Validation and Reproducibility of Total Plaque Thickness in Carotid and Femoral Arteries Using Ultrasound.

OBJECTIVE: Plaque burden quantification by ultrasound improves cardiovascular (CV) risk prediction. However, measuring total plaque volume (TPV) with 3-D ultrasound, the current gold standard, is time consuming. In the present study we investigated the reproducibility of weighted total plaque thickness (wTPT) measured by 2-D ultrasound and its correlation with TPV. METHODS: Participants in an ongoing study of subclinical atherosclerosis and CV risk with no known atherosclerotic CV disease but who were found to have one or more plaques in carotid or femoral arteries by 2-D ultrasound were included. A total of 34 women and 26 men (mean age: 59.4 y, standard deviation: 8.7) underwent primary 2-D and 3-D ultrasound examinations. Participants then underwent a 2-D ultrasound examination by another radiologist blinded to the first radiologist's findings. Finally, all participants underwent a follow-up 2-D ultrasound by the first radiologist. RESULTS: Comparison of wTPT measurements between the 2-D studies revealed no significant difference (mean difference: 0.29 mm, 95% confidence interval [CI]: -0.48 to 1.17). Inter-observer and intra-observer analyses revealed intraclass correlation coefficients of 0.97 (95% CI: 0.96-0.98) and 1.0 (95% CI: 0.99-1.00), respectively. wTPT correlated with TPV (Spearman's ρ = 0.98, 95% CI: 0.96-0.99). Elapsed time for assessing wTPT was less than that for TPV (mean difference: 36.1 min, 95% CI: 26.0-46.3). CONCLUSION: wTPT had high reproducibility and correlation with TPV while requiring substantially less time. Future studies addressing the role of wTPT in predicting CV disease are needed.

Halved contrast medium dose in lower limb dual-energy computed tomography angiography-a randomized controlled trial.

OBJECTIVES: To compare vascular attenuation (VA) of an experimental half iodine-load dual-layer spectral detector CT (SDCT) lower limb computed tomography angiography (CTA) with control (standard iodine-load conventional 120-kilovolt peak (kVp) CTA). METHODS: Ethical approval and consent were obtained. In this parallel RCT, CTA examinations were randomized into experimental or control. Patients received 0.7 vs 1.4 mL/kg of iohexol 350 mgI/mL in the experimental- vs the control group. Two experimental virtual monoenergetic image (VMI) series at 40 and 50 kiloelectron volts (keV) were reconstructed. PRIMARY OUTCOME: VA. SECONDARY OUTCOMES: image noise (noise), contrast- and signal-to-noise ratio (CNR and SNR), and subjective examination quality (SEQ). RESULTS: A total of 106 vs 109 were randomized and 103 vs 108 were analyzed in the experimental vs, control groups, respectively. VA was higher on experimental 40 keV VMI than on control (p < 0.0001), but lower on 50 keV VMI (p < 0.022). Noise was higher on experimental 40 keV VMI than on control (p = 0.00022), but lower on 50 keV VMI (p = 0.0033). CNR and SNR were higher than the control on experimental 40 keV VMI (both p < 0.0001) and 50 keV (p = 0.0058 and p = 0.0023, respectively). SEQ was better on both VMIs in the experimental group than in the control (both p < 0.0001). CONCLUSIONS: Half iodine-load SDCT lower limb CTA at 40 keV achieved higher VA than the control. CNR, SNR, noise, and SEQ were higher at 40 keV, while 50 keV showed lower noise. CLINICAL RELEVANCE STATEMENT: Spectral detector CT with low-energy virtual monoenergetic imaging performed halved iodine contrast medium (CM) lower limb CT-angiography with sustained objective and subjective quality. This facilitates CM reduction, improvement of low CM-dosage examinations, and examination of patients with more severe kidney impairment. TRIAL REGISTRATION: Retrospectively registered 5 August 2022 at clinicaltrials.gov NCT05488899. KEY POINTS: • Contrast medium dosage may be halved in lower limb dual-energy CT angiography with virtual monoenergetic images at 40 keV, which may reduce contrast medium consumption in the face of a global shortage. • Experimental half-iodine-load dual-energy CT angiography at 40 keV showed higher vascular attenuation, contrast-to-noise ratio, signal-to-noise ratio, and subjective examination quality than standard iodine-load conventional. • Half-iodine dual-energy CT angiography protocols may allow us to reduce the risk of PC-AKI, examine patients with more severe kidney impairment, and provide higher quality examinations or salvage poor examinations when impaired kidney function limits the CM dose.

