Search

We present a framework with two options for selecting a subgroup of training cases for intracranial aneurysm (IA) treatment. Option one, general training, describes training cases that represent a variety of different IAs that represent the diversity of real world cases. This can be achieved via instance selection (IS), which reduces size of a dataset by eliminating redundancies via outlier removal and clustering. Option two, a specific training scenario, describes training that is specialized based on IA features of a specific case, for which we present the novel reverse instance selection (RIS), which introduces similarity to the specific case to the IS methodology. We evaluated our IS and RIS by comparing them to subsets selected based on similarity (SIM) and random sampling (RndS). RIS outperformed SIM and RndS in 79% of experiments. IS outperformed RndS in 33% of experiments. In both scenarios, we observed that our approaches, which balance the weaknesses of SIM and RndS, perform best for small subset sizes close to the database cluster size. Our IS and RIS are flexible in regards to the underlying machine learning and weighting of metrics for evaluation, thus providing a way to select a representative and diverse subset not just for IAs, but also for different kinds of data.

Intracranial mirror aneurysms are pathologic dilatations of cerebral arteries that occur at identical locations in both hemispheres. They do not always have identical sizes, shapes, or rupture states, despite identical patient-specific risk factors. As a subgroup of multiple aneurysms and thus small amounts of data, clinical research about their causes and development has not been conclusive. Using eight patients with 16 mirror aneurysms with different rupture states, we investigated differences between the hemispheres. We virtually performed stitching: detaching the aneurysm domes to place them on the opposite parent vessel to assess the influence of aneurysm morphology and upstream parent vessel shape on hemodynamics. In the unaltered datasets, we found that blood flow within aneurysms in the left hemisphere has higher wall shear stress and velocity. After stitching, we found that both upstream parent vessel as well as aneurysm morphology influence hemodynamics. Morphologic parameters like size, undulation index, and aneurysm angle had a higher influence. In a short rupture risk analysis we found a correlation between rupture and aspect ratio and non-sphericity index. Overall, we found possible indicators for mirror aneurysm analysis that illustrated the importance of morphological analysis, though more datasets are necessary for conclusive statements about special treatment for mirror aneurysms.

Background: The rotational direction (RD) of helical blood flow can be classified as either a clockwise (RD + ) or counter-clockwise (RD - ) flow. We hypothesized that this simple classification might not be sufficient for analysis in vivo and a simultaneous existence of RD +/- may occur. We utilized volumetric velocity-sensitive cardiovascular magnetic resonance imaging (4D flow MRI) to analyze rotational blood flow in the thoracic aorta., Methods: Forty volunteers (22 females; mean age, 41±16 years) and seventeen patients with bicuspid aortic valves (BAVs) (9 females; mean age, 42±14 years) were prospectively included. The RDs and the calculation of the rotating blood volumes (RBVs) in the thoracic aorta were performed using a pathline-projection strategy., Results: We could confirm a mainly clockwise RD in the ascending, descending aorta and in the aortic arch. Furthermore, we found a simultaneous existence of RD + /RD - . The RD +/- -volume in the ascending aorta was significantly higher in BAV patients, the mean RD + /RD - percentage was approximately 80%/20% vs. 60%/40% in volunteers (P<0.01). The maximum RBV always occurred during systole. There was significantly more clockwise than counter-clockwise rotational flow in the ascending aorta (P<0.01) and the aortic arch (P<0.01), but no significant differences in the descending aorta (P=0.48)., Conclusions: A simultaneous occurrence of RD + /RD - indicates that a simple categorization in either of both is insufficient to describe blood flow in vivo . Rotational flow in the ascending aorta and in the aortic arch differs significantly from flow in the descending aorta. BAV patients show significantly more clockwise rotating volume in the ascending aorta compared to healthy volunteers., Competing Interests: Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-23-183/coif). AA was supported by an ESOR/ESCR 2019 cardiac fellowship. The other authors have no conflicts of interest to declare., (2023 Quantitative Imaging in Medicine and Surgery. All rights reserved.)

