[Show abstract][Hide abstract] ABSTRACT: The objective of this study was to investigate the improvement in diagnostic quality of an iterative model-based reconstruction (IMBR) algorithm for low-tube-voltage (80-kVp) and low-tube-current in abdominal computed tomography angiography (CTA). A total of 11 patients were imaged on a 256-slice multidetector computed tomography for visualization of the aorta. For all patients, three different reconstructions from the low-tube-voltage data are generated: filtered backprojection (FBP), IMBR, and a mixture of both [Formula: see text]. To determine the diagnostic value of IMBR-based reconstructions, the image quality was assessed. With IMBR-based reconstructions, image noise could be significantly reduced, which was confirmed by a highly improved contrast-to-noise ratio. In the image quality assessment, radiologists were able to reliably detect more third-order and higher aortic branches in the IMBR reconstructions compared to FBP reconstructions. The effective dose level was, on average, 3.0 mSv for 80-kVp acquisitions. Low-tube-voltage CTAs significantly improve vascular contrast as presented by others; however, this effect in combination with IMBR enabled yet another substantial improvement of diagnostic quality. For IMBR, a significant improvement of image quality and a decreased radiation dose at low-tube-voltage can be reported.
[Show abstract][Hide abstract] ABSTRACT: Purpose
Evaluation of a new software tool for generation of simulated low-dose computed tomography (CT) images from an original higher dose scan.
Materials and Methods
Original CT scan data (100 mAs, 80 mAs, 60 mAs, 40 mAs, 20 mAs, 10 mAs; 100 kV) of a swine were acquired (approved by the regional governmental commission for animal protection). Simulations of CT acquisition with a lower dose (simulated 10–80 mAs) were calculated using a low-dose simulation algorithm. The simulations were compared to the originals of the same dose level with regard to density values and image noise. Four radiologists assessed the realistic visual appearance of the simulated images.
Image characteristics of simulated low dose scans were similar to the originals. Mean overall discrepancy of image noise and CT values was −1.2% (range −9% to 3.2%) and −0.2% (range −8.2% to 3.2%), respectively, p>0.05. Confidence intervals of discrepancies ranged between 0.9–10.2 HU (noise) and 1.9–13.4 HU (CT values), without significant differences (p>0.05). Subjective observer evaluation of image appearance showed no visually detectable difference.
Simulated low dose images showed excellent agreement with the originals concerning image noise, CT density values, and subjective assessment of the visual appearance of the simulated images. An authentic low-dose simulation opens up opportunity with regard to staff education, protocol optimization and introduction of new techniques.
[Show abstract][Hide abstract] ABSTRACT: Neoadjuvant chemotherapy is an accepted standard of care for locally advanced esophagogastric cancer. As only a subgroup benefits, a response-based tailored treatment would be of interest. The aim of our study was the evaluation of the prognostic and predictive value of clinical response in esophagogastric adenocarcinomas.
Clinical response based on a combination of endoscopy and computed tomography (CT) scan was evaluated retrospectively within a prospective database in center A and then transferred to center B. A total of 686/740 (A) and 184/210 (B) patients, staged cT3/4, cN0/1 underwent neoadjuvant chemotherapy and were then re-staged by endoscopy and CT before undergoing tumor resection. Of 184 patients, 118 (B) additionally had an interim response assessment 4-6 weeks after the start of chemotherapy.
In A, 479 patients (70 %) were defined as clinical nonresponders, 207 (30 %) as responders. Median survival was 38 months (nonresponders: 27 months, responders: 108 months, log-rank, p < 0.001). Clinical and histopathological response correlated significantly (p < 0.001). In multivariate analysis, clinical response was an independent prognostic factor (HR for death 1.4, 95 %CI 1.0-1.8, p = 0.032). In B, 140 patients (76 %) were nonresponders and 44 (24 %) responded. Median survival was 33 months, (nonresponders: 27 months, responders: not reached, p = 0.003). Interim clinical response evaluation (118 patients) also had prognostic impact (p = 0.008). Interim, preoperative clinical response and histopathological response correlated strongly (p < 0.001).