Automatic segmentation of human knee anatomy by a convolutional neural network applying a 3D MRI protocol.

BACKGROUND: To study deep learning segmentation of knee anatomy with 13 anatomical classes by using a magnetic resonance (MR) protocol of four three-dimensional (3D) pulse sequences, and evaluate possible clinical usefulness. METHODS: The sample selection involved 40 healthy right knee volumes from adult participants. Further, a recently injured single left knee with previous known ACL reconstruction was included as a test subject. The MR protocol consisted of the following 3D pulse sequences: T1 TSE, PD TSE, PD FS TSE, and Angio GE. The DenseVNet neural network was considered for these experiments. Five input combinations of sequences (i) T1, (ii) T1 and FS, (iii) PD and FS, (iv) T1, PD, and FS and (v) T1, PD, FS and Angio were trained using the deep learning algorithm. The Dice similarity coefficient (DSC), Jaccard index and Hausdorff were used to compare the performance of the networks. RESULTS: Combining all sequences collectively performed significantly better than other alternatives. The following DSCs (±standard deviation) were obtained for the test dataset: Bone medulla 0.997 (±0.002), PCL 0.973 (±0.015), ACL 0.964 (±0.022), muscle 0.998 (±0.001), cartilage 0.966 (±0.018), bone cortex 0.980 (±0.010), arteries 0.943 (±0.038), collateral ligaments 0.919 (± 0.069), tendons 0.982 (±0.005), meniscus 0.955 (±0.032), adipose tissue 0.998 (±0.001), veins 0.980 (±0.010) and nerves 0.921 (±0.071). The deep learning network correctly identified the anterior cruciate ligament (ACL) tear of the left knee, thus indicating a future aid to orthopaedics. CONCLUSIONS: The convolutional neural network proves highly capable of correctly labeling all anatomical structures of the knee joint when applied to 3D MR sequences. We have demonstrated that this deep learning model is capable of automatized segmentation that may give 3D models and discover pathology. Both useful for a preoperative evaluation.

Assessment of preoperative axillary nodal disease burden: breast MRI in locally advanced breast cancer before, during and after neoadjuvant endocrine therapy.

