Refine
Document Type
- Article (14)
Language
- English (14)
Has Fulltext
- yes (14)
Is part of the Bibliography
- no (14)
Keywords
- Algorithms (2)
- Artificial intelligence (2)
- Bone density (2)
- Cancer (2)
- Multidetector computed tomography (2)
- Neoplasms (2)
- Osteoporosis (2)
- Spine (2)
- Tomography (x-ray computed) (2)
- Age determination by skeleton (1)
Institute
- Medizin (14)
Purpose: To determine whether machine learning assisted-texture analysis of multi-energy virtual monochromatic image (VMI) datasets from dual-energy CT (DECT) can be used to differentiate metastatic head and neck squamous cell carcinoma (HNSCC) lymph nodes from lymphoma, inflammatory, or normal lymph nodes.
Materials and methods: A retrospective evaluation of 412 cervical nodes from 5 different patient groups (50 patients in total) having undergone DECT of the neck between 2013 and 2015 was performed: (1) HNSCC with pathology proven metastatic adenopathy, (2) HNSCC with pathology proven benign nodes (controls for (1)), (3) lymphoma, (4) inflammatory, and (5) normal nodes (controls for (3) and (4)). Texture analysis was performed with TexRAD® software using two independent sets of contours to assess the impact of inter-rater variation. Two machine learning algorithms (Random Forests (RF) and Gradient Boosting Machine (GBM)) were used with independent training and testing sets and determination of accuracy, sensitivity, specificity, PPV, NPV, and AUC.
Results: In the independent testing (prediction) sets, the accuracy for distinguishing different groups of pathologic nodes or normal nodes ranged between 80 and 95%. The models generated using texture data extracted from the independent contour sets had substantial to almost perfect agreement. The accuracy, sensitivity, specificity, PPV, and NPV for correctly classifying a lymph node as malignant (i.e. metastatic HNSCC or lymphoma) versus benign were 92%, 91%, 93%, 95%, 87%, respectively.
Conclusion: Machine learning assisted-DECT texture analysis can help distinguish different nodal pathology and normal nodes with a high accuracy.
Background: Bone age (BA) assessment performed by artificial intelligence (AI) is of growing interest due to improved accuracy, precision and time efficiency in daily routine. The aim of this study was to investigate the accuracy and efficiency of a novel AI software version for automated BA assessment in comparison to the Greulich-Pyle method.
Methods: Radiographs of 514 patients were analysed in this retrospective study. Total BA was assessed independently by three blinded radiologists applying the GP method and by the AI software. Overall and gender-specific BA assessment results, as well as reading times of both approaches, were compared, while the reference BA was defined by two blinded experienced paediatric radiologists in consensus by application of the Greulich-Pyle method.
Results: Mean absolute deviation (MAD) and root mean square deviation (RSMD) were significantly lower between AI-derived BA and reference BA (MAD 0.34 years, RSMD 0.38 years) than between reader-calculated BA and reference BA (MAD 0.79 years, RSMD 0.89 years; p < 0.001). The correlation between AI-derived BA and reference BA (r = 0.99) was significantly higher than between reader-calculated BA and reference BA (r = 0.90; p < 0.001). No statistical difference was found in reader agreement and correlation analyses regarding gender (p = 0.241). Mean reading times were reduced by 87% using the AI system.
Conclusions: A novel AI software enabled highly accurate automated BA assessment. It may improve efficiency in clinical routine by reducing reading times without compromising the accuracy compared with the Greulich-Pyle method.
BACKGROUND: Evaluation of latest generation automated attenuation-based tube potential selection (ATPS) impact on image quality and radiation dose in contrast-enhanced chest-abdomen-pelvis computed tomography examinations for gynaecologic cancer staging.
METHODS: This IRB approved single-centre, observer-blinded retrospective study with a waiver for informed consent included a total of 100 patients with contrast-enhanced chest-abdomen-pelvis CT for gynaecologic cancer staging. All patients were examined with activated ATPS for adaption of tube voltage to body habitus. 50 patients were scanned on a third-generation dual-source CT (DSCT), and another 50 patients on a second-generation DSCT. Predefined image quality setting remained stable between both groups at 120 kV and a current of 210 Reference mAs. Subjective image quality assessment was performed by two blinded readers independently. Attenuation and image noise were measured in several anatomic structures. Signal-to-noise ratio (SNR) was calculated. For the evaluation of radiation exposure, CT dose index (CTDIvol) values were compared.