Preoperative clinical response was an independent prognostic factor in center A, while in center B its prognostic value could only be confirmed in univariate analysis. The accordance with histopathological response was good in both centers, and interim clinical response evaluation showed comparable results to preoperative evaluation.
[Show abstract][Hide abstract] ABSTRACT: To develop a prototype algorithm for automatic spine segmentation in MDCT images and use it to automatically detect osteoporotic vertebral fractures.
Cross-sectional routine thoracic and abdominal MDCT images of 71 patients including 8 males and 9 females with 25 osteoporotic vertebral fractures and longitudinal MDCT images of 9 patients with 18 incidental fractures in the follow-up MDCT were retrospectively selected. The spine segmentation algorithm localised and identified the vertebrae T5-L5. Each vertebra was automatically segmented by using corresponding vertebra surface shape models that were adapted to the original images. Anterior, middle, and posterior height of each vertebra was automatically determined; the anterior-posterior ratio (APR) and middle-posterior ratio (MPR) were computed. As the gold standard, radiologists graded vertebral fractures from T5 to L5 according to the Genant classification in consensus.
Using ROC analysis to differentiate vertebrae without versus with prevalent fracture, AUC values of 0.84 and 0.83 were obtained for APR and MPR, respectively (p < 0.001). Longitudinal changes in APR and MPR were significantly different between vertebrae without versus with incidental fracture (ΔAPR: -8.5 % ± 8.6 % versus -1.6 % ± 4.2 %, p = 0.002; ΔMPR: -11.4 % ± 7.7 % versus -1.2 % ± 1.6 %, p < 0.001).
This prototype algorithm may support radiologists in reporting currently underdiagnosed osteoporotic vertebral fractures so that appropriate therapy can be initiated.
• This spine segmentation algorithm automatically localised, identified, and segmented the vertebrae in MDCT images. • Osteoporotic vertebral fractures could be automatically detected using this prototype algorithm. • The prototype algorithm helps radiologists to report underdiagnosed osteoporotic vertebral fractures.
No preview · Article · Jan 2014 · European Radiology
[Show abstract][Hide abstract] ABSTRACT: Recent data suggest primary resection as the preferable approach in patients with signet ring cell gastric cancer (SRC). The aim of our retrospective exploratory study was to evaluate the influence of SRC on prognosis and response in esophagogastric adenocarcinoma treated with neoadjuvant chemotherapy.
A total of 723 locally advanced esophagogastric adenocarcinomas (cT3/4 N any) documented in a prospective database from two academic centers were classified according to the WHO definition for SRC (more than 50 % SRC) and analyzed for their association with response and prognosis after neoadjuvant treatment.
A total of 235 tumors (32.5 %) contained SRC. Median survival of SRC was 26.3 compared with 46.6 months (p < 0.001) for non-SRC. SRC were significantly associated with female gender, gastric localization, advanced ypT and R1/2 categories, and lower risk of surgical complications and anastomotic leakage (each p < 0.001). Clinical (21.1 vs. 33.7 %, p = 0.001) and histopathological response (less than 10 % residual tumor: 16.3 vs. 28.9 %, p < 0.001) were significantly less frequent in SRC. Clinical response (p = 0.003) and complete histopathological response (pCR) (3.4 %) (p = 0.003) were associated with improved prognosis in SRC. Clinical response, surgical complications, ypTN categories, but not SRC were independent prognostic factors in forward Cox regression analysis in R0 resected patients. Risk of peritoneal carcinomatosis was increased (p < 0.001), while local (p = 0.015) and distant metastases (p = 0.02) were less frequent than in non-SRC.
Prognosis of SRC is unfavorable. Although response to neoadjuvant chemotherapy is rare in SRC, it is associated with improved outcome. Thus, chemotherapy might not generally be abandoned in SRC. A stratification based on SRC should be included in clinical trials.
No preview · Article · Jan 2014 · Annals of Surgical Oncology
[Show abstract][Hide abstract] ABSTRACT: PURPOSE
Evaluation of a new software tool for generation of simulated low-dose computed tomography (CT) images from an original higher dose scan.