BACKGROUND: Axillary lymph node (LN) metastasis is one of the most important predictors of recurrence and survival in breast cancer, and accurate assessment of LN involvement is crucial. Determining extent of residual disease is key for surgical planning after neoadjuvant therapy. The aim of the study was to evaluate the diagnostic reliability of MRI for nodal disease in locally advanced breast cancer patients treated with neoadjuvant endocrine therapy (NET). METHODS: Thirty-three clinically node-positive locally advanced breast cancer patients who underwent NET and surgery were prospectively enrolled. Two radiologists reviewed the axillary nodes at 3 separate time points MRI examinations at baseline (before the first treatment regimen), interim (following at least 2 months after the first cycle and prior to crossing-over), and preoperative (after the final administration of therapy and immediately before surgery). According to LN status after surgery, imaging features and diagnostic performance were analyzed. RESULTS: All 33 patients had a target LN reduction, the greatest treatment benefit from week 8 to week 16. There was a positive correlation between the maximal diameter of the most suspicious LN measured by MRI and pathology during and after NET, being highest at therapy completion (r = 0.6, P ≤ .001). Mean and median differences of maximal diameter of the most suspicious LN were higher with MRI than with pathology. Seven of 33 patients demonstrated normal posttreatment MRI nodal status (yrN0). Of these 7 yrN0, 3 exhibited no metastasis on final pathology (ypN0), 2 ypN1 and 2 ypN2. Reciprocally, MRI diagnosed 3 cases of ypN0 as yrN + . Diffusion -weighted imaging (DWI) was the only axillary node characteristic significant when associated with pathological node status (χ2(4) = 8.118, P = .072). CONCLUSION: Performance characteristics of MRI were not completely sufficient to preclude surgical axillary staging. To our knowledge, this is the first study on MRI LN assessment following NET in locally advanced breast cancer, and further studies with larger sample sizes are required to consolidate the results of this preliminary study. TRIAL REGISTRATION: Institutional Review Board approval was obtained (this current manuscript is from a prospective, open-label, randomized single-center cohort substudy of the NEOLETEXE trial). NEOLETEXE, a phase 2 clinical trial, was registered on March 23rd, 2015 in the National trial database of Norway and approved by the Regional Ethical Committee of the South-Eastern Health Region in Norway; registration number: REK-SØ-84-2015 .

Low diagnostic accuracy and inter-observer agreement on CT and MRI in diagnosis of spinal fractures in multiple myeloma.

Skeletal disease is common in multiple myeloma. We investigated the inter-observer agreement and diagnostic accuracy of spinal fractures diagnosed by computer tomography (CT) and magnetic resonance imaging (MRI) from 12 myeloma patients. Two radiologists independently assessed the images. CT, MRI, and other images were combined to a gold standard. The inter-observer agreement was assessed with Cohen's kappa. Radiologist 1 diagnosed 20 malignant spinal fractures on CT and 26 on MRI, while radiologist 2 diagnosed 12 malignant spinal fractures on CT and 22 on MRI. In comparison the gold standard diagnosed 10 malignant spinal fractures. The sensitivity for malignant fractures varied from 0.5 to 1 for CT and MRI, and the specificity varied from 0.17 to 0.67. On MRI, the specificity for malignant spinal fractures was 0.17 for both radiologists. The inter-observer agreement for malignant spinal fractures on CT was -0.42 (Cohen's kappa) and -0.13 for MRI, while for osteoporotic fractures it was -0.24 for CT and 0.53 for MRI. We conclude that malignant spinal fractures were over-diagnosed on CT and MRI. The inter-observer agreement was extremely poor.

Explainable Artificial Intelligence for Human-Machine Interaction in Brain Tumor Localization.

Primary malignancies in adult brains are globally fatal. Computer vision, especially recent developments in artificial intelligence (AI), have created opportunities to automatically characterize and diagnose tumor lesions in the brain. AI approaches have provided scores of unprecedented accuracy in different image analysis tasks, including differentiating tumor-containing brains from healthy brains. AI models, however, perform as a black box, concealing the rational interpretations that are an essential step towards translating AI imaging tools into clinical routine. An explainable AI approach aims to visualize the high-level features of trained models or integrate into the training process. This study aims to evaluate the performance of selected deep-learning algorithms on localizing tumor lesions and distinguishing the lesion from healthy regions in magnetic resonance imaging contrasts. Despite a significant correlation between classification and lesion localization accuracy (R = 0.46, p = 0.005), the known AI algorithms, examined in this study, classify some tumor brains based on other non-relevant features. The results suggest that explainable AI approaches can develop an intuition for model interpretability and may play an important role in the performance evaluation of deep learning models. Developing explainable AI approaches will be an essential tool to improve human-machine interactions and assist in the selection of optimal training methods.

Correlation between MRI morphological response patterns and histopathological tumor regression after neoadjuvant endocrine therapy in locally advanced breast cancer: a randomized phase II trial.

