Image quality of abdominal photon-counting CT with reduced contrast media dose: Evaluation of reduced contrast media protocols during the COVID19 pandemic supply shortage.

Rationale and objectives: Aim of this study was to assess the impact of contrast media dose (CMD) reduction on diagnostic quality of photon-counting detector CT (PCD-CT) and energy-integrating detector CT (EID-CT). Methods: CT scans of the abdominal region with differing CMD acquired in portal venous phase on a PCD-CT were included and compared to EID-CT scans. Diagnostic quality and contrast intensity were rated. Additionally, readers had to assign the scans to reduced or regular CMD. Regions-of-interest (ROIs) were placed in defined segments of portal vein, inferior vena cava, liver, spleen, kidneys, abdominal aorta and muscular tissue. Signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) were calculated. Results: Overall 158 CT scans performed on a PCD-CT and 68 examinations on an EID-CT were analyzed. Overall diagnostic quality showed no significant differences for PCD-CT with standard CMD which scored a median 5 (IQR:5-5) and PCD-CT with 70% CMD scoring 5 (4-5). (For PCD-CT, 71.69% of the examinations with reduced CMD were assigned to regular CMD by the readers, for EID-CT 9.09%. Averaged for all measurements SNR for 50% CMD was reduced by 19% in PCD-CT (EID-CT 34%) and CNR by 48% (EID-CT 56%). Virtual monoenergetic images (VMI)50keV for PCD-CT images acquired with 50% CMD showed an increase in SNR by 72% and CNR by 153%. Conclusions: Diagnostic interpretability of PCD-CT examinations with reduction of up to 50% CMD is maintained. PCD-CT deducted scans especially with 70% CMD were often not recognized as CMD reduced scans. Compared to EID-CT less decline in SNR and CNR is observed for CMD reduced PCD-CT images. Employing VMI50keV for CMD-reduced PCD-CT images compensated for the effects.

Determining individual glomerular proteinuria and periglomerular infiltration in a cleared murine kidney by a 3-dimensional fast marching algorithm.

Three-dimensional (3D) imaging has advanced basic research and clinical medicine. However, limited resolution and imperfections of real-world 3D image material often preclude algorithmic image analysis. Here, we present a methodologic framework for such imaging and analysis for functional and spatial relations in experimental nephritis. First, optical tissue-clearing protocols were optimized to preserve fluorescence signals for light sheet fluorescence microscopy and compensated attenuation effects using adjustable 3D correction fields. Next, we adapted the fast marching algorithm to conduct backtracking in 3D environments and developed a tool to determine local concentrations of extractable objects. As a proof-of-concept application, we used this framework to determine in a glomerulonephritis model the individual proteinuria and periglomerular immune cell infiltration for all glomeruli of half a mouse kidney. A correlation between these parameters surprisingly did not support the intuitional assumption that the most inflamed glomeruli are the most proteinuric. Instead, the spatial density of adjacent glomeruli positively correlated with the proteinuria of a given glomerulus. Because proteinuric glomeruli appear clustered, this suggests that the exact location of a kidney biopsy may affect the observed severity of glomerular damage. Thus, our algorithmic pipeline described here allows analysis of various parameters of various organs composed of functional subunits, such as the kidney, and can theoretically be adapted to processing other image modalities.

Use of virtual monoenergetic images for reduction of extensive dental implant associated artifacts in photon-counting detector CT.

Aim of this study was to assess the impact of virtual monoenergetic images (VMI) on dental implant artifacts in photon-counting detector computed tomography (PCD-CT) compared to standard reconstructed polychromatic images (PI). 30 scans with extensive (≥ 5 dental implants) dental implant-associated artifacts were retrospectively analyzed. Scans were acquired during clinical routine on a PCD-CT. VMI were reconstructed for 100-190 keV (10 keV steps) and compared to PI. Artifact extent and assessment of adjacent soft tissue were rated using a 5-point Likert grading scale for qualitative assessment. Quantitative assessment was performed using ROIs in most pronounced hypodense and hyperdense artifacts, artifact-impaired soft tissue, artifact-free fat and muscle tissue. A corrected attenuation was calculated as difference between artifact-impaired tissue and tissue without artifacts. Qualitative assessment of soft palate and cheeks improved for all VMI compared to PI (Median PI: 1 (Range: 1-3) and 1 (1-3); e.g. VMI130 keV 2 (1-5); p < 0.0001 and 2 (1-4); p < 0.0001). In quantitative assessment, VMI130 keV showed best results with a corrected attenuation closest to 0 (PI: 30.48 ± 98.16; VMI130 keV: - 0.55 ± 73.38; p = 0.0026). Overall, photon-counting deducted VMI reduce the extent of dental implant-associated artifacts. VMI of 130 keV showed best results and are recommended to support head and neck CT scans.