METHOD AND MATERIALS
Original contrast-enhanced and non-enhanced CT examinations (120 kVp; 100 mAs, 80 mAs, 60 mAs, 40 mAs, 20 mAs, and 10 mAs) of a swine were acquired. Simulations of CT images with a lower radiation exposure (range 10-80 mAs) were calculated using a low-dose simulation algorithm that simulates accurately both photon noise and electronic noise that would be present in a scan at lower dose. Simulated non-enhanced images were compared to the original non-enhanced CT data of the same radiation dose level regarding density values and image noise. Four radiologists assessed the visual appearance of the simulated contrast-enhanced CT data.
Image characteristics of simulated low-dose scans were similar to the original acquisitions. Mean overall discrepancy of image noise and CT values between original and simulated CT images was 0.2 % (range -0.6 % to 0.8 %) and -0.3 % (range -2.1 % to 0.8 %), respectively, p > 0.05. Subjective observer evaluation of image appearance showed no visually detectable difference.
Simulated low dose images showed excellent agreement with the original scan data concerning image noise, CT density values, and subjective assessment of the visual appearance of the simulated images.
An authentic low-dose simulation from actual CT examinations opens up important opportunity with regard to staff education, protocol optimization and introduction of new reconstruction techniques.
[Show abstract][Hide abstract] ABSTRACT: PURPOSE
To investigate the improvement in diagnostic quality of iterative model reconstruction (IMR) algorithm for sub-mSv computed tomography angiography of the pulmonary arteries (CTA).
METHOD AND MATERIALS
Eighteen patients (single-center, IRB approved) were imaged on a Philips Brilliance iCT (Philips, Cleveland, OH) for visualization of the pulmonary arteries, 8 with and 10 without pulmonary artery embolism. All scans were performed at 120 kVp (average effective doses 4.34±1.99mSv). Acquisitions with reduced radiation exposure were simulated from the original CT data to 15% of the tube current, resulting in a sub-mSv average dose of 0.65±0.30mSv. Filtered backprojection (FBP) was used to reconstruct the original data (protocol A); sub-mSv data were reconstructed using FBP (protocol B) and IMR (protocol C). The performance of IMR was assessed with respect to the image quality metrics image noise and contrast-to-noise ratio (CNR) and with respect to effective dose of each protocol. Two blinded readers determined subjective image quality and assessed the detectability of pulmonary artery embolism, where ground truth was obtained from protocol A.
With IMR noise could be subjectively removed, while the image texture (look and feel) of these images differed from FBP reconstructions. Specifically, with IMR, the noise is significantly reduced by a factor up to 20 (B vs. C). This is reflected by an improvement in the contrast-to-noise ratio and improved image quality with a median image quality score of 3 (IMR, B) vs. 1 (FBP, C), p < 0.05. With respect to diagnostics the angiographic datasets protocol A and C were identical, while B was worse: To detect pulmonary artery embolism in IMR and FBP low dose images, the sensitivity was 100% for IMR and 62.5% for FBP while specificity was 100% for both protocols.
This simulation study indicates that by using IMR for reconstruction, pulmonary artery embolism can be detected accurately in scans with sub-mSv dose levels.
IMR has the potential to reduce patient dose and improve image quality in clinical day-to-day routine.
[Show abstract][Hide abstract] ABSTRACT: Evaluation of 15,000 computed tomography (CT) examinations to investigate if iterative reconstruction (IR) reduces sustainably radiation exposure.
Information from 15,000 CT examinations was collected, including all aspects of the exams such as scan parameter, patient information, and reconstruction instructions. The examinations were acquired between January 2010 and December 2012, while after 15 months a first generation IR algorithm was installed. To collect the necessary information from PACS, RIS, MPPS and structured reports a Dose Monitoring System was developed. To harvest all possible information an optical character recognition system was integrated, for example to collect information from the screenshot CT-dose report. The tool transfers all data to a database for further processing such as the calculation of effective dose and organ doses. To evaluate if IR provides a sustainable dose reduction, the effective dose values were statistically analyzed with respect to protocol type, diagnostic indication, and patient population.
