Assessment of image quality and impact of deep learning-based software in non-contrast head CT scans.

In this retrospective study, we aimed to assess the objective and subjective image quality of different reconstruction techniques and a deep learning-based software on non-contrast head computed tomography (CT) images. In total, 152 adult head CT scans (77 female, 75 male; mean age 69.4 ± 18.3 years) obtained from three different CT scanners using different protocols between March and April 2021 were included. CT images were reconstructed using filtered-back projection (FBP), iterative reconstruction (IR), and post-processed using a deep learning-based algorithm (PS). Post-processing significantly reduced noise in FBP-reconstructed images (up to 15.4% reduction) depending on the protocol, leading to improvements in signal-to-noise ratio of up to 19.7%. However, when deep learning-based post-processing was applied to FBP images compared to IR alone, the differences were inconsistent and partly non-significant, which appeared to be protocol or site specific. Subjective assessments showed no significant overall improvement in image quality for all reconstructions and post-processing. Inter-rater reliability was low and preferences varied. Deep learning-based denoising software improved objective image quality compared to FBP in routine head CT. A significant difference compared to IR was observed for only one protocol. Subjective assessments did not indicate a significant clinical impact in terms of improved subjective image quality, likely due to the low noise levels in full-dose images.

Impact of AI-Based Post-Processing on Image Quality of Non-Contrast Computed Tomography of the Chest and Abdomen.

Non-contrast computed tomography (CT) is commonly used for the evaluation of various pathologies including pulmonary infections or urolithiasis but, especially in low-dose protocols, image quality is reduced. To improve this, deep learning-based post-processing approaches are being developed. Therefore, we aimed to compare the objective and subjective image quality of different reconstruction techniques and a deep learning-based software on non-contrast chest and low-dose abdominal CTs. In this retrospective study, non-contrast chest CTs of patients suspected of COVID-19 pneumonia and low-dose abdominal CTs suspected of urolithiasis were analysed. All images were reconstructed using filtered back-projection (FBP) and were post-processed using an artificial intelligence (AI)-based commercial software (PixelShine (PS)). Additional iterative reconstruction (IR) was performed for abdominal CTs. Objective and subjective image quality were evaluated. AI-based post-processing led to an overall significant noise reduction independent of the protocol (chest or abdomen) while maintaining image information (max. difference in SNR 2.59 ± 2.9 and CNR 15.92 ± 8.9, p < 0.001). Post-processing of FBP-reconstructed abdominal images was even superior to IR alone (max. difference in SNR 0.76 ± 0.5, p ≤ 0.001). Subjective assessments verified these results, partly suggesting benefits, especially in soft-tissue imaging (p < 0.001). All in all, the deep learning-based denoising-which was non-inferior to IR-offers an opportunity for image quality improvement especially in institutions using older scanners without IR availability. Further studies are necessary to evaluate potential effects on dose reduction benefits.

Deep learning-based scan range optimization can reduce radiation exposure in coronary CT angiography.

OBJECTIVES: Cardiac computed tomography (CT) is essential in diagnosing coronary heart disease. However, a disadvantage is the associated radiation exposure to the patient which depends in part on the scan range. This study aimed to develop a deep neural network to optimize the delimitation of scan ranges in CT localizers to reduce the radiation dose. METHODS: On a retrospective training cohort of 1507 CT localizers randomly selected from calcium scoring and angiography scans and acquired between 2010 and 2017, optimized scan ranges were delimited by two radiologists in consensus. A neural network was trained to reproduce the scan ranges and was tested on two randomly selected and independent validation cohorts: an internal cohort of 233 CT localizers (January 2018-June 2020) and an external cohort from a nearby hospital of 298 CT localizers (July 2020-December 2020). Localizers where a bypass surgery was visible were excluded. The effective radiation dose to the patient was simulated using a Monte Carlo simulation. Scan ranges of radiographers, radiologists, and the network were compared using an equivalence test; likewise, the reduction in effective dose was tested using a superior test. RESULTS: The network replicated the radiologists' scan ranges with a Dice score of 96.5 ± 0.02 (p < 0.001, indicating equivalence). The generated scan ranges resulted in an effective dose reduction of 10.0% (p = 0.002) in the internal cohort and 12.6% (p < 0.001) in the external cohort compared to the scan ranges delimited by radiographers in clinical routine. CONCLUSIONS: Automatic delimitation of the scan range can result in a radiation dose reduction to the patient. CLINICAL RELEVANCE STATEMENT: Fully automated delimitation of the scan range using a deep neural network enables a significant reduction in radiation exposure during CT coronary angiography compared to manual examination planning. It can also reduce the workload of the radiographers. KEY POINTS: • Scan range delimitation for coronary computed tomography angiography could be performed with high accuracy by a deep neural network. • Automated scan ranges showed a high agreement of 96.5% with the scan ranges of radiologists. • Using a Monte Carlo simulation, automated scan ranges reduced the effective dose to the patient by up to 12.6% (0.9 mSv) compared to the scan ranges of radiographers in clinical routine.

