among 250 evaluated patients was 10 6-15. Inter-rater reliability of the CCS was excellent ICC 0.97 (0.97-0.98). Best attenuation threshold for identification of involved lung was -522 HU. While the relationship of deep-learning- and threshold-based quantification was linear and strong (r
=0.84), both automated quantification methods translated to the semi-quantitative CCS in a non-linear fashion, with an increasing slope towards higher CCS (r
= 0.80, r
=0.63).
The manual semi-quantitative CCS underestimates the extent of COVID pneumonia in higher score ranges, which limits its clinical usefulness in cases of severe disease. Clinical implementation of fully automated methods, such as deep-learning or threshold-based approaches (best threshold in our multi-center dataset: -522 HU), might save time of trained personnel, abolish inter-reader variability, and allow for truly quantitative, linear assessment of COVID lung involvement.
Diagnosing a coronavirus disease 2019 (COVID-19) infection with high specificity in chest computed tomography (CT) imaging is considered possible due to distinctive imaging features of COVID-19 ...pneumonia. Since other viral non-COVID pneumonia show mostly a different distribution pattern, it is reasonable to assume that the patterns observed caused by the novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) are a consequence of its genetically encoded molecular properties when interacting with the respiratory tissue. As more mutations of the initial SARS-CoV-2 wild-type with varying aggressiveness have been detected in the course of 2021, it became obvious that its genome is in a state of transformation and therefore a potential modification of the specific morphological appearance in CT may occur. The aim of this study was to quantitatively analyze the morphological differences of the SARS-CoV-2-B.1.1.7 mutation and wildtype variant in CT scans of the thorax.
We analyzed a dataset of 140 patients, which was divided into pneumonias caused by n=40 wildtype variants, n=40 B.1.1.7 variants, n=20 bacterial pneumonias, n=20 viral (non-COVID) pneumonias, and a test group of n=20 unremarkable CT examinations of the thorax. Semiautomated 3D segmentation of the lung tissue was performed for quantification of lung pathologies. The extent, ratio, and specific distribution of inflammatory affected lung tissue in each group were compared in a multivariate group analysis.
Lung segmentation revealed significant difference between the extent of ground glass opacities (GGO) or consolidation comparing SARS-CoV-2 wild-type and B.1.1.7 variant. Wildtype and B.1.1.7 variant showed both a symmetric distribution pattern of stage-dependent GGO and consolidation within matched COVID-19 stages. Viral non-COVID pneumonias had significantly fewer consolidations than the bacterial, but also than the COVID-19 B.1.1.7 variant groups.
CT based segmentation showed no significant difference between the morphological appearance of the COVID-19 wild-type variant and the SARS-CoV-2 B.1.1.7 mutation. However, our approach allowed a semiautomatic quantification of bacterial and viral lung pathologies. Quantitative CT image analyses, such as the one presented, appear to be an important component of pandemic preparedness considering an organism with ongoing genetic change, to describe a potential arising change in CT morphological appearance of possible new upcoming COVID-19 variants of concern.