IR has the potential to reduce radiation dose significantly. Before clinical introduction of IR the average effective dose was 10.1±7.8mSv and with IR 8.9±7.1mSv (p*=0.01). Especially in CTA, with the possibility to use kV reduction protocols, such as in aortic CTAs (before IR: average14.2±7.8mSv; median11.4mSv /with IR:average9.9±7.4mSv; median7.4mSv), or pulmonary CTAs (before IR: average9.7±6.2mSV; median7.7mSv /with IR: average6.4±4.7mSv; median4.8mSv) the dose reduction effect is significant(p*=0.01). On the contrary for unenhanced low-dose scans of the cranial (for example sinuses) the reduction is not significant (before IR:average6.6±5.8mSv; median3.9mSv/with IR:average6.0±3.1mSV; median3.2mSv).
The dose aspect remains a priority in CT research. Iterative reconstruction algorithms reduce sustainably and significantly radiation dose in the clinical routine. Our results illustrate that not only in studies with a limited number of patients but also in the clinical routine, IRs provide long-term dose saving.
[Show abstract][Hide abstract] ABSTRACT: The diagnostic use of computer tomography angiography (CTA) to identify perforating blood vessels for abdominal free-flap breast reconstruction was extended to estimate the three-dimensional (3D) preoperative flap volume and to compare it with the real intraoperative flap weights in order to (1) evaluate the accuracy of CTA-based 3D flap volume prediction, and (2) to analyze abdominal tissue estimation for required breast volume reconstruction.
Preoperative CTA was performed in 54 patients undergoing unilateral breast reconstruction with a free, deep, inferior epigastric artery perforator flap. 3D flap volumes ([Formula: see text]) based on CTA data were calculated and compared with the actual intraoperative flap weight (g). In addition, a breast volume to flap volume ratio was calculated to analyze whether the estimated 3D abdominal flap volume would match that of the breast to be removed.
40 CTA data sets (74.1 %) fulfilled the technical requirements for a reliable determination of flap volume. 3D CTA flap volume prediction showed no relevant differences to the actual flap weight (p = 0.44) and high correlations (r = 0.998, [Formula: see text]), allowing a prediction accuracy within 0.29 [Formula: see text] 3.0 % (range: from [Formula: see text]8.77 to 5.67 %) of the real flap weight. Significantly larger flap volumes were harvested compared with the actually required breast volumes ([Formula: see text]), leading to an average of 21 % of the remnant flap tissue potentially being discarded.
CTA-based 3D flap volume prediction provides accurate preoperative guidelines concerning the needed amount of abdominal tissue that can be harvested to achieve acceptable symmetry.
No preview · Article · Oct 2013 · International Journal of Computer Assisted Radiology and Surgery
[Show abstract][Hide abstract] ABSTRACT: To improve the detection of liver lesions in patients with hepatocellular carcinoma (HCC) via an iodine contrast enhancement tool.
Thirty-two patients with clinically proven HCCs underwent imaging with a three-phase protocol on a 256-slice MDCT. The contrast enhancement in the reconstructed slices was improved via a post-processing tool. Mean image noise was measured in four different regions: liver lesion, healthy liver, subcutaneous fat and bone. For each image set the image noise and contrast-to-noise ratio (CNR) were assessed. For subjective image assessment, four experienced radiologists evaluated the diagnostic quality.
While employing the post-processing algorithm, CNR between the liver lesion and healthy liver tissue improves significantly by a factor of 1.78 (CNRwithout vC = 2.30 ± 1.92/CNRwith vC = 4.11 ± 3.05) (P* = 0.01). All results could be achieved without a strengthening of artefacts; mean HU values of subcutaneous fat and bone did not significantly change. Subjective image analysis illustrated a significant improvement when employing post-processing for clinically relevant criteria such as diagnostic confidence.
With post-processing we see a significantly improved detection of arterial uptake in hepatic lesions compared with non-processed data. The improvement in CNR was confirmed by subjective image assessment for small lesions and for lesions with limited uptake.
• Enhancement with iodine-based contrast agents is an essential part of CT. • A new post-processing tool significantly improves the diagnostics of hepatocellular carcinoma. • It also improves detection of small lesions with limited iodine uptake.