Determining body height and weight from thoracic and abdominal CT localizers in pediatric and young adult patients using deep learning.

In this retrospective study, we aimed to predict the body height and weight of pediatric patients using CT localizers, which are overview scans performed before the acquisition of the CT. We trained three commonly used networks (EfficientNetV2-S, ResNet-18, and ResNet-34) on a cohort of 1009 and 1111 CT localizers of pediatric patients with recorded body height and weight (between January 2013 and December 2019) and validated them in an additional cohort of 116 and 127 localizers (acquired in 2020). The best-performing model was then tested in an independent cohort of 203 and 225 CT localizers (acquired between January 2021 and March 2023). In addition, a cohort of 1401 and 1590 localizers from younger adults (acquired between January 2013 and December 2013) was added to the training set to determine if it could improve the overall accuracy. The EfficientNetV2-S using the additional adult cohort performed best with a mean absolute error of 5.58 ± 4.26 cm for height and 4.25 ± 4.28 kg for weight. The relative error was 4.12 ± 4.05% for height and 11.28 ± 12.05% for weight. Our study demonstrated that automated estimation of height and weight in pediatric patients from CT localizers can be performed.

Pediatric age estimation from thoracic and abdominal CT scout views using deep learning.

Age assessment is regularly used in clinical routine by pediatric endocrinologists to determine the physical development or maturity of children and adolescents. Our study investigates whether age assessment can be performed using CT scout views from thoracic and abdominal CT scans using a deep neural network. Hence, we retrospectively collected 1949 CT scout views from pediatric patients (acquired between January 2013 and December 2018) to train a deep neural network to predict the chronological age from CT scout views. The network was then evaluated on an independent test set of 502 CT scout views (acquired between January 2019 and July 2020). The trained model showed a mean absolute error of 1.18 ± 1.14 years on the test data set. A one-sided t-test to determine whether the difference between the predicted and actual chronological age was less than 2.0 years was statistically highly significant (p < 0.001). In addition, the correlation coefficient was very high (R = 0.97). In conclusion, the chronological age of pediatric patients can be assessed with high accuracy from CT scout views using a deep neural network.

Patterns of cardiac involvement in different muscular dystrophies assessed by magnetic resonance imaging.

BACKGROUND: In muscular dystrophies, it is not only skeletal muscles that can be affected, but also the myocardium. This cardiac involvement can represent a major cause of morbidity and mortality. PURPOSE: To investigate cardiac involvement in Duchenne (DMD), Becker (BMD), and limb girdle muscular dystrophy (LGMD) patients, and carriers of DMD/BMD by cardiac magnetic resonance (CMR) imaging and to search for differences in the pattern of cardiac involvement. MATERIAL AND METHODS: All patients with genetically or histologically proven DMD, BMD, and LGMD, or confirmed carriers of DMD/BMD who had undergone CMR at our clinic between January 2008 and November 2018 were retrospectively included and re-evaluated for regional and global left ventricular function, increased trabecularization, and late enhancement. RESULTS: A total of 26 DMD, 10 BMD, 11 LGMD, and seven DMD/BMD carriers were included. Only one carrier of DMD presented with normal CMR results; all other participants showed cardiac abnormalities. Regional wall motion abnormalities (RWMA; prevalence in LGMD patients: 55%) and late enhancement (prevalence in LGMD patients: 82%) were frequent. RWMA were accentuated basal inferolateral in DMD/BMD carriers, while in LGMD they were accentuated apical. In all groups late enhancement was located mainly subepicardial/midmyocardial with a basal inferolateral accentuation. Apart from the different RWMA distribution, no further group-specific differences were found. CONCLUSION: We found a high rate of cardiac involvement not only in DMD/BMD, but also in LGMD and DMD/BMD carriers with a different RWMA accentuation (apical in LGMD and basal inferolateral in DMD/BMD) as a single group-specific difference.