Deep learning-based assessment of CT markers of sarcopenia and myosteatosis for outcome assessment in patients with advanced pancreatic cancer after high-intensity focused ultrasound treatment.

OBJECTIVES: To evaluate the prognostic value of CT-based markers of sarcopenia and myosteatosis in comparison to the Eastern Cooperative Oncology Group (ECOG) score for survival of patients with advanced pancreatic cancer treated with high-intensity focused ultrasound (HIFU). MATERIALS AND METHODS: For 142 retrospective patients, the skeletal muscle index (SMI), skeletal muscle radiodensity (SMRD), fatty muscle fraction (FMF), and intermuscular fat fraction (IMFF) were determined on superior mesenteric artery level in pre-interventional CT. Each marker was tested for associations with sex, age, body mass index (BMI), and ECOG. The prognostic value of the markers was examined in Kaplan-Meier analyses with the log-rank test and in uni- and multivariable Cox proportional hazards (CPH) models. RESULTS: The following significant associations were observed: Male patients had higher BMI and SMI. Patients with lower ECOG had lower BMI and SMI. Patients with BMI lower than 21.8 kg/m2 (median) also showed lower SMI and IMFF. Patients younger than 63.3 years (median) were found to have higher SMRD, lower FMF, and lower IMFF. In the Kaplan-Meier analysis, significantly lower survival times were observed in patients with higher ECOG or lower SMI. Increased patient risk was observed for higher ECOG, lower BMI, and lower SMI in univariable CPH analyses for 1-, 2-, and 3-year survival. Multivariable CPH analysis for 1-year survival revealed increased patient risk for higher ECOG, lower SMI, lower IMFF, and higher FMF. In multivariable analysis for 2- and 3-year survival, only ECOG and FMF remained significant. CONCLUSION: CT-based markers of sarcopenia and myosteatosis show a prognostic value for assessment of survival in advanced pancreatic cancer patients undergoing HIFU therapy. CLINICAL RELEVANCE STATEMENT: The results indicate a greater role of myosteatosis for additional risk assessment beyond clinical scores, as only FMF was associated with long-term survival in multivariable CPH analyses along ECOG and also showed independence to ECOG in group analysis. KEY POINTS: • This study investigates the prognostic value of CT-based markers of sarcopenia and myosteatosis for patients with pancreatic cancer treated with high-intensity focused ultrasound. • Markers for sarcopenia and myosteatosis showed a prognostic value besides clinical assessment of the physical status by the Eastern Cooperative Oncology Group score. In contrast to muscle size measurements, the myosteatosis marker fatty muscle fraction demonstrated independence to the clinical score. • The results indicate that myosteatosis might play a greater role for additional patient risk assessments beyond clinical assessments of physical status.

Deeplasia: deep learning for bone age assessment validated on skeletal dysplasias.

BACKGROUND: Skeletal dysplasias collectively affect a large number of patients worldwide. Most of these disorders cause growth anomalies. Hence, evaluating skeletal maturity via the determination of bone age (BA) is a useful tool. Moreover, consecutive BA measurements are crucial for monitoring the growth of patients with such disorders, especially for timing hormonal treatment or orthopedic interventions. However, manual BA assessment is time-consuming and suffers from high intra- and inter-rater variability. This is further exacerbated by genetic disorders causing severe skeletal malformations. While numerous approaches to automate BA assessment have been proposed, few are validated for BA assessment on children with skeletal dysplasias. OBJECTIVE: We present Deeplasia, an open-source prior-free deep-learning approach designed for BA assessment specifically validated on patients with skeletal dysplasias. MATERIALS AND METHODS: We trained multiple convolutional neural network models under various conditions and selected three to build a precise model ensemble. We utilized the public BA dataset from the Radiological Society of North America (RSNA) consisting of training, validation, and test subsets containing 12,611, 1,425, and 200 hand and wrist radiographs, respectively. For testing the performance of our model ensemble on dysplastic hands, we retrospectively collected 568 radiographs from 189 patients with molecularly confirmed diagnoses of seven different genetic bone disorders including achondroplasia and hypochondroplasia. A subset of the dysplastic cohort (149 images) was used to estimate the test-retest precision of our model ensemble on longitudinal data. RESULTS: The mean absolute difference of Deeplasia for the RSNA test set (based on the average of six different reference ratings) and dysplastic set (based on the average of two different reference ratings) were 3.87 and 5.84 months, respectively. The test-retest precision of Deeplasia on longitudinal data (2.74 months) is estimated to be similar to a human expert. CONCLUSION: We demonstrated that Deeplasia is competent in assessing the age and monitoring the development of both normal and dysplastic bones.