Full-text · Article · Sep 2013 · European Radiology
[Show abstract][Hide abstract] ABSTRACT: Neoadjuvant chemotherapy for locally advanced gastric cancer leads to major histopathological response in less than 30 % of patients. Data on interim endoscopic response assessment do not exist. This exploratory prospective study evaluates early endoscopy after 50 % of the chemotherapy as predictor for later response and prognosis.
Forty-seven consecutive patients were included (45 resected; 33 R0 resections). All patients received baseline endoscopy and CT scans, after 50 % of their chemotherapy (EGD-1, CT-1) and after completion of chemotherapy (EGD-2, CT-2). Interim endoscopic response (EGD-1) was assessed after having received 50 % (6 weeks) of the planned 12 weeks of neoadjuvant chemotherapy. Post-chemotherapy response was clinically assessed by a combination of CT scan (CT-2) and endoscopy (EGD-2). Histopathological response was determined by a standardized scoring system (Becker criteria). Endoscopic response was defined as a reduction of >75 % of the tumor mass.
Twelve patients were responders at EGD-1 and 13 at EGD-2. Nine patients (19.1 %) were clinical responders and 7 patients (15.6 %) were histopathological responders after chemotherapy. Specificity, accuracy, and negative predictive value of the interim EGD-1 for subsequent histopathological response were 31/38 (82 %), 36/47 (76 %), and 31/33 (93 %); and for recurrence or death, 28/30 (93.3 %), 38/47 (80.9 %), and 28/35 (80.0 %). Response at EGD-1 was significantly associated with histopathological response (p = 0.010), survival (p < 0.001), and recurrence-free survival (p = 0.009).
Interim endoscopy after 6 weeks predicts response and prognosis. Therefore, tailoring treatment according to interim endoscopic assessment could be feasible, but the findings of this study should be validated in a larger patient cohort.
[Show abstract][Hide abstract] ABSTRACT: This study aimed to evaluate a novel segmentation software for automated liver volumetry and segmentation regarding segmentation speed and interobserver variability.
Computed tomographic scans of 20 patients without underlying liver disease and 10 patients with liver metastasis from colorectal cancer were analyzed by a novel segmentation software. Liver segmentation was performed after manual placement of specific landmarks into 9 segments according to the Couinaud model as well as into 4 segments, the latter being import for surgery planning. Time for segmentation was measured and the obtained segmental and total liver volumes between the different readers were compared calculating intraclass correlations (ICCs). Volumes of liver tumor burden were evaluated similarly.
Liver segmentation could be performed rapidly 3 minutes or less. Comparison of total liver volumes revealed a perfect ICC of greater than 0.997. Segmental liver volumes within the 9-part segmentation provided fair to moderate correlation for the left lobe and good to excellent correlations for the right lobe. When applying a 4-part segmentation relevant to clinical practice, strong to perfect agreement was observed. Similarly tumor volumes showed perfect ICC (>0.998).
Rapid determination of total and segmental liver volumes can be obtained using a novel segmentation software suitable for daily clinical practice.
No preview · Article · Jul 2013 · Journal of computer assisted tomography
[Show abstract][Hide abstract] ABSTRACT: The current common and dogmatic opinion is that whole-body computed tomography (WBCT) should not be performed in major trauma patients in shock. We aimed to assess whether WBCT during trauma-room treatment has any effect on the mortality of severely injured patients in shock.
In a retrospective multicenter cohort study involving 16719 adult blunt major trauma patients we compared the survival of patients who were in moderate, severe or no shock (systolic blood pressure 90-110,<90 or >110 mmHg) at hospital admission and who received WBCT during resuscitation to those who did not. Using data derived from the 2002-2009 version of TraumaRegister®, we determined the observed and predicted mortality and calculated the standardized mortality ratio (SMR) as well as logistic regressions.
9233 (55.2%) of the 16719 patients received WBCT. The mean injury severity score was 28.8±12.1. The overall mortality rate was 17.4% (SMR = 0.85, 95%CI 0.81-0.89) for patients with WBCT and 21.4% (SMR = 0.98, 95%CI 0.94-1.02) for those without WBCT (p<0.001). 4280 (25.6%) patients were in moderate shock and 1821 (10.9%) in severe shock. The mortality rate for patients in moderate shock with WBCT was 18.1% (SMR 0.85, CI95% 0.78-0.93) compared to 22.6% (SMR 1.03, CI95% 0.94-1.12) to those without WBCT (p<0.001, p = 0.002 for the SMRs). The mortality rate for patients in severe shock with WBCT was 42.1% (SMR 0.99, CI95% 0.92-1.06) compared to 54.9% (SMR 1.10, CI95% 1.02-1.16) to those without WBCT (p<0.001, p = 0.049 for the SMRs). Adjusted logistic regression analyses showed that WBCT is an independent predictor for survival that significantly increases the chance of survival in patients in moderate shock (OR = 0.73; 95%CI 0.60-0.90, p = 0.002) as well as in severe shock (OR = 0.67; 95%CI 0.52-0.88, p = 0.004). The number needed to scan related to survival was 35 for all patients, 26 for those in moderate shock and 20 for those in severe shock.
WBCT during trauma resuscitation significantly increased the survival in haemodynamically stable as well as in haemodynamically unstable major trauma patients. Thus, the application of WBCT in haemodynamically unstable severely injured patients seems to be safe, feasible and justified if performed quickly within a well-structured environment and by a well-organized trauma team.
[Show abstract][Hide abstract] ABSTRACT: BACKGROUND: The role of surgery for patients with metastatic esophagogastric adenocarcinoma (EGC) is not defined. The purpose of this study was to define selection criteria for patients who may benefit from resection following systemic chemotherapy. METHODS: From 1987 to 2007, 160 patients presenting with synchronous metastatic EGC (cT3/4 cNany cM0/1 finally pM1) were treated with chemotherapy followed by resection of the primary tumor and metastases. Clinical and histopathological data, site and number of metastases were analyzed. A prognostic score was established and validated in a second cohort from another academic center (n = 32). RESULTS: The median survival (MS) in cohort 1 was 13.6 months. Significant prognostic factors were grading (p = 0.046), ypT- (p = 0.001), ypN- (p = 0.011) and R-category (p = 0.015), lymphangiosis (p = 0.021), clinical (p = 0.004) and histopathological response (p = 0.006), but not localization or number of metastases. The addition of grading (G1/2:0 points; G3/4:1 points), clinical response (responder: 0; nonresponder: 1) and R-category (complete:0; R1:1; R2:2) defines two groups of patients with significantly different survival (p = 0.001) [low risk group (Score 0/1), n = 22: MS 35.3 months, 3-year-survival 47.6%); high risk group (Score 2/3/4) n = 126: MS 12.0 months, 3-year-survival 14.2%]. The score showed a strong trend in the validation cohort (p = 0.063) [low risk group (MS not reached, 3-year-survival 57.1%); high risk group (MS 19.9 months, 3-year-survival 6.7%)]. CONCLUSION: We observed long-term survival after resection of metastatic EGC. A simple clinical score may help to identify a subgroup of patients with a high chance of benefit from resection. However, the accurate estimation of achieving a complete resection, which is an integral element of the score, remains challenging.
No preview · Article · Jan 2013 · European journal of surgical oncology: the journal of the European Society of Surgical Oncology and the British Association of Surgical Oncology
[Show abstract][Hide abstract] ABSTRACT: Purpose: Patient-specific models of organs improve the planning of surgical resections and the intraoperative approach. In particular, pancreatic cancer surgery can benefit from a manufactured, patient-specific model. Therewith, the complex anatomical structures of and around the pancreas, especially the vessels, can be visualized. Methods: Several CT data sets showing the pancreas / pancreatic cancer were segmented. A special software system was used which allows segmentation of different tissues and export of three dimensional model data. A patient-specific model containing the pancreas, pancreatic cancer, arteries and veins was created for each case using rapid prototyping technologies. Different prototyping techniques were applied to show details and anatomical structures within the pancreas. Results: The patient-specific models of the pancreas / pancreatic cancer allowed better visualization of the three-dimensional anatomical structure as well as the perception of size and relations of dimensions and volume. Additionally, the vessels and the cancer within the pancreas were highlighted. Conclusions: The described procedure of creating patient-specific models of the pancreas promises advantages for pre-and intraoperative planning of pancreatic cancer surgery.