Diagnostic reference levels for chest computed tomography in children as a function of patient size.

BACKGROUND: Radiation exposures from computed tomography (CT) in children are inadequately studied. Diagnostic reference levels (DRLs) can help optimise radiation doses. OBJECTIVE: To determine local DRLs for paediatric chest CT performed mainly on modern dual-source, multi-slice CT scanners as a function of patient size. MATERIALS AND METHODS: Five hundred thirty-eight chest CT scans in 345 children under 15 years (y) of age (median age: 8 y, interquartile range [IQR]: 4-13 y) performed on four different CT scanners (38% on third-generation and 43% on second-generation dual-source CT) between November 2013 and December 2020 were retrospectively analysed. Examinations were grouped by water-equivalent diameter as a measure of patient size. DRLs for volume CT dose index (CTDIvol) and dose-length product (DLP) were determined for six different patient sizes and compared to national and European DRLs. RESULTS: The DRLs for CTDIvol and DLP are determined for each patient size group as a function of water-equivalent diameter as follows: (I) < 13 cm (n = 22; median: age 7 months): 0.4 mGy, 7 mGy·cm; (II) 13 cm to less than 17 cm (n = 151; median: age 3 y): 1.2 mGy, 25 mGy·cm; (III) 17 cm to less than 21 cm (n = 211; median: age 8 y): 1.7 mGy, 44 mGy·cm; (IV) 21 cm to less than 25 cm (n = 97; median: age 14 y): 3.0 mGy, 88 mGy·cm; (V) 25 cm to less than 29 cm (n = 42; median: age 14 y): 4.5 mGy, 135 mGy·cm; (VI) ≥ 29 cm (n = 15; median: age 14 y): 8.0 mGy, 241 mGy·cm. Compared with corresponding age and weight groups, our size-based DRLs for DLP are 54% to 71% lower than national and 23% to 85% lower than European DRLs. CONCLUSION: We developed DRLs for paediatric chest CT as a function of patient size with substantially lower values than national and European DRLs. Precise knowledge of size-based DRLs may assist other institutions in further dose optimisation in children.

Determining the anatomical site in knee radiographs using deep learning.

An important quality criterion for radiographs is the correct anatomical side marking. A deep neural network is evaluated to predict the correct anatomical side in radiographs of the knee acquired in anterior-posterior direction. In this retrospective study, a ResNet-34 network was trained on 2892 radiographs from 2540 patients to predict the anatomical side of knees in radiographs. The network was evaluated in an internal validation cohort of 932 radiographs of 816 patients and in an external validation cohort of 490 radiographs from 462 patients. The network showed an accuracy of 99.8% and 99.9% on the internal and external validation cohort, respectively, which is comparable to the accuracy of radiographers. Anatomical side in radiographs of the knee in anterior-posterior direction can be deduced from radiographs with high accuracy using deep learning.

Pediatric age estimation from radiographs of the knee using deep learning.