Development of image-based decision support systems utilizing information extracted from radiological free-text report databases with text-based transformers.

OBJECTIVES: To investigate the potential and limitations of utilizing transformer-based report annotation for on-site development of image-based diagnostic decision support systems (DDSS). METHODS: The study included 88,353 chest X-rays from 19,581 intensive care unit (ICU) patients. To label the presence of six typical findings in 17,041 images, the corresponding free-text reports of the attending radiologists were assessed by medical research assistants ("gold labels"). Automatically generated "silver" labels were extracted for all reports by transformer models trained on gold labels. To investigate the benefit of such silver labels, the image-based models were trained using three approaches: with gold labels only (MG), with silver labels first, then with gold labels (MS/G), and with silver and gold labels together (MS+G). To investigate the influence of invested annotation effort, the experiments were repeated with different numbers (N) of gold-annotated reports for training the transformer and image-based models and tested on 2099 gold-annotated images. Significant differences in macro-averaged area under the receiver operating characteristic curve (AUC) were assessed by non-overlapping 95% confidence intervals. RESULTS: Utilizing transformer-based silver labels showed significantly higher macro-averaged AUC than training solely with gold labels (N = 1000: MG 67.8 [66.0-69.6], MS/G 77.9 [76.2-79.6]; N = 14,580: MG 74.5 [72.8-76.2], MS/G 80.9 [79.4-82.4]). Training with silver and gold labels together was beneficial using only 500 gold labels (MS+G 76.4 [74.7-78.0], MS/G 75.3 [73.5-77.0]). CONCLUSIONS: Transformer-based annotation has potential for unlocking free-text report databases for the development of image-based DDSS. However, on-site development of image-based DDSS could benefit from more sophisticated annotation pipelines including further information than a single radiological report. CLINICAL RELEVANCE STATEMENT: Leveraging clinical databases for on-site development of artificial intelligence (AI)-based diagnostic decision support systems by text-based transformers could promote the application of AI in clinical practice by circumventing highly regulated data exchanges with third parties. KEY POINTS: • The amount of data from a database that can be used to develop AI-assisted diagnostic decision systems is often limited by the need for time-consuming identification of pathologies by radiologists. • The transformer-based structuring of free-text radiological reports shows potential to unlock corresponding image databases for on-site development of image-based diagnostic decision support systems. • However, the quality of image annotations generated solely on the content of a single radiology report may be limited by potential inaccuracies and incompleteness of this report.

Post-interventional infectious complications in percutaneous transabdominal lymphatic interventions: an observational study.

The purpose of this retrospective study was to evaluate the occurrence of infectious complications and inflammatory reactions after transabdominal lymphatic-interventions. 63 lymphatic-interventions were performed in 60 patients (male/female: 35/25; mean age 56 [9-85] years) [chylothorax n = 48, chylous ascites n = 7, combined chylothorax/chylous ascites n = 5]. Post-interventional clinical course and laboratory findings were analyzed in the whole cohort as well as subgroups without (group A; n = 35) and with peri-interventional antibiotics (group B; n = 25) (pneumonia n = 16, drainage-catheter inflammation n = 5, colitis n = 1, cystitis n = 1, transcolonic-access n = 2). No septic complications associated with the intervention occurred. Leucocytes increased significantly, peaking on post-interventional day-1 (8.6 ± 3.9 × 106 cells/mL vs. 9.8 ± 4.7 × 106 cells/mL; p = 0.009) and decreased thereafter (day-10: 7.3 ± 2.7 × 106 cells/mL, p = 0.005). CRP-values were pathological in 89.5% of patients already at baseline (40.1 ± 63.9 mg/L) and increased significant on day-3 (77.0 ± 78.8 mg/L, p < 0.001). Values decreased thereafter (day-15: 25.3 ± 34.4 mg/L, p = 0.04). In subgroup B, 13/25 patients had febrile episodes post-interventionally (pneumonia n = 11, cystitis n = 1, drainage-catheter inflammation n = 1). One patient developed biliary peritonitis despite continued antibiotics and underwent cholecystectomy. Baseline leucocytes and CRP-levels were higher in group B than A, but with comparable post-interventional profiles. Clinically relevant infectious complications associated with transabdominal lymphatic-interventions are rare irrespective of peri-interventional antibiotic use. Post-interventional elevation of leucocytes and CRP are observed with normalization over 10-15 days.

Direct deep learning-based survival prediction from pre-interventional CT prior to transcatheter aortic valve replacement.

PURPOSE: To investigate survival prediction in patients undergoing transcatheter aortic valve replacement (TAVR) using deep learning (DL) methods applied directly to pre-interventional CT images and to compare performance with survival models based on scalar markers of body composition. METHOD: This retrospective single-center study included 760 patients undergoing TAVR (mean age 81 ± 6 years; 389 female). As a baseline, a Cox proportional hazards model (CPHM) was trained to predict survival on sex, age, and the CT body composition markers fatty muscle fraction (FMF), skeletal muscle radiodensity (SMRD), and skeletal muscle area (SMA) derived from paraspinal muscle segmentation of a single slice at L3/L4 level. The convolutional neural network (CNN) encoder of the DL model for survival prediction was pre-trained in an autoencoder setting with and without a focus on paraspinal muscles. Finally, a combination of DL and CPHM was evaluated. Performance was assessed by C-index and area under the receiver operating curve (AUC) for 1-year and 2-year survival. All methods were trained with five-fold cross-validation and were evaluated on 152 hold-out test cases. RESULTS: The CNN for direct image-based survival prediction, pre-trained in a focussed autoencoder scenario, outperformed the baseline CPHM (CPHM: C-index = 0.608, 1Y-AUC = 0.606, 2Y-AUC = 0.594 vs. DL: C-index = 0.645, 1Y-AUC = 0.687, 2Y-AUC = 0.692). Combining DL and CPHM led to further improvement (C-index = 0.668, 1Y-AUC = 0.713, 2Y-AUC = 0.696). CONCLUSIONS: Direct DL-based survival prediction shows potential to improve image feature extraction compared to segmentation-based scalar markers of body composition for risk assessment in TAVR patients.

Long-Term Clinical Outcome of Abdomino-Thoracic Lymphatic Interventions of Traumatic and Non-Traumatic Lymphatic Leakage in Adults.

The aim of this study was to retrospectively evaluate the long-term results of lymphatic interventions in adults with abdomino-thoracic lymphatic pathologies. Management of abdomino-thoracic chylous effusions in adults undergoing X-ray-lymphangiography with or without lymph-vessel embolization (LVE) from 2010-2018 was reviewed. Patients underwent lymphangiography alone when imaging showed normal findings or lymphatic obstruction without leakage or reflux; otherwise, LVE was performed (leakage, reflux, obstruction with leakage or reflux, lymphatic masses). Technical and clinical success, complications, and long-term outcomes were assessed. 78 patients (47 male, median age 56.3 years) were treated for chylous effusions (60.3% traumatic, 39.7% non-traumatic). Lymphangiography showed leakage (48.7%), reflux (14.1%), obstruction (28.2%), lymphatic masses (5.1%), and normal findings (3.8%). Embolization was performed in 49/78 (62.8%) cases. Overall, treatment was clinically successful in 74.4% (mean follow-up of 28 months), with significant differences between LVE and lymphangiography (91.8% vs. 44.8%; p < 0.001), traumatic and non-traumatic etiologies (89.4% vs. 51.6%; p < 0.001), and leakage locations (p = 0.003). The clinical success of LVE did not differ between leakage etiologies or locations. Complications occurred in 5 patients (2/5 needed treatment). Patients survived significantly longer after successful treatment (2679 vs. 927 days; p = 0.044) and without malignancy (3214 vs. 1550 days; p = 0.043). Lymphatic interventions are safe and effective. LVE should be attempted whenever feasible, as success is high (>90%). Successful intervention has a positive effect on patient survival.

Combining iterative metal artifact reduction and virtual monoenergetic images severely reduces hip prosthesis-associated artifacts in photon-counting detector CT.