PURPOSE: To correlate MRI morphological response patterns with histopathological tumor regression grading system based on tumor cellularity in locally advanced breast cancer (LABC)-treated neoadjuvant with third-generation aromatase inhibitors. METHODS: Fifty postmenopausal patients with ER-positive/HER-2-negative LABC treated with neoadjuvant letrozole and exemestane given sequentially in an intra-patient cross-over regimen for at least 4 months with MRI response monitoring at baseline as well as after at least 2 and 4 months on treatment. The MRI morphological response pattern was classified into 6 categories: 0/complete imaging response; I/concentric shrinkage; II/fragmentation; III/diffuse; IV/stable; and V/progressive. Histopathological tumor regression was assessed based on the recommendations from The Royal College of Pathologists regarding tumor cellularity. RESULTS: Following 2 and 4 months with therapy, the most common MRI pattern was pattern II (24/50 and 21/50, respectively). After 4 months on therapy, the most common histopathological tumor regression grade was grade 3 (21/50). After 4 months an increasing correlation is observed between MRI patterns and histopathology. The overall correlation, between the largest tumor diameter obtained from MRI and histopathology, was moderate and positive (r = 0.50, P-value = 2e-04). Among them, the correlation was highest in type IV (r = 0.53). CONCLUSION: The type II MRI pattern "fragmentation" was more frequent in the histopathological responder group; and types I and IV in the non-responder group. Type II pattern showed the best endocrine responsiveness and a relatively moderate correlation between sizes obtained from MRI and histology, whereas type IV pattern indicated endocrine resistance but the strongest correlation between MRI and histology.

Predictors of pneumothorax and chest drainage after percutaneous CT-guided lung biopsy: A prospective study.

OBJECTIVES: We present an analysis of predictors of pneumothorax, and pneumothorax requiring chest drainage after CT-guided lung biopsy, in one of the largest Scandinavian dataset presented. METHODS: We prospectively registered 875 biopsy procedures from 786 patients in one institution from January 27, 2012, to March 1, 2017, and recorded complications including pneumothorax with or without chest drainage, and multiple variables we assumed could be associated with complications. We performed multivariable logistic regression analysis to identify predictors of pneumothorax and pneumothorax requiring chest drainage. RESULTS: Of the biopsied lesions, 65% were malignant, 29% benign, and 6% inconclusive. Pneumothorax occurred in 39% of the procedures and chest drainage was performed in 10%. In multivariable analysis, significant predictors of pneumothorax were emphysema (OR 1.92), smaller lesion size (OR 0.83, per 1 cm increase in lesion size), lateral body position during procedure (OR 2.00), longer needle time (OR 1.09, per minute), repositioning of coaxial needle with new insertion through pleura (OR 3.04), insertion through interlobar fissure (OR 5.21), and shorter distance to pleura (OR 0.79, per 1 cm increase in distance). Predictors of chest drainage were emphysema (OR 4.01), lateral body position (OR 2.61), and needle insertion through interlobar fissure (OR 4.17). CONCLUSION: Predictors of pneumothorax were emphysema, lateral body position, needle insertion through interlobar fissure, repositioning of coaxial needle with new insertion through pleura, and shorter distance to pleura. The finding of lateral body position as a predictor of pneumothorax is not earlier described. Emphysema, lateral body position, and needle insertion through interlobar fissure were also predictors of chest drainage. KEY POINTS: • Pneumothorax is a frequent complication to CT-guided lung biopsy; a smaller fraction of these complications needs chest drainage. • Predictors for pneumothorax are emphysema, smaller lesion size, lateral body position, longer needle time, repositioning of coaxial needle with new insertion through pleura, needle insertion through the interlobar fissure, and shorter distance to pleura. • Predictors for requirement for chest drainage post CT-guided lung biopsy are emphysema, lateral body position, and needle insertion through the interlobar fissure.

Breast sarcoidosis: Clinical features, imaging, and histological findings.