OBJECTIVES: Age estimation, especially in pediatric patients, is regularly used in different contexts ranging from forensic over medicolegal to clinical applications. A deep neural network has been developed to automatically estimate chronological age from knee radiographs in pediatric patients. METHODS: In this retrospective study, 3816 radiographs of the knee from pediatric patients from a German population (acquired between January 2008 and December 2018) were collected to train a neural network. The network was trained to predict chronological age from the knee radiographs and was evaluated on an independent validation cohort of 423 radiographs (acquired between January 2019 and December 2020) and on an external validation cohort of 197 radiographs. RESULTS: The model showed a mean absolute error of 0.86 ± 0.72 years and 0.9 ± 0.71 years on the internal and external validation cohorts, respectively. Separating age classes (< 14 years from ≥ 14 years and < 18 years from ≥ 18 years) showed AUCs between 0.94 and 0.98. CONCLUSIONS: The chronological age of pediatric patients can be estimated with good accuracy from radiographs of the knee using a deep neural network. KEY POINTS: • Radiographs of the knee can be used for age estimations in pediatric patients using a standard deep neural network. • The network showed a mean absolute error of 0.86 ± 0.72 years in an internal validation cohort and of 0.9 ± 0.71 years in an external validation cohort. • The network can be used to separate the age classes < 14 years from ≥ 14 years with an AUC of 0.97 and < 18 years from ≥ 18 years with an AUC of 0.94.

A spear to the heart-the accidental discovery of a giant cement embolism in the right heart: a case report.

BACKGROUND: The incidence of recognized cardiopulmonary cement embolism in the context of percutaneous vertebroplasty varies between 0% and 23%. In most cases, only small fragments embolize in the pulmonary arteries or the right heart cavities. The latter can cause potential harm by right ventricular perforation. CASE SUMMARY: A 57-year-old patient was admitted to our department of cardiology due to exertional dyspnoea and chest pain. In the course of further diagnostic tests, a huge cement embolus was accidentally discovered in the right ventricle. The unusual size and length and the threat of ventricular perforation make this case so unique. DISCUSSION: Large cement embolisms in kyphoplasty settings are possible and associated with the risk of fulminant complications.

IMAGE QUALITY STUDY OF RADIATION-REDUCED COMBINED CHEST AND ABDOMEN/PELVIS CT COMPARED WITH A STANDARD PROTOCOL.

The purpose of this study is to examine radiation doses and image quality of a low-dose (LD) protocol for chest and abdomen/pelvis (CAP) CT compared with a standard (STD) protocol. A total of 361 patients were included between October 2019 and April 2020; 104 patients with LD-protocol (100 kV, ref mAs 80 (chest)/145 (abdomen/pelvis)) and 257 patients with STD-protocol (100 kV, ref mAs 100 (chest)/180 (abdomen/pelvis)) at second-generation dual-source CT. Radiation doses for CTDIvol and DLP, and objective and subjective image qualities of 50 examinations from each group were evaluated. The LD-protocol applied significantly lower radiation doses compared with the STD-protocol (p < 0.001), achieving a dose reduction by 37% for the median DLP in chest, 19% in abdomen/pelvis and 22% in total. Median total DLP was 342 mGy·cm (LD) vs. 436 mGy·cm (STD). The LD-CAP CT protocol achieved a significant dose reduction far below national diagnostic reference levels, ensuring acceptable and good image quality.

Generating Virtual Short Tau Inversion Recovery (STIR) Images from T1- and T2-Weighted Images Using a Conditional Generative Adversarial Network in Spine Imaging.

Short tau inversion recovery (STIR) sequences are frequently used in magnetic resonance imaging (MRI) of the spine. However, STIR sequences require a significant amount of scanning time. The purpose of the present study was to generate virtual STIR (vSTIR) images from non-contrast, non-fat-suppressed T1- and T2-weighted images using a conditional generative adversarial network (cGAN). The training dataset comprised 612 studies from 514 patients, and the validation dataset comprised 141 studies from 133 patients. For validation, 100 original STIR and respective vSTIR series were presented to six senior radiologists (blinded for the STIR type) in independent A/B-testing sessions. Additionally, for 141 real or vSTIR sequences, the testers were required to produce a structured report of 15 different findings. In the A/B-test, most testers could not reliably identify the real STIR (mean error of tester 1-6: 41%; 44%; 58%; 48%; 39%; 45%). In the evaluation of the structured reports, vSTIR was equivalent to real STIR in 13 of 15 categories. In the category of the number of STIR hyperintense vertebral bodies (p = 0.08) and in the diagnosis of bone metastases (p = 0.055), the vSTIR was only slightly insignificantly equivalent. By virtually generating STIR images of diagnostic quality from T1- and T2-weighted images using a cGAN, one can shorten examination times and increase throughput.