RESULTS: Diagnostic image quality was obtained in all patients. The median CTDIvol (6.1 mGy, range 3.9-22 mGy) was 40 % lower when using the algorithm compared with the previous ATCM protocol (median 10.2 mGy · cm, range 5.8-22.8 mGy). A reduction in potential to 90 kV occurred in 19 cases, a reduction to 100 kV in 23 patients and a reduction to 110 kV in 3 patients of our experimental cohort. These patients received significantly lower radiation exposure compared to the former used protocol.
CONCLUSION: Latest generation automated ATPS on third-generation DSCT provides good diagnostic image quality in chest-abdomen-pelvis CT while average radiation dose is reduced by 40 % compared to former ATPS protocol on second-generation DSCT.
Background: Evaluation of automated attenuation-based tube potential selection and its impact on image quality and radiation dose in CT (computed tomography) examinations for cancer staging.
Methods: A total of 110 (59 men, 51 women) patients underwent chest-abdomen-pelvis CT examinations; 55 using a fixed tube potential of 120 kV/current of 210 Reference mAs (using CareDose4D), and 55 using automated attenuation-based tube potential selection (CAREkV) also using a current of 210 Reference mAs. This evaluation was performed as a single-centre, observer-blinded retrospective analysis. Image quality was assessed by two readers in consensus. Attenuation, image noise, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) were measured or calculated for objective image evaluation. For the evaluation of radiation exposure, dose-length-product (DLP) values were compared and Size-specific dose estimates (SSDE) values were calculated.
Results: Diagnostic image quality was obtained from all patients. The median DLP (703.5 mGy · cm, range 390–2203 mGy · cm) was 7.9% lower when using the algorithm compared with the standard 120 kV protocol (median 756 mGy · cm, range 345–2267 mGy · cm). A reduction in potential to 100 kV occurred in 32 cases; therefore, these patients received significantly lower radiation exposure compared with the 120 kV protocol.
Conclusion: Automated attenuation-based tube potential selection produces good diagnostic image quality in chest-abdomen-pelvis CT and reduces the patient’s overall radiation dose by 7.9% compared to the standard 120 kV protocol.
Purpose: To determine the value of the 2D multiple-echo data image combination (MEDIC) sequence relative to the short-tau inversion recovery (STIR) sequence regarding the depiction of chondral lesions in the patellofemoral joint.
Materials and methods: During a period of 6 month patients with acute pain at the anterior aspect of the knee, joint effusion and suspected chondral lesion defect in the patellofemoral joint underwent MRI including axial MEDIC and STIR imaging. Patients with chondral lesions in the patellofemoral joint on at least one sequence were included. The MEDIC and STIR sequence were quantitatively compared regarding the patella cartilage-to-effusion contrast-to-noise ratio (CNR) and qualitatively regarding the depiction of chondral lesions independently scored by two radiologists on a 3-point scale (1 = not depicted; 2 = blurred depicted; 3 = clearly depicted) using the Wilcoxon-Mann-Whitney-Test. For the analysis of inter-observer agreement the Cohen's Weighted Kappa test was used.
Results: 30 of 58 patients (male: female, 21:9; age: 44 ± 12 yrs) revealed cartilage lesions (fissures, n = 5 including fibrillation; gaps, n = 15; delamination, n = 7; osteoarthritis, n = 3) and were included in this study. The STIR-sequence was significantly (p < 0.001) superior to the MEDIC-sequence regarding both, the patella cartilage-to-effusion CNR (mean CNR: 232 ± 61 vs. 40 ± 16) as well as the depiction of chondral lesion (mean score: 2.83 ± 0.4 vs. 1.75 ± 0.7) with substantial inter-observer agreement in the rating of both sequences (κ = 0.76–0.89).
Conclusion: For the depiction of chondral lesions in the patellofemoral joint, the axial STIR-sequence should be chosen in preference to the axial MEDIC-sequence.