Breast sarcoidosis is an extremely rare entity (about 1%). Conventional imaging significantly contributes to the detection of breast lesions, but it has been unable to establish a definite diagnosis. Histological examination should be mandatory, over imaging assessments, in order to confirm an early diagnosis and to avoid unjustified treatments. Malignancy should be excluded as a primary differential diagnosis. However, in the presence of granulomas, it is important to recognize other granulomatous disorders such as tuberculosis, Wegener's granulomatosis, or idiopathic granulomatous mastitis, since therapeutic strategies differ. This report clarifies the current clinical assessments and differential diagnosis of breast sarcoidosis.

Temporomandibular joint pain and associated magnetic resonance findings: a retrospective study with a control group.

BACKGROUND: To better understand and evaluate clinical usefulness of magnetic resonance imaging (MRI) in diagnosis and treatment of temporomandibular disorders (TMD), parameters for the evaluation are useful. PURPOSE: To assess a clinically suitable staging system for evaluation of MRI of the temporomandibular joint (TMJ) and correlate the findings with age and some clinical symptoms of the TMJ. MATERIAL AND METHODS: Retrospective analysis of 79 consecutive patients with clinical temporomandibular disorder or diagnosed inflammatory arthritis. Twenty-six healthy volunteers were included as controls. Existing data included TMJ pain, limited mouth opening (<30 mm) and corresponding MRI evaluations of the TMJs. RESULTS: The patients with clinical TMD complaints had statistically significantly more anterior disc displacement (ADD), disc deformation, caput flattening, surface destructions, osteophytes, and caput edema diagnosed by MRI compared to the controls. Among the arthritis patients, ADD, effusion, caput flattening, surface destructions, osteophytes, and caput edema were significantly more prevalent compared to the healthy volunteers. In the control group, disc deformation and presence of osteophytes significantly increased with age, and a borderline significance was found for ADD and surface destructions on the condylar head. No statistically significant associations were found between investigated clinical and MRI parameters. CONCLUSION: This study presents a clinically suitable staging system for comparable MRI findings in the TMJs. Our results indicate that some findings are due to age-related degenerative changes rather than pathological changes. Results also show that clinical findings such as pain and limited mouth opening may not be related to changes diagnosed by MRI.

Accuracy of breast MRI in patients receiving neoadjuvant endocrine therapy: comprehensive imaging analysis and correlation with clinical and pathological assessments.

PURPOSE: To assess the accuracy of magnetic resonance imaging (MRI) measurements in locally advanced oestrogen receptor-positive and human epidermal growth factor receptor 2-negative breast tumours before, during and after neoadjuvant endocrine treatment (NET) for evaluation of tumour response in comparison with clinical and pathological assessments. METHODS: This prospective study enrolled postmenopausal patients treated neoadjuvant with letrozole and exemestane given sequentially in an intra-patient cross-over regimen. Fifty-four patients were initially recruited, but only 35 fulfilled the inclusion criteria and confirmed to participate with a median age of 77. Tumours were scanned with MRI prior to treatment, during the eighth week of treatment and prior to surgery. Additionally, changes in longest diameter on clinical examination (CE) and tumour size at pathology were determined. Pre- and post-operative measurements of tumour size were compared in order to evaluate tumour response. RESULTS: The correlation between post-treatment MRI size and pathology was moderate and higher with a correlation coefficient (r) 0.64 compared to the correlation between CE and pathology r = 0.25. Post-treatment MRI and clinical results had a negligible bias towards underestimation of lesion size. Tumour size on MRI and CE had 0.82 cm and 0.52 cm lower mean size than tumour size measured by pathology, respectively. CONCLUSIONS: The higher correlation between measurements of residual disease obtained on MRI and those obtained with pathology validates the accuracy of imaging assessment during NET. MRI was found to be more accurate for estimating complete responses than clinical assessments and warrants further investigation in larger cohorts to validate this finding.