Aim of this study was to assess the impact of virtual monoenergetic images (VMI) in combination and comparison with iterative metal artifact reduction (IMAR) on hip prosthesis-associated artifacts in photon-counting detector CT (PCD-CT). Retrospectively, 33 scans with hip prosthesis-associated artifacts acquired during clinical routine on a PCD-CT between 08/2022 and 09/2022 were analyzed. VMI were reconstructed for 100-190 keV with and without IMAR, and compared to polychromatic images. Qualitatively, artifact extent and assessment of adjacent soft tissue were rated by two radiologists using 5-point Likert items. Quantitative assessment was performed measuring attenuation and standard deviation in most pronounced hypodense and hyperdense artifacts, artifact-impaired bone, muscle, vessels, bladder and artifact-free corresponding tissue. To quantify artifacts, an adjusted attenuation was calculated as the difference between artifact-impaired tissue and corresponding tissue without artifacts. Qualitative assessment improved for all investigated image reconstructions compared to polychromatic images (PI). VMI100keV in combination with IMAR achieved best results (e.g. diagnostic quality of the bladder: median PI: 1.5 (range 1-4); VMI100keV+IMAR: 5 (3-5); p < 0.0001). In quantitative assessment VMI100keV with IMAR provided best artifact reduction with an adjusted attenuation closest to 0 (e.g. bone: PI: 302.78; VMI100keV+IMAR: 51.18; p < 0.0001). The combination of VMI and IMAR significantly reduces hip prosthesis-associated artifacts in PCD-CT and improves the diagnostic quality of surrounding tissue.

Computed tomography derived cervical fat-free muscle fraction as an imaging-based outcome marker in patients with acute ischemic stroke: a pilot study.

BACKGROUND: Outcome assessment in stroke patients is essential for evidence-based stroke care planning. Computed tomography (CT) is the mainstay of diagnosis in acute stroke. This study aimed to investigate whether CT-derived cervical fat-free muscle fraction (FFMF) as a biomarker of muscle quality is associated with outcome parameters after acute ischemic stroke. METHODS: In this retrospective study, 66 patients (mean age: 76 ± 13 years, 30 female) with acute ischemic stroke in the anterior circulation who underwent CT, including CT-angiography, and endovascular mechanical thrombectomy of the middle cerebral artery between August 2016 and January 2020 were identified. Based on densitometric thresholds, cervical paraspinal muscles covered on CT-angiography were separated into areas of fatty and lean muscle and FFMF was calculated. The study cohort was binarized based on median FFMF (cutoff value: < 71.6%) to compare clinical variables and outcome data between two groups. Unpaired t test and Mann-Whitney U test were used for statistical analysis. RESULTS: National Institute of Health Stroke Scale (NIHSS) (12.2 ± 4.4 vs. 13.6 ± 4.5, P = 0.297) and modified Rankin scale (mRS) (4.3 ± 0.9 vs. 4.4 ± 0.9, P = 0.475) at admission, and pre-stroke mRS (1 ± 1.3 vs. 0.9 ± 1.4, P = 0.489) were similar between groups with high and low FFMF. NIHSS and mRS at discharge were significantly better in patients with high FFMF compared to patients with low FFMF (NIHSS: 4.5 ± 4.4 vs. 9.5 ± 6.7; P = 0.004 and mRS: 2.9 ± 2.1 vs.3.9 ± 1.8; P = 0.049). 90-day mRS was significantly better in patients with high FFMF compared to patients with low FFMF (3.3 ± 2.2 vs. 4.3 ± 1.9, P = 0.045). CONCLUSION: Cervical FFMF obtained from routine clinical CT might be a new imaging-based muscle quality biomarker for outcome prediction in stroke patients.

Linear multi-scale modeling of diffusion MRI data: A framework for characterization of oriented structures across length scales.

Diffusion-weighted magnetic resonance imaging (DW-MRI) has evolved to provide increasingly sophisticated investigations of the human brain's structural connectome in vivo. Restriction spectrum imaging (RSI) is a method that reconstructs the orientation distribution of diffusion within tissues over a range of length scales. In its original formulation, RSI represented the signal as consisting of a spectrum of Gaussian diffusion response functions. Recent technological advances have enabled the use of ultra-high b-values on human MRI scanners, providing higher sensitivity to intracellular water diffusion in the living human brain. To capture the complex diffusion time dependence of the signal within restricted water compartments, we expand upon the RSI approach to represent restricted water compartments with non-Gaussian response functions, in an extended analysis framework called linear multi-scale modeling (LMM). The LMM approach is designed to resolve length scale and orientation-specific information with greater specificity to tissue microstructure in the restricted and hindered compartments, while retaining the advantages of the RSI approach in its implementation as a linear inverse problem. Using multi-shell, multi-diffusion time DW-MRI data acquired with a state-of-the-art 3 T MRI scanner equipped with 300 mT/m gradients, we demonstrate the ability of the LMM approach to distinguish different anatomical structures in the human brain and the potential to advance mapping of the human connectome through joint estimation of the fiber orientation distributions and compartment size characteristics.

Association between single-slice and whole heart measurements of epicardial and pericardial fat in cardiac MRI.

BACKGROUND: Epicardial (ECF) and pericardial fat (PCF) are important prognostic markers for various cardiac diseases. However, volumetry of the fat compartments is time-consuming. PURPOSE: To investigate whether total volume of ECF and PCF can be estimated by axial single-slice measurements and in a four-chamber view. MATERIAL AND METHODS: A total of 113 individuals (79 patients and 34 healthy) were included in this retrospective magnetic resonance imaging (MRI) study. The total volume of ECF and PCF was determined using a 3D-Dixon sequence. Additionally, the area of ECF and PCF was obtained in single axial layers at five anatomical landmarks (left coronary artery, right coronary artery, right pulmonary artery, mitral valve, coronary sinus) of the Dixon sequence and in a four-chamber view of a standard cine sequence. Pearson's correlation coefficient was calculated between the total volume and each single-slice measurement. RESULTS: Axial single-slice measurements of ECF and PCF correlated strongly with the total fat volumes at all landmarks (ECF: r = 0.85-0.94, P < 0.001; PCF: r = 0.89-0.94, P < 0.001). The best correlation was found at the level of the left coronary artery for ECF and PCF (r = 0.94, P < 0.001). Correlation between single-slice measurement in the four-chamber view and the total ECF and PCF volume was lower (r = 0.75 and r = 0.8, respectively, P < 0.001). CONCLUSION: Single-slice measurements allow an estimation of ECF and PCF volume. This time-efficient analysis allows studies of larger patient cohorts and the opportunistic determination of ECF/PCF from routine examinations.

Deep learning for standardized, MRI-based quantification of subcutaneous and subfascial tissue volume for patients with lipedema and lymphedema.

OBJECTIVES: To contribute to a more in-depth assessment of shape, volume, and asymmetry of the lower extremities in patients with lipedema or lymphedema utilizing volume information from MR imaging. METHODS: A deep learning (DL) pipeline was developed including (i) localization of anatomical landmarks (femoral heads, symphysis, knees, ankles) and (ii) quality-assured tissue segmentation to enable standardized quantification of subcutaneous (SCT) and subfascial tissue (SFT) volumes. The retrospectively derived dataset for method development consisted of 45 patients (42 female, 44.2 ± 14.8 years) who underwent clinical 3D DIXON MR-lymphangiography examinations of the lower extremities. Five-fold cross-validated training was performed on 16,573 axial slices from 40 patients and testing on 2187 axial slices from 5 patients. For landmark detection, two EfficientNet-B1 convolutional neural networks (CNNs) were applied in an ensemble. One determines the relative foot-head position of each axial slice with respect to the landmarks by regression, the other identifies all landmarks in coronal reconstructed slices using keypoint detection. After landmark detection, segmentation of SCT and SFT was performed on axial slices employing a U-Net architecture with EfficientNet-B1 as encoder. Finally, the determined landmarks were used for standardized analysis and visualization of tissue volume, distribution, and symmetry, independent of leg length, slice thickness, and patient position. RESULTS: Excellent test results were observed for landmark detection (z-deviation = 4.5 ± 3.1 mm) and segmentation (Dice score: SCT = 0.989 ± 0.004, SFT = 0.994 ± 0.002). CONCLUSIONS: The proposed DL pipeline allows for standardized analysis of tissue volume and distribution and may assist in diagnosis of lipedema and lymphedema or monitoring of conservative and surgical treatments. KEY POINTS: • Efficient use of volume information that MRI inherently provides can be extracted automatically by deep learning and enables in-depth assessment of tissue volumes in lipedema and lymphedema. • The deep learning pipeline consisting of body part regression, keypoint detection, and quality-assured tissue segmentation provides detailed information about the volume, distribution, and asymmetry of lower extremity tissues, independent of leg length, slice thickness, and patient position.

Influence of Image Processing on Radiomic Features From Magnetic Resonance Imaging.

OBJECTIVE: Before implementing radiomics in routine clinical practice, comprehensive knowledge about the repeatability and reproducibility of radiomic features is required. The aim of this study was to systematically investigate the influence of image processing parameters on radiomic features from magnetic resonance imaging (MRI) in terms of feature values as well as test-retest repeatability. MATERIALS AND METHODS: Utilizing a phantom consisting of 4 onions, 4 limes, 4 kiwifruits, and 4 apples, we acquired a test-retest dataset featuring 3 of the most commonly used MRI sequences on a 3 T scanner, namely, a T1-weighted, a T2-weighted, and a fluid-attenuated inversion recovery sequence, each at high and low resolution. After semiautomatic image segmentation, image processing with systematic variation of image processing parameters was performed, including spatial resampling, intensity discretization, and intensity rescaling. For each respective image processing setting, a total of 45 radiomic features were extracted, corresponding to the following 7 matrices/feature classes: conventional indices, histogram matrix, shape matrix, gray-level zone length matrix, gray-level run length matrix, neighboring gray-level dependence matrix, and gray-level cooccurrence matrix. Systematic differences of individual features between different resampling steps were assessed using 1-way analysis of variance with Tukey-type post hoc comparisons to adjust for multiple testing. Test-retest repeatability of radiomic features was measured using the concordance correlation coefficient, dynamic range, and intraclass correlation coefficient. RESULTS: Image processing influenced radiological feature values. Regardless of the acquired sequence and feature class, significant differences ( P < 0.05) in feature values were found when the size of the resampled voxels was too large, that is, bigger than 3 mm. Almost all higher-order features depended strongly on intensity discretization. The effects of intensity rescaling were negligible except for some features derived from T1-weighted sequences. For all sequences, the percentage of repeatable features (concordance correlation coefficient and dynamic range ≥ 0.9) varied considerably depending on the image processing settings. The optimal image processing setting to achieve the highest percentage of stable features varied per sequence. Irrespective of image processing, the fluid-attenuated inversion recovery sequence in high-resolution overall yielded the highest number of stable features in comparison with the other sequences (89% vs 64%-78% for the respective optimal image processing settings). Across all sequences, the most repeatable features were generally obtained for a spatial resampling close to the originally acquired voxel size and an intensity discretization to at least 32 bins. CONCLUSION: Variation of image processing parameters has a significant impact on the values of radiomic features as well as their repeatability. Furthermore, the optimal image processing parameters differ for each MRI sequence. Therefore, it is recommended that these processing parameters be determined in corresponding test-retest scans before clinical application. Extensive repeatability, reproducibility, and validation studies as well as standardization are required before quantitative image analysis and radiomics can be reliably translated into routine clinical care.

Percutaneous transhepatic biliary drainage: a retrospective single-center study of 372 patients.

BACKGROUND: Complication rates in percutaneous transhepatic biliary drainage (PTBD) are non-uniform and vary considerably. In addition, the impact of peri-procedural risk factors is under-investigated. PURPOSE: To compare success and complication rates of PTBD in patients with and without accompanying technical risk factors. MATERIAL AND METHODS: A single-center retrospective study was conducted from January 2004 to December 2016. Patients receiving PTBD due to biliary obstruction or biliary leakage were included. Technical risk factors (non-distended bile ducts, ascites, obesity, anasarca, non-compliance) were assessed. Complications were classified according to the Society of Interventional Radiology. RESULTS: In total, 372 patients were included (57.3% men, 42.7% women; mean age = 66 years). Overall, 466 PTBDs were performed. Of the patients, 70.1% presented with malignancy and biliary obstruction; 26.8% had benign biliary obstruction; 3.1% had biliary leakage. Technical risk factors were reported in 57 (15.3%) patients. Overall technical success of initial PTBD was 98.7%, primary technical success was 97.9%. In patients with non-dilatated bile ducts, primary technical success was 68.2%. Overall complication rate was 15.0% (8.1% major complications, 6.9% minor complications). Neither major nor minor complications were more frequent in patients with technical risk factors (P > 0.05). In left-sided PTBD, hemorrhage was more frequent (P = 0.015). Patients with malignancy were significantly more affected by drainage-related complications (P = 0.004; odds ratio = 2.03). The mortality rate was 0.5% (n = 2). CONCLUSION: PTBD is a safe and effective method for the treatment of biliary obstruction and biliary leaks. Complication rates are low, even in procedures with risk factors.

Are We There Yet? The Value of Deep Learning in a Multicenter Setting for Response Prediction of Locally Advanced Rectal Cancer to Neoadjuvant Chemoradiotherapy.