Background: Parameters of body composition have prognostic potential in patients with oncologic diseases. The aim of the present study was to analyse the prognostic potential of radiomics-based parameters of the skeletal musculature and adipose tissues in patients with advanced hepatocellular carcinoma (HCC)., Methods: Radiomics features were extracted from a cohort of 297 HCC patients as post hoc sub-study of the SORAMIC randomized controlled trial. Patients were treated with selective internal radiation therapy (SIRT) in combination with sorafenib or with sorafenib alone yielding two groups: (1) sorafenib monotherapy (n = 147) and (2) sorafenib and SIRT (n = 150). The main outcome was 1-year survival. Segmentation of muscle tissue and adipose tissue was used to retrieve 881 features. Correlation analysis and feature cleansing yielded 292 features for each patient group and each tissue type. We combined 9 feature selection methods with 10 feature set compositions to build 90 feature sets. We used 11 classifiers to build 990 models. We subdivided the patient groups into a train and validation cohort and a test cohort, that is, one third of the patient groups., Results: We used the train and validation set to identify the best feature selection and classification model and applied it to the test set for each patient group. Classification yields for patients who underwent sorafenib monotherapy an accuracy of 75.51% and area under the curve (AUC) of 0.7576 (95% confidence interval [CI]: 0.6376-0.8776). For patients who underwent treatment with SIRT and sorafenib, results are accuracy = 78.00% and AUC = 0.8032 (95% CI: 0.6930-0.9134)., Conclusions: Parameters of radiomics-based analysis of the skeletal musculature and adipose tissue predict 1-year survival in patients with advanced HCC. The prognostic value of radiomics-based parameters was higher in patients who were treated with SIRT and sorafenib., (© 2023 The Authors. Journal of Cachexia, Sarcopenia and Muscle published by Wiley Periodicals LLC.)

Background: For prediction of many types of clinical outcome, the skeletal muscle mass can be used as an independent biomarker. Manual segmentation of the skeletal muscles is time-consuming, therefore we present a deeplearning-based approach for the identification of muscle mass at the L3 level in clinical routine computed tomographic (CT) data., Patients and Methods: We conducted a retrospective study of 130 patient datasets. Individual CT slice analysis at the L3 level was fed into a U-Net architecture. As a result, we obtained segmentations of the musculus rectus abdominis, abdominal wall muscles, musculus psoas major, musculus quadratus lumborum and musculus erector spinae in the CT-slice at the L3 level., Results: The Dice score was 0.95±0.02, 0.86±0.12, 0.93±0.05, 0.92±0.05, 0.86±0.08 for the erector spine, rectus, abdominal wall, psoas and quadratus lumborum muscles, respectively. For the overall skeletal muscle mass, the test data achieved a Dice score of 0.95±0.03., Conclusion: Our network achieved Dice scores larger than 0.86 for each of the five different muscle types and 0.95 for the overall skeletal muscle mass. The subdivision of muscle types can serve as a basis for obtaining future biomarkers. Our network is publicly available so that it might be beneficial for others to improve the clinical workflow within examination of routine CT scans., (Copyright © 2022, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.)

Numerous studies assess intracranial aneurysm rupture risk based on morphological and hemodynamic parameter analysis in addition to clinical information such as aneurysm localization, age, and sex. However, intracranial aneurysms mostly occur with a saccular shape located either lateral to the parent artery or at a bifurcation. In contrast, fusiform intracranial aneurysms (FIAs), i.e., aneurysms with a non-saccular, dilated form, occur in approximately 3-13% of all cases and therefore have not yet been as thoroughly studied. To improve the understanding of FIA hemodynamics, this pilot study contains morphological analyses and image-based blood flow simulations in three patient-specific cases. For a precise and realistic comparison to the pre-pathological state, each dilation was manually removed and the time-dependent blood flow simulations were repeated. Additionally, a validated fast virtual stenting approach was applied to evaluate the effect of virtual endovascular flow-diverter deployment focusing on relevant hemodynamic quantities. For two of the three patients, post-interventional information was available and included in the analysis. The results of this numerical pilot study indicate that complex flow structures, i.e., helical flow phenomena and the presence of high oscillating flow features, predominantly occur in FIAs with morphologically differing appearances. Due to the investigation of the individual healthy states, the original flow environment could be restored which serves as a reference for the virtual treatment target. It was shown that the realistic deployment led to a considerable stabilization of the individual hemodynamics in all cases. Furthermore, a quantification of the stent-induced therapy effect became feasible for the treating physician. The results of the morphological and hemodynamic analyses in this pilot study show that virtual stenting can be used in FIAs to quantify the effect of the planned endovascular treatment., Competing Interests: DB is a Proctor/Advisor: Phenox, Balt, Acandis, Thromb X, Perflow, Kaneka, Penumbra, outside of the submitted work. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest., (Copyright © 2022 Saalfeld, Stahl, Korte, Miller Marsh, Preim, Beuing, Cherednychenko, Behme and Berg.)

Background: Preoperative three-dimensional (3D) reconstructions for liver surgery planning have been shown to be effective in reduction of blood loss and operation time. However, the role of the 'presentation modality' is not well investigated. We present the first study to compare 3D PDFs, 3D printed models (PR) and virtual reality (VR) 3D models with regard to anatomical orientation and personal preferences in a high volume liver surgery center., Methods: Thirty participants, 10 medical students, 10 residents, 5 fellows and 5 hepatopancreatobiliary (HPB) experts, assigned the tumor-bearing segments of 20 different patient's individual liver reconstructions. Liver models were presented in a random order in all modalities. Time needed to specify the tumor location was recorded. In addition, a score was calculated factoring in correct, wrong and missing segment assignments. Furthermore, standardized test/questionnaires for spatial thinking and seeing, vegetative side effects and usability were completed., Results: Participants named significantly more correct segments in VR (P=0.040) or PR (P=0.036) compared to PDF. Tumor assignment was significantly shorter with 3D PR models compared to 3D PDF (P<0.001) or VR application (P<0.001). Regardless of the modality, HPB experts were significantly faster (24±8 vs. 35±11 sec; P=0.014) and more often correct (0.87±0.12 vs. 0.83±0.15; P<0.001) than medical students. Test results for spatial thinking and seeing had no influence on time but on correctness of tumor assignment. Regarding usability and user experience the VR application achieved the highest scores without causing significant vegetative symptoms and was also the most preferred method (n=22, 73.3%) because of the multiple functions like scaling and change of transparency. Ninety percent (n=27) stated that this application can positively influence the operation planning., Conclusions: 3D PR models and 3D VR models enable a better and partially faster anatomical orientation than reconstructions presented as 3D PDFs. User's preferred the VR application over the PR models and PDF. A prospective trial is needed to evaluate the different presentation modalities regarding intra- and postoperative outcomes., Competing Interests: Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at http://dx.doi.org/10.21037/atm-21-512). Dr. FH reports grants from German Federal Ministry of Education and Research, other from Medical Centre of the Johannes Gutenberg University of Mainz, during the conduct of the study. Dr. PS, Dr. CH, Dr. BP, Dr. WK, Dr. HL, and Dr. TH report grants from German Federal Ministry of Education and Research, during the conduct of the study. AP has no conflicts of interest to declare., (2021 Annals of Translational Medicine. All rights reserved.)

Introduction: A multitude of approaches have been postulated for assessing the risk of intracranial aneurysm rupture. However, the amount of potential predictive factors is not applicable in clinical practice and they are rejected in favor of the more practical PHASES score. For the subgroup of multiple intracranial aneurysms (MIAs), the PHASES score might severely underestimate the rupture risk, as only the aneurysm with the largest diameter is considered for risk evaluation., Methods: In this study, we investigated 38 patients harboring a total number of 87 MIAs with respect to their morphological and hemodynamical characteristics. For the determination of the best suited parameters regarding their predictive power for aneurysm rupture, we conducted three phases of statistical evaluation. The statistical analysis aimed to identify parameters that differ significantly between ruptured and unruptured aneurysms, show smallest possible correlations among each other and have a high impact on rupture risk prediction., Results: Significant differences between ruptured and unruptured aneurysms were found in 16 out of 49 parameters. The lowest correlation were found for gamma, aspect ratio (AR1), aneurysm maximal relative residence time (Aneurysm&#95;RRT&#95;max) and aneurysm mean relative residence time. The data-driven parameter selection yielded a significant correlation of only two parameters (AR1 and the Aneurysm&#95;RRT&#95;max) with rupture state (area under curve = 0.75)., Conclusion: A high number of established morphological and hemodynamical parameters seem to have no or only low effect on prediction of aneurysm rupture in patients with MIAs. For best possible rupture risk assessment of patients with MIAs, only the morphological parameter AR1 and the hemodynamical parameter Aneurysm&#95;RRT&#95;max need to be included in the prediction model., Competing Interests: Conflict of interest statement: The authors declare that there is no conflict of interest., (© The Author(s), 2020.)

4D flow MRI enables quantitative assessment of helical flow. Current methods are susceptible to noise. To evaluate helical flow patterns in healthy volunteers and patients with bicuspid aortic valves (BAV) at 1.5 T and 3 T using pressure-based helix-extraction in 4D flow MRI. Two intraindividual 4D flow MRI examinations were performed at 1.5 T and 3 T in ten healthy volunteers (5 females, 32 ± 3 years) and 2 patients with BAV using different acceleration techniques (kt-GRAPPA and centra). Several new quantitative parameters for the evaluation of volumes [ml], lengths [mm] as well as temporal parameters [ms] of helical flow were introduced and analyzed using the software tool Bloodline. We found good correlations between measurements in volunteers at 1.5 T and 3 T regarding helical flow volumes (R = 0.98) and temporal existence (R = 0.99) of helices in the ascending aorta. Furthermore, we found significantly larger (11.7 vs. 77.6 ml) and longer lasting (317 vs. 769 ms) helices in patients with BAV than in volunteers. The assessed parameters do not depend on the magnetic field strength used for the acquisition. The technique of pressure-based extraction of 4D flow MRI pattern is suitable for differentiation of normal and pathological flow.

To compare two broadly used 4D-flow- with a 2D-flow-sequence in healthy volunteers, regarding absolute flow parameters, image quality (IQ), and eddy current correction (ECC). Forty volunteers (42 ± 11.8 years, 22 females) were examined with a 3T scanner. Thoracic aortic flow was assessed using a 3D-T2w-SPACE-STIR-sequence for morphology and two accelerated 4D-flow sequences for comparison, one with k-t undersampling and one with standard GRAPPA parallel-imaging. 2D-flow was used as reference standard. The custom-made software tool Bloodline enabled flow measurements for all analyses at the same location. Quantitative flow analyses were performed with and without ECC. One reader assessed pathline IQ (IQ-PATH) and occurrence of motion artefacts (IQ-ART) on a 3-point grading scale, the higher the better. k-t GRAPPA allowed a significant mean scan time reduction of 46% (17:56 ± 5:26 min vs. 10:40 ± 3:15 min) and provided significantly fewer motion artefacts than standard GRAPPA (IQ-ART 1.57 ± 0.55 vs. 0.84 ± 0.48; p < 0.001). Neither 4D-flow sequence significantly differed in flow volume nor peak velocity results with or without ECC. Nevertheless, the correlation between both 4D-flow sequences and 2D-flow was better with ECC; the k-t GRAPPA sequence performed best (R = 0.96 vs. 0.90). k-t GRAPPA 4D-flow was not inferior to a standard GRAPPA-sequence, showed fewer artefacts, comparable IQ and was almost two-fold faster.

Background: Four-dimensional (4D) flow magnetic resonance imaging (MRI) sequences with advanced parallel imaging have the potential to reduce scan time with equivalent image quality and accuracy compared with standard two-dimensional (2D) flow MRI. We compared 4D flow to standard 2D flow sequences using a constant and pulsatile flow phantom at 3 T., Methods: Two accelerated 4D flow sequences (GRAPPA2 and k-t-GRAPPA5) were evaluated regarding the concordance of flow volumes, flow velocities, and reproducibility as well as dependency on measuring plane and velocity encoding (V enc ). The calculated flow volumes and peak velocities of the phantom were used as reference standard. Flow analysis was performed using the custom-made software "Bloodline"., Results: No significant differences in flow volume were found between the 2D, both 4D flow MRI sequences, and the pump reference (p = 0.994) or flow velocities (p = 0.998) in continuous and pulsatile flow. An excellent correlation (R = 0.99-1.0) with a reference standard and excellent reproducibility of measurements (R = 0.99) was achieved for all sequences. A V enc overestimated by up to two times had no impact on flow measurements. However, misaligned measuring planes led to an increasing underestimation of flow volume and mean velocity in 2D flow accuracy, while both 4D flow measurements were not affected. Scan time was significantly shorter for k-t-GRAPPA5 (1:54 ± 0:01 min, mean ± standard deviation) compared to GRAPPA2 (3:56 ± 0:02 min) (p = 0.002)., Conclusions: Both 4D flow sequences demonstrated equal agreement with 2D flow measurements, without impact of V enc overestimation and plane misalignment. The highly accelerated k-t-GRAPPA5 sequence yielded results similar to those of GRAPPA2.

Rationale and Objectives: The quantitative assessment of neck lymph nodes in the context of malignant tumors requires an efficient segmentation technique for lymph nodes in tomographic three-dimensional (3D) datasets. We present a stable 3D mass-spring model for lymph node segmentation in computed tomography (CT) datasets., Materials and Methods: For the first time our model concurrently represents the characteristic gray value range, directed contour information, and shape knowledge, which leads to a robust and efficient segmentation process., Results: Our model design and the segmentation accuracy were both evaluated with 40 lymph nodes from five clinical CT datasets containing malignant tumors of the neck., Conclusion: The segmentation accuracy proved to be comparable to that of manual segmentations by experienced users and significantly reduced the time and interaction needed for the lymph node segmentation.