A national study of the causes, consequences and amelioration of adverse events in the use of MRI, CT, and conventional radiography in Norway.

BACKGROUND: Rapid technological developments, increased complexity, and increased demand have made patient safety a challenge in radiology. PURPOSE: To uncover the causes and consequences behind patient injury compensation claims in the use of MRI, CT, and conventional X-ray examinations, and to determine the system factors that need to be focused on in order to prevent these events. MATERIAL AND METHODS: This descriptive cross-sectional study uses data acquired from The Norwegian System of Patient Injury Compensation. A total of 240 cases from 2012-2016 were included. RESULTS: According to our study, the main factors contributing to patient injury compensation claims in radiology were false-negative findings (48.7%), misinterpretation (13.1%), and "satisfaction of search" (12%). Another source of error was routines (8.7%), mainly where the patient should have been (further) examined using another modality. Other causes were related to communication (7.6%), procedures (2.9%), technical factors (2.5%), organizational and management factors (1.5%), competence (0.7%), location of the lesion (0.7%), patient factors (0.7%), false-positive findings (0.4%), and work environment (0.4%). These events led to delayed diagnosis and/or treatment in the range of 0-3650 days. CONCLUSION: Errors of perception (false negative and "satisfaction of search") and cognitive errors (misinterpretation) were the main reasons behind patient injury compensation claims in radiology. We suggest that a combination of double-reading, specialization, increased collaboration between professionals, as well as a reduction of unnecessary examinations should be considered to reduce adverse events in radiology.

Radiological evaluation of primary breast sarcoidosis presenting as bilateral breast lesions.

Sarcoidosis is an idiopathic multisystemic inflammatory disease that may exceptionally involve the breast and can have imaging features suspicious for benign or malignant lesions. Biopsy should be required to distinguish between breast sarcoidosis and malignancy, because clinical, mammographic and sonographic findings can be ambiguous or inconclusive. This case discusses the radiological manifestations and the value of different diagnostic features, and names the most relevant differential diagnosis.

CT guided neurolytic blockade of the coeliac plexus in patients with advanced and intractably painful pancreatic cancer.

BACKGROUND AND AIMS: Pain caused by infiltrating pancreatic cancer is complex in nature and may therefore be difficult to treat. In addition to conventional analgesics, neurolytic blockade of the coeliac plexus is often recommended. However, different techniques are advocated, and procedures vary, and the results may therefore be difficult to compare. Therefore strong evidence for the effect of this treatment is still lacking, and more studies are encouraged. Our aim was to describe our technique and procedures using a Computer Tomograph (CT) guided procedure with a dorsal approach and present the results. METHODS: The procedures were performed in collaboration between a radiologist and an anaesthesiologist. All patients had advanced pancreatic cancer. The patients were placed in prone position on pillows, awake and monitored. Optimal placement of injection needles was guided by CT, and the radiologist injected a small dose of contrast as a control. When optimal needle position, the anaesthesiologist took over and completed the procedure. At first 40 mg methylprednisolone was injected to prevent inflammation. Thereafter a mixture of 99% ethanol diluted to 50% by ropivacaine 7.5 mg/mL to a total amount of 20-30 mL per needle was slowly injected. Repeated aspiration was performed during injection to avoid intravasal injection. Pain treatment and pain score was recorded and compared before and after the treatment. RESULTS: Eleven procedures in 10 patients were performed. Age 49-75, mean 59 years. Median rest life time was 36 days (11-140). Significant reduction of analgesics was observed 1 week after the procedure, and most patients also reported reduction of pain. No serious side effects were observed. CONCLUSIONS: CT guided neurolytic celiac plexus blockade is a safe and effective treatment for intractable pain caused by advanced pancreatic cancer. Not all patients experience a significant effect, but the side effects are minor, and the procedure should therefore be offered patients experiencing intractable cancer related pain.