This retrospective study aims to evaluate the generalizability of a promising state-of-the-art multitask deep learning (DL) model for predicting the response of locally advanced rectal cancer (LARC) to neoadjuvant chemoradiotherapy (nCRT) using a multicenter dataset. To this end, we retrained and validated a Siamese network with two U-Nets joined at multiple layers using pre- and post-therapeutic T2-weighted (T2w), diffusion-weighted (DW) images and apparent diffusion coefficient (ADC) maps of 83 LARC patients acquired under study conditions at four different medical centers. To assess the predictive performance of the model, the trained network was then applied to an external clinical routine dataset of 46 LARC patients imaged without study conditions. The training and test datasets differed significantly in terms of their composition, e.g., T-/N-staging, the time interval between initial staging/nCRT/re-staging and surgery, as well as with respect to acquisition parameters, such as resolution, echo/repetition time, flip angle and field strength. We found that even after dedicated data pre-processing, the predictive performance dropped significantly in this multicenter setting compared to a previously published single- or two-center setting. Testing the network on the external clinical routine dataset yielded an area under the receiver operating characteristic curve of 0.54 (95% confidence interval [CI]: 0.41, 0.65), when using only pre- and post-therapeutic T2w images as input, and 0.60 (95% CI: 0.48, 0.71), when using the combination of pre- and post-therapeutic T2w, DW images, and ADC maps as input. Our study highlights the importance of data quality and harmonization in clinical trials using machine learning. Only in a joint, cross-center effort, involving a multidisciplinary team can we generate large enough curated and annotated datasets and develop the necessary pre-processing pipelines for data harmonization to successfully apply DL models clinically.

Assessment of liver cirrhosis severity with extracellular volume fraction MRI.

We aimed to investigate the diagnostic utility of MRI extracellular volume fraction (ECV) for the assessment of liver cirrhosis severity as defined by Child-Pugh class. In this retrospective study, 90 patients (68 cirrhotic patients and 22 controls), who underwent multiparametric liver MRI, were identified. Hepatic T1 relaxation times and ECV were assessed. Clinical scores of liver disease severity were calculated. One-way analysis of variance (ANOVA) followed by Tukey's multiple comparison test, Spearman's correlation coefficient, and receiver operating characteristic (ROC) analysis were used for statistical analysis. In cirrhotic patients, hepatic native T1 increased depending on Child-Pugh class (620.5 ± 78.9 ms (Child A) vs. 666.6 ± 73.4 ms (Child B) vs. 828.4 ± 91.2 ms (Child C), P < 0.001). ECV was higher in cirrhotic patients compared to the controls (40.1 ± 11.9% vs. 25.9 ± 4.5%, P < 0.001) and increased depending of Child-Pugh class (33.3 ± 6.0% (Child A) vs. 39.6 ± 4.9% (Child B) vs. 52.8 ± 1.2% (Child C), P < 0.001). ECV correlated with Child-Pugh score (r = 0.64, P < 0.001). ECV allowed differentiating between Child-Pugh classes A and B, and B and C with an AUC of 0.785 and 0.944 (P < 0.001, respectively). The diagnostic performance of ECV for differentiating between Child-Pugh classes A and B, and B and C was higher compared to hepatic native T1 (AUC: 0.651 and 0.910) and MELD score (AUC: 0.740 and 0.795) (P < 0.05, respectively). MRI-derived ECV correlated with Child-Pugh score and had a high diagnostic performance for the discrimination of different Child-Pugh classes. ECV might become a valuable non-invasive biomarker for the assessment of liver cirrhosis severity.

Pilot Animal Study on Robotic-Assisted Endovascular Visceral Interventions.

PURPOSE: To evaluate technical feasibility and safety of common endovascular visceral interventions using a vascular robotic platform through preclinical study. MATERIAL AND METHODS: The CorPath GRX Robotic System (Corindus Inc, Waltham, Massachusetts) was tested in an anesthetized pig for its ability to navigate various commercially available devices in the abdominal vasculature and to perform routine endovascular visceral procedures. After manually placing a guiding catheter in the celiac trunk, several visceral branches were probed with microcatheters and -wires under robotic assistance, and embolization with liquids (lipiodol), detachable coils and plugs were performed. Furthermore, the origin of the celiac trunk was stented before accessing the left hypogastric artery for pelvic embolization. RESULTS: All procedures were performed with technical success and without any complications. Navigating the catheters and wires via the steering console proved intuitive. Coil, plug and stent deployment were exclusively controlled by remote with remarkable precision and stability. CONCLUSION: Robotic-assisted visceral embolization and stenting as well as pelvic embolization using the CorPath GRX System is feasible and safe. Application of the platform in the abdominal vasculature is demonstrated for the first time. Considering the precision and the potential for reducing the operator's radiation exposure, further research in this area is highly encouraged to enable translation into clinical practice.