Automatic Scan Range Delimitation in Chest CT Using Deep Learning.

PURPOSE: To develop and evaluate fully automatic scan range delimitation for chest CT by using deep learning. MATERIALS AND METHODS: For this retrospective study, scan ranges were annotated by two expert radiologists in consensus in 1149 (mean age, 65 years ± 16 [standard deviation]; 595 male patients) chest CT topograms acquired between March 2002 and February 2019 (350 with pleural effusion, 376 with atelectasis, 409 with neither, 14 with both). A conditional generative adversarial neural network was trained on 1000 randomly selected topograms to generate virtual scan range delimitations. On the remaining 149 topograms the software-based scan delimitations, scan lengths, and estimated radiation exposure were compared with those from clinical routine. For statistical analysis an equivalence test (two one-sided t tests) was used, with equivalence limits of 10 mm. RESULTS: The software-based scan ranges were similar to the radiologists' annotations, with a mean Dice score coefficient of 0.99 ± 0.01 and an absolute difference of 1.8 mm ± 1.9 and 3.3 mm ± 5.6 at the upper and lower boundary, respectively. An equivalence test indicated that both scan range delimitations were similar (P < .001). The software-based scan delimitation led to shorter scan ranges compared with those used in clinical routine (298.2 mm ± 32.7 vs 327.0 mm ± 42.0; P < .001), resulting in a lower simulated total radiation exposure (3.9 mSv ± 3.0 vs 4.2 mSv ± 3.3; P < .001). CONCLUSION: A conditional generative adversarial neural network was capable of automating scan range delimitation with high accuracy, potentially leading to shorter scan times and reduced radiation exposure.Keywords: Adults and Pediatrics, CT, Computer Applications-Detection/Diagnosis, Convolutional Neural Network (CNN), Lung, Radiation Safety, Segmentation, Supervised learning, Thorax © RSNA, 2021Supplemental material is available for this article.

Detecting the pulmonary trunk in CT scout views using deep learning.

For CT pulmonary angiograms, a scout view obtained in anterior-posterior projection is usually used for planning. For bolus tracking the radiographer manually locates a position in the CT scout view where the pulmonary trunk will be visible in an axial CT pre-scan. We automate the task of localizing the pulmonary trunk in CT scout views by deep learning methods. In 620 eligible CT scout views of 563 patients between March 2003 and February 2020 the region of the pulmonary trunk as well as an optimal slice ("reference standard") for bolus tracking, in which the pulmonary trunk was clearly visible, was annotated and used to train a U-Net predicting the region of the pulmonary trunk in the CT scout view. The networks' performance was subsequently evaluated on 239 CT scout views from 213 patients and was compared with the annotations of three radiographers. The network was able to localize the region of the pulmonary trunk with high accuracy, yielding an accuracy of 97.5% of localizing a slice in the region of the pulmonary trunk on the validation cohort. On average, the selected position had a distance of 5.3 mm from the reference standard. Compared to radiographers, using a non-inferiority test (one-sided, paired Wilcoxon rank-sum test) the network performed as well as each radiographer (P < 0.001 in all cases). Automated localization of the region of the pulmonary trunk in CT scout views is possible with high accuracy and is non-inferior to three radiographers.

Assessing the Role of Pericardial Fat as a Biomarker Connected to Coronary Calcification-A Deep Learning Based Approach Using Fully Automated Body Composition Analysis.