Purpose: To investigate the diagnostic performance of noise-optimized virtual monoenergetic images (VMI+) in dual-energy CT (DECT) of portal vein thrombosis (PVT) compared to standard reconstructions. Method: This retrospective, single-center study included 107 patients (68 men; mean age, 60.1 ± 10.7 years) with malignant or cirrhotic liver disease and suspected PVT who had undergone contrast-enhanced portal-phase DECT of the abdomen. Linearly blended (M_0.6) and virtual monoenergetic images were calculated using both standard VMI and noise-optimized VMI+ algorithms in 20 keV increments from 40 to 100 keV. Quantitative measurements were performed in the portal vein for objective contrast-to-noise ratio (CNR) calculation. The image series showing the greatest CNR were further assessed for subjective image quality and diagnostic accuracy of PVT detection by two blinded radiologists. Results: PVT was present in 38 subjects. VMI+ reconstructions at 40 keV revealed the best objective image quality (CNR, 9.6 ± 4.3) compared to all other image reconstructions (p < 0.01). In the standard VMI series, CNR peaked at 60 keV (CNR, 4.7 ± 2.1). Qualitative image parameters showed the highest image quality rating scores for the 60 keV VMI+ series (median, 4) (p ≤ 0.03). The greatest diagnostic accuracy for the diagnosis of PVT was found for the 40 keV VMI+ series (sensitivity, 96%; specificity, 96%) compared to M_0.6 images (sensitivity, 87%; specificity, 92%), 60 keV VMI (sensitivity, 87%; specificity, 97%), and 60 keV VMI+ reconstructions (sensitivity, 92%; specificity, 97%) (p ≤ 0.01). Conclusions: Low-keV VMI+ reconstructions resulted in significantly improved diagnostic performance for the detection of PVT compared to other DECT reconstruction algorithms.
Objectives: To evaluate the predictive value of volumetric bone mineral density (BMD) assessment of the lumbar spine derived from phantomless dual-energy CT (DECT)-based volumetric material decomposition as an indicator for the 2-year occurrence risk of osteoporosis-associated fractures. Methods: L1 of 92 patients (46 men, 46 women; mean age, 64 years, range, 19–103 years) who had undergone third-generation dual-source DECT between 01/2016 and 12/2018 was retrospectively analyzed. For phantomless BMD assessment, dedicated DECT postprocessing software using material decomposition was applied. Digital files of all patients were sighted for 2 years following DECT to obtain the incidence of osteoporotic fractures. Receiver operating characteristic (ROC) analysis was used to calculate cut-off values and logistic regression models were used to determine associations of BMD, sex, and age with the occurrence of osteoporotic fractures. Results: A DECT-derived BMD cut-off of 93.70 mg/cm3 yielded 85.45% sensitivity and 89.19% specificity for the prediction to sustain one or more osteoporosis-associated fractures within 2 years after BMD measurement. DECT-derived BMD was significantly associated with the occurrence of new fractures (odds ratio of 0.8710, 95% CI, 0.091–0.9375, p < .001), indicating a protective effect of increased DECT-derived BMD values. Overall AUC was 0.9373 (CI, 0.867–0.977, p < .001) for the differentiation of patients who sustained osteoporosis-associated fractures within 2 years of BMD assessment. Conclusions: Retrospective DECT-based volumetric BMD assessment can accurately predict the 2-year risk to sustain an osteoporosis-associated fracture in at-risk patients without requiring a calibration phantom. Lower DECT-based BMD values are strongly associated with an increased risk to sustain fragility fractures.
Key Points: Dual-energy CT–derived assessment of bone mineral density can identify patients at risk to sustain osteoporosis-associated fractures with a sensitivity of 85.45% and a specificity of 89.19%. The DECT-derived BMD threshold for identification of at-risk patients lies above the American College of Radiology (ACR) QCT guidelines for the identification of osteoporosis (93.70 mg/cm 3 vs 80 mg/cm 3 ).
Objectives: To investigate the diagnostic accuracy of color-coded contrast-enhanced dual-energy CT virtual noncalcium (VNCa) reconstructions for the assessment of lumbar disk herniation compared to unenhanced VNCa imaging.
Methods: A total of 91 patients were retrospectively evaluated (65 years ± 16; 43 women) who had undergone third-generation dual-source dual-energy CT and 3.0-T MRI within an examination interval up to 3 weeks between November 2019 and December 2020. Eight weeks after assessing unenhanced color-coded VNCa reconstructions for the presence and degree of lumbar disk herniation, corresponding contrast-enhanced portal venous phase color-coded VNCa reconstructions were independently analyzed by the same five radiologists. MRI series were additionally analyzed by one highly experienced musculoskeletal radiologist and served as reference standard.
Results: MRI depicted 210 herniated lumbar disks in 91 patients. VNCa reconstructions derived from contrast-enhanced CT scans showed similar high overall sensitivity (93% vs 95%), specificity (94% vs 95%), and accuracy (94% vs 95%) for the assessment of lumbar disk herniation compared to unenhanced VNCa images (all p > .05). Interrater agreement in VNCa imaging was excellent for both, unenhanced and contrast-enhanced CT (κ = 0.84 vs κ = 0.86; p > .05). Moreover, ratings for diagnostic confidence, image quality, and noise differed not significantly between unenhanced and contrast-enhanced VNCa series (all p > .05).
Conclusions: Color-coded VNCa reconstructions derived from contrast-enhanced dual-energy CT yield similar diagnostic accuracy for the depiction of lumbar disk herniation compared to unenhanced VNCa imaging and therefore may improve opportunistic retrospective lumbar disk herniation assessment, particularly in case of staging CT examinations.
Key Points
• Color-coded dual-source dual-energy CT virtual noncalcium (VNCa) reconstructions derived from portal venous phase yield similar high diagnostic accuracy for the assessment of lumbar disk herniation compared to unenhanced VNCa CT series (94% vs 95%) with MRI serving as a standard of reference.
• Diagnostic confidence, image quality, and noise levels differ not significantly between unenhanced and contrast-enhanced portal venous phase VNCa dual-energy CT series.
• Dual-source dual-energy CT might have the potential to improve opportunistic retrospective lumbar disk herniation assessment in CT examinations performed for other indications through reconstruction of VNCa images.
Purpose: To evaluate the effect of reduced z-axis scan coverage on diagnostic performance and radiation dose of neck CT in patients with suspected cervical abscess.
Methods: Fifty-one patients with suspected cervical abscess were included and underwent contrast-enhanced neck CT on a 2nd or 3rd generation dual-source CT system. Image acquisition ranged from the aortic arch to the upper roof of the frontal sinuses (CTstd). Subsequently, series with reduced z-axis coverage (CTred) were reconstructed starting at the aortic arch up to the orbital floor. CTstd and CTred were independently assessed by two radiologists for the presence/absence of cervical abscesses and for incidental and alternative findings. In addition, diagnostic accuracy for the depiction of the cervical abscesses was calculated for both readers. Furthermore, DLP (dose-length-product), effective dose (ED) and organ doses were calculated and compared for CTred and CTstd, using a commercially available dose management platform.
Results: A total of 41 abscesses and 3 incidental/alternative findings were identified in CTstd. All abscesses and incidental/alternative findings could also be detected on CTred resulting in a sensitivity and specificity of 1.0 for both readers. DLP, ED and organ doses of the brain, the eye lenses, the red bone marrow and the salivary glands of CTred were significantly lower than for CTstd (p<0.001).
Conclusions: Reducing z-axis coverage of neck CT allows for a significant reduction of effective dose and organ doses at similar diagnostic performance as compared to CTstd.
Objective: To investigate the accuracy, efficiency and radiation dose of a novel laser navigation system (LNS) compared to those of free-handed punctures on computed tomography (CT).
Materials and methods: Sixty punctures were performed using a phantom body to compare accuracy, timely effort, and radiation dose of the conventional free-handed procedure to those of the LNS-guided method. An additional 20 LNS-guided interventions were performed on another phantom to confirm accuracy. Ten patients subsequently underwent LNS-guided punctures.
Results: The phantom 1-LNS group showed a target point accuracy of 4.0 ± 2.7 mm (freehand, 6.3 ± 3.6 mm; p = 0.008), entrance point accuracy of 0.8 ± 0.6 mm (freehand, 6.1 ± 4.7 mm), needle angulation accuracy of 1.3 ± 0.9° (freehand, 3.4 ± 3.1°; p < 0.001), intervention time of 7.03 ± 5.18 minutes (freehand, 8.38 ± 4.09 minutes; p = 0.006), and 4.2 ± 3.6 CT images (freehand, 7.9 ± 5.1; p < 0.001). These results show significant improvement in 60 punctures compared to freehand. The phantom 2-LNS group showed a target point accuracy of 3.6 ± 2.5 mm, entrance point accuracy of 1.4 ± 2.0 mm, needle angulation accuracy of 1.0 ± 1.2°, intervention time of 1.44 ± 0.22 minutes, and 3.4 ± 1.7 CT images. The LNS group achieved target point accuracy of 5.0 ± 1.2 mm, entrance point accuracy of 2.0 ± 1.5 mm, needle angulation accuracy of 1.5 ± 0.3°, intervention time of 12.08 ± 3.07 minutes, and used 5.7 ± 1.6 CT-images for the first experience with patients.
Conclusion: Laser navigation system improved accuracy, duration of intervention, and radiation dose of CT-guided interventions.