Assistance from Automated ASPECTS Software Improves Reader Performance.

PURPOSE: To compare physicians' ability to read Alberta Stroke Program Early CT Score (ASPECTS) in patients with a large vessel occlusion within 6 hours of symptom onset when assisted by a machine learning-based automatic software tool, compared with their unassisted score. MATERIALS AND METHODS: 50 baseline CT scans selected from two prior studies (CRISP and GAMES-RP) were read by 3 experienced neuroradiologists who were provided access to a follow-up MRI. The average ASPECT score of these reads was used as the reference standard. Two additional neuroradiologists and 6 non-neuroradiologist readers then read the scans both with and without assistance from the software reader-augmentation program and reader improvement was determined. The primary hypothesis was that the agreement between typical readers and the consensus of 3 expert neuroradiologists would be improved with software augmented vs. unassisted reads. Agreement was based on the percentage of the individual ASPECT regions (50 cases, 10 regions each; N=500) where agreement was achieved. RESULTS: Typical non-neuroradiologist readers agreed with the expert consensus read in 72% of the 500 ASPECTS regions, evaluated without software assistance. The automated software alone agreed in 77%. When the typical readers read the scan in conjunction with the software, agreement improved to 78% (P<0.0001, test of proportions). The software program alone achieved correlations for total ASPECT scores that were similar to the expert readers who had access to the follow-up MRI scan to help enhance the quality of their reads. CONCLUSION: Typical readers had statistically significant improvement in their scoring of scans when the scan was read in conjunction with the automated software, achieving agreement rates that were comparable to neuroradiologists.

MDCT diagnosis of acute pulmonary embolism in the emergent setting.

To compare utilization of CT pulmonary angiogram (CTA) for diagnosis of pulmonary embolism (PE) in an emergency department (ED) with unstructured CT ordering to published rates of CT positivity in other EDs including those employing decision support and to identify pathways for improved utilization via collaboration with our pathology and ED colleagues. Two hundred seventeen patients over a 2.5-month time period who received a CTA for PE were reviewed with exclusion of pediatric patients and all sub-optimal, non-diagnostic, or equivocal scans; 21 were excluded leaving a sample of 196 patients. The rate of PE diagnosis and association of PE positivity with selected factors (D-dimer testing) was assessed. The percentage of cases positive for PE was 10.7 % (21/196) which is similar to the frequently published rate of 10 % in other emergency departments including settings that have studied the use of decision support. D-dimer testing was performed in 40.3 % of cases. In 29.6 % (58/196) of subjects, D-dimer was positive, 10.7 % (21/196) was negative, and 59.7 % (117/196) was not assessed. Prevalence of PE among D-dimer negative (0 %, 0/21) was lower versus positive D-dimer (12.1 %, 7/58) and unknown D-dimer patients (12.0 %, 14/117). D-dimer had 100 % (21/21) negative predictive value for the diagnosis of PE. While this suggests that D-dimer is useful to rule-out PE, due to the small number of patients with PE, the 95 % confidence intervals are wide and the post-test likelihood of PE could be as high as 14 %. The rate of CT positivity for PE in an ED with unstructured CT ordering is similar to that in other published series including as series in which decision support was used. While D-dimer had high negative predictive value, large studies are needed to confirm this high sensitivity and potentially increase its use in ruling out PE without CT and to reduce CT ordering particularly in patients with sufficiