Development of multi-compartment model of the liver using image-based meshing software

Problems in the assessment of treatment response

To find out if the use of different virtual monoenergetic data sets enabled by DECT technology might have a negative impact on post-processing applications, specifically in case of the "unfolded ribs" algorithm. Metal or beam hardening artifacts are suspected to generate image artifacts and thus reduce diagnostic accuracy. This paper tries to find out how the generation of "unfolded rib" CT image reformates is influenced by different virtual monoenergetic CT images and looks for possible improvement of the post-processing tool. Between March 2021 and April 2021, thin-slice dual-energy CT image data of the chest were used creating "unfolded rib" reformates. The same data sets were analyzed in three steps: first the gold standard with the original algorithm on mixed image data sets followed by the original algorithm on different keV levels (40-120keV) and finally using a modified algorithm which in the first step used segmentation based on mixed image data sets, followed by segmentation based on different keV levels. Image quality (presence of artifacts), lesion and fracture detectability were assessed for all series. Both, the original and the modified algorithm resulted in more artifact-free image data sets compared to the gold standard. The modified algorithm resulted in significantly more artifact-free image data sets at the keV-edges (40-120keV) compared the original algorithm. Especially "black artifacts" and pseudo-lesions, potentially inducing false positive findings, could be reduced in all keV level with the modified algorithm. Detection of focal sclerotic, lytic or mixed (k = 0.990-1.000) lesions was very good for all keV levels. The Fleiss-kappa test for detection of fresh and old rib fractures was ≥ 0.997. The use of different virtual monoenergetic keVs for the "unfolded rib" algorithm is generating different artifacts. Segmentation-based artifacts could be eliminated by the proposed new algorithm, showing the best results at 70-80keV.

A systematic review and meta-analysis of computed tomography in the diagnosis of pediatric foreign body aspiration

Data suggest that clinicians, when evaluating pediatric patients with blunt head trauma, may be overordering head computed tomography (CT). Prior decision instruments (DIs) aimed at aiding clinicians in safely forgoing CTs may be paradoxically increasing CT utilization. This study evaluated a novel DI that aims for high sensitivity while also improving specificity over prior instruments. We conducted a planned secondary analysis of the NEXUS Head CT DI among patients less than 18 years old. The rule required patients satisfy seven criteria to achieve "low-risk" classification. Patients were assigned "high-risk" status if they fail to meet one or more criteria. Our primary outcome was the ability of the rule to identify all patients requiring neurosurgical intervention. The study enrolled 1,018 blunt head injury pediatric patients. The DI assigned high-risk status to 27 of 27 patients requiring neurosurgical intervention (sensitivity= 100.0%, 95% confidence interval [CI]= 87.2%-100%]). The instrument assigned low-risk status to 330 of 991 patients who did not require neurosurgical intervention (specificity= 33.3%, 95% CI= 30.3%-36.3%). None of the 991 low-risk patients required neurosurgical intervention (negative predictive value [NPV]= 100%, 95% CI= 99.6%-100%). The DI correctly assigned high-risk status to 48 of the 49 patients with significant intracranial injuries, yielding a sensitivity of 98.0% (95% CI= 89.1%-99.9%). The instrument assigned low-risk status to 329 of 969 patients who did not have significant injuries to yield a specificity of 34.0% (95% CI= 31.0%-37.0%). Significant injuries were absent in 329 of the 330 patients assigned low-risk status to yield a NPV of 99.7% (95% CI= 98.3%-100%). The Pediatric NEXUS Head CT DI reliably identifies blunt trauma patients who require head CT imaging and could significantly reduce the use of CT imaging.

The characterization of adnexal masses has been facilitated remarkably by the use of pelvic ultrasound during the last decade. Its ease of use, accessibility, relatively low cost, and the recent introduction of color and duplex scanning has made this modality invaluable. Please see other chapters in this volume for a discussion of the use of ultrasound for differentiating benign from malignant ovarian masses. For patients with advanced ovarian cancer and a known pelvic mass which can be palpated on bimanual exam, however, ultrasonography is rarely helpful other than to confirm the presence of ascites. While magnetic resonance imaging (MRI) provides precise information regarding the location of disease and invasion of tissue planes, such detailed anatomic relationships are generally not clinically relevant for treatment planning in patients with advanced ovarian carcinoma, particularly given the associated costs. This is in contradistinction to cervical cancer, in which MRI has been shown to be sensitive for the detection of parametrial involvement, a parameter which significantly alters the plan of treatment if present (Yu et al. 1998). 18F fluorodeoxyglucose-positron emission tomography (FDG-PET) is highly sensitive for cancerous lesions and may ultimately prove useful for the detection of recurrences, but the cost/benefit ratio is poor for the preoperative evaluation of primary ovarian cancer. Computed tomo­graphy (CT) offers advantages over other techniques: relatively low cost, fast scan times, wide availability, and evaluation of the entire abdominal cavity. Furthermore, the use of intravenous and oral contrast improves visualization of retroperitoneal anatomy (Figure 8.1). For these reasons, CT has become a common diagnostic procedure to assess the extent of disease and plan surgical interventions in patients with advanced ovarian cancer. This chapter addresses the use of CT imaging in this cohort of patients and its potential to predict surgical outcome.

SAT0551 Sensitivity of Chest Radiography in the Early Diagnosis of Sarcoidosis: is it Really Should be Done?

Aortic Dissection as a Complication of Celiac Plexus Block

Purpose: Present methods of generating four-dimensional (4D) computed tomography (CT) image data sets that bin either projections or reconstructed images based on the phase of the respiratory cycle might display artifacts caused by irregularities in the respiratory cycle. Binning based solely on displacement has the potential for resolving these artifacts but yields three-dimensional image data sets that are unevenly spaced in time, making the application of the 4D data set to treatment planning more difficult. We propose a method for displacement-based binning that sets the binning points at displacement points corresponding to approximately equally-spaced phases. Method and Materials: The present approach extracts the file used to monitor the patient's respiratory cycle and identifies the points on the respiratory cycle corresponding to true end-inspiration. The time intervals between each set of end inspiration points were divided into a specified number of phases (typically 10) equally spaced in time. The displacements corresponding to these phase points were averaged among the respiratory cycles included in the image acquisition, and the times in each respiratory cycle corresponding to the mean values were recorded and sent to the reconstruction program. A MATLAB program was written to generate these times and applied to a typical respiratory cycle. Results: When evenly-spaced displacements were used for displacement binning, the mean time intervals among the 10 phases varied by as much as a factor of 5. The present methodology reduced the variation in mean time intervals among phases to approximately 20%. Conclusion: f binning for acquisition of 4D CT image data sets is to be based on displacement of the respiratory trace, then, to ensure approximately equally-spaced time intervals, the displacement for a specific phase averaged over all respiratory cycles should be used. Conflict of Interest: Supported in part by a Sponsored Research Agreement with Philips Medical Systems, Inc.

With the advent of slip ring technology and the introduction of spiral (helical) computed tomography (CT) [1], the decade of the 1990s saw CT’s role in imaging the urinary tract expand. Now, with the introduction of multi-detector technology, CT will become an even more important imaging tool in the urinary tract [2]. Multi-detector technology gives CT scanners the capability of creating four slices per gantry rotation. The four slices are derived from four channels of data, which emanate from an array of multiple detectors; the number of which depends on the manufacturer. Therefore, relative to single-slice CT (SCT), multi-slice CT (MSCT) results in a four-fold increase in data per gantry revolution. This results in four main advantages. The first is speed, because there is a four-fold increase in speed relative to SCT. Some manufacturers utilize two gantry rotations per second. This results in an eight-fold increase in speed. The second main advantage is that MSCT scanners allow us to obtain images over a larger area with thin collimation. This results in improved spatial resolution, not only in the z-axis with reconstructed slice thicknesses as low as 0.5 mm but near isotropic resolution in non-axial planes. The third advantage relates to scan volume. These scanners can scan larger volumes during a single breath hold, with the ability to scan from head to toe in a matter of seconds. The fourth, and perhaps most important advantage, is the flexibility provided both in acquisition protocols and in reconstruction algorithms. The radiologist has multiple options to choose from when designing CT protocols. Acquisition parameters radiologists must consider include collimation, rotation speed, pitch, and dose. Concerning reconstruction options, unlike SCT, image data sets with variable slice thicknesses can be obtained, so long as the thickness is not less than the collimation. For example, data acquisition can be formed with 1-mm collimation (or four 1-mm slices per gantry rotation) and reconstructed using both 1-mm and 2.5-mm slices or, data acquisition can be formed with 2.5-mm collimation (or four 2.5-mm slices per gantry rotation) and reconstructed using both 2.5-mm and 5-mm slices. The thin slice data can be used subsequently for multi-planar reconstruction, and the thicker slices can be viewed axially.

With continuing developments in computed tomography (CT) technology and its increasing use of CT imaging, the ionizing radiation dose from CT is becoming a major public concern particularly for high-dose applications such as cardiac imaging. We recently proposed a novel interior tomography approach for x-ray dose reduction that is very different from all the previously proposed methods. Our method only uses the projection data for the rays passing through the desired region of interest. This method not only reduces x-ray dose but scatter as well. In this paper, we quantify the reduction in the amount of x-ray dose and scattered radiation that could be achieved using this method. Results indicate that interior tomography may reduce the x-ray dose by 18% to 58% and scatter to the detectors by 19% to 59% as the FOV is reduced from 50 to 8.6 cm.

The purpose of this study was to determine the clinical opportunities for the use of computed tomography (CT) imaging for inferring bone quality and to critically analyze the correlation between dual x-ray absorptiometry (DXA) and diagnostic CT as reported in the literature. A systematic review of the MEDLINE database was performed in February 2016 using the PubMed interface. The inclusion criteria were English language, studies performed using living human subjects, studies pertaining to orthopaedics, use of conventional diagnostic CT scans, studies that measured cancellous bone, and studies that reported Hounsfield unit (HU) measurements directly rather than a computed bone mineral density. Thirty-seven studies that reported on a total of 9,109 patients were included. Of these, 10 studies correlated HU measurements of trabecular bone with DXA-based bone assessment. Reported correlation coefficients ranged between 0.399 and 0.891, and 5 of the studies reported appropriate threshold HU levels for diagnosing osteoporosis or osteopenia. Direct HU measurement from diagnostic CT scans has the potential to be used opportunistically for osteoporosis screening, but in its current state it is not ready for clinical implementation. There is a lack of exchangeability among different machines that limits its broad applicability. Future research efforts should focus on identifying thresholds at specific anatomic regions in high-risk patients in order to have the greatest impact on patients. However, using diagnostic CT to infer region-specific osteoporosis could be extraordinarily valuable to orthopaedic surgeons and primary care physicians, and merits further research.

Preoperative nerve imaging using computed tomography in patients with heterotopic ossification of the elbow

Introduction: Peripheral artery disease (PAD) is characterized by atherosclerosis of lower extremity arteries that promotes reduced muscle perfusion and skeletal muscle atrophy, which contributes to functional impairment in PAD patients. Computed tomography (CT) imaging offers a non-invasive method for quantifying peripheral muscle characteristics; however, manual image segmentation of muscles is time-consuming, and regional muscle analysis remains understudied in PAD. Therefore, we sought to develop and validate a deep learning approach for calf muscle segmentation to assess serial regional changes in muscle density of PAD patients over an 18-month study period. Methods: Patients with PAD (n=89) were prospectively enrolled for lower extremity non-contrast CT imaging. A subset of patients (n=45) was recruited for an additional follow-up CT scan 18 months later. Calf muscle groups, including the gastrocnemius, soleus, and tibialis anterior, were manually segmented from axial CT images of each patient’s symptomatic limb, and mean muscle densities for each muscle were quantified based on CT Hounsfield units. Manual segmentations were used to train an nnU-Net deep learning model. Data augmentation techniques were performed to enhance generalization. The dataset was split into an 80/20 ratio for the training and test sets. Dice coefficients were calculated to evaluate the overlap/agreement between manual and deep learning segmentations. Paired t-tests were performed to assess the differences in muscle densities between the baseline and 18-month follow-up measures. Results: The deep learning model achieved high segmentation performance, with dice scores of 0.90 ± 0.02 for the gastrocnemius and soleus, and 0.89 ± 0.02 for the tibialis anterior. Deep learning-guided serial CT image analysis detected a significant reduction in muscle densities across all muscle groups at 18-month follow-up when compared to baseline measurements (p<0.05). Conclusion: Deep learning analysis can enable rapid, regional analysis of lower extremity skeletal muscle characteristics, reducing the time for muscle-by-muscle analysis in PAD patients from hours to seconds. Future implementation of regional muscle analysis in PAD may assist vascular medicine specialists with identifying regional muscle wasting that is related to patient walking impairment and/or symptoms as well as aid in serial monitoring of PAD patients following revascularization or supervised exercise for claudication.

Real-time Ultrasound (US) image fusion with a pre-acquired second imaging dataset - Computed Tomography (CT), Magnetic Resonance Imaging (MRI) and/or CT/PET - has become widely used in recent years for both diagnosis and image-guided interventional procedures. Liver and kidneys are the main focused anatomical districts, related to abdominal application. There are still nowadays some drawbacks, regarding the adoption of the fusion imaging technique in everyday practice especially regarding its ease of use and the time needed in order to obtain a precise real-time fusion between US and the second imaging modality. The present work is a preliminary study on the feasibility and practical use of an Automatic registration algorithm for CT-US real-time fusion imaging. Data obtained by tests performed on a Doppler phantom, for the assessment of the precision of the registration procedure and in-vivo Automatic registration tests, are presented.

Deep learning based model for classification of COVID −19 images for healthcare research progress