(1) Background: Epi- and Paracardial Adipose Tissue (EAT, PAT) have been spotlighted as important biomarkers in cardiological assessment in recent years. Since biomarker quantification is an increasingly important method for clinical use, we wanted to examine fully automated EAT and PAT quantification for possible use in cardiovascular risk stratification. (2) Methods: 966 patients with intermediate Framingham risk scores for Coronary Artery Disease referred for coronary calcium scans were included in clinical routine retrospectively. The Coronary Artery Calcium Score (CACS) was extracted and tissue quantification was performed by a deep learning network. (3) Results: The Computed Tomography (CT) segmentations predicted by the network indicated no significant correlation between EAT volume and EAT radiodensity when compared to Agatston score (r = 0.18, r = -0.09). CACS 0 category patients showed significantly lower levels of total EAT and PAT volumes and higher EAT and PAT densities than CACS 1-99 category patients (p < 0.01). Notably, this difference did not reach significance regarding EAT attenuation in male patients. Women older than 50 years, thus more likely to be postmenopausal, were shown to be at higher risk of coronary calcification (p < 0.01, OR = 4.59). CACS 1-99 vs. CACS ≥100 category patients remained below significance level (EAT volume: p = 0.087, EAT attenuation: p = 0.98). (4) Conclusions: Our study proves the feasibility of a fully automated adipose tissue analysis in clinical cardiac CT and confirms in a large clinical cohort that volume and attenuation of EAT and PAT are not correlated with CACS. Broadly available deep learning based rapid and reliable tissue quantification should thus be discussed as a method to assess this biomarker as a supplementary risk predictor in cardiac CT.

QRS-T angle in patients with Hypertrophic Cardiomyopathy - A comparison with Cardiac Magnetic Resonance Imaging.

Objective: We sought to investigate the possible association of a wide QRS-T angle on the surface EKG and myocardial fibrosis on contrast-enhanced cardiovascular magnetic (CMR) imaging in patients with hypertrophic cardiomyopathy (HCM). Background: Risk stratification in HCM patients is challenging. Late gadolinium enhancement (LGE) visualizes myocardial fibrosis with unique spatial resolution and is a strong and independent prognosticator in these patients. The QRS-T angle from the surface EKG is a promising prognostic marker in various cardiac pathologies. Methods: 70 patients with HCM obtained a standardized digital 12-lead EKG for the calculation of the QRS-T angle and underwent comprehensive CMR imaging for visualization of fibrosis by LGE. Patients were divided into groups according to the absence or presence of fibrosis on CMR. Results: 43 of 70 patients with HCM showed LGE on CMR following contrast administration. HCM patients with LGE (fibrosis) had wider QRS-T angles as compared to the patient group without LGE (100±54 vs. 46±31; <0.001). A QRS-T angle of 90 degrees or more was a strong predictor (OR 32.84, CI 4.08-264.47; p <0.001) of HCM with LGE. Conclusion: There is a strong association of a wide QRS-T angle and myocardial fibrosis in patients with HCM.

Bacterial contamination of automated MRI contrast injectors in clinical routine.

Aim: To quantify the frequency of bacterial contamination of the injected contrast agent/saline solution by an automated contrast injection system, and to evaluate whether usage of a novel tube system can reduce it. Methods: For bacterial contamination quantification two identical automated piston pump MRI contrast injectors were used in combination with a standard tube system. 3-5 ml of the contrast agent/saline solution was collected from the system prior to its connection to the patients' venous cannula in 104 consecutive patients. To test, whether a novel tube system reduces contamination, a tube system with shielded screw connections was used with the same contrast injectors and contrast agent/saline samples were collected in further 101 patients. Specimens were microbiologically analyzed. Frequencies of contamination were compared using Fisher exact test. Results: With the standard tube system, bacterial contamination was observed in 5.8% (6 out of 104 specimens). With the novel tube system, contamination was observed in 2.0% (2 out of 101 specimens, p=0.280). Staphylococcus epidermidis was the most common germ (5 cases) followed by Micrococcus luteus (2 cases) and Oligella ureolytica (1 case). Conclusion: Bacterial contaminations of MRI contrast injectors occurred in a non-negligible frequency especially with S. epidermidis. A trend towards reduced bacterial contamination was seen when a novel tube system with shielded screw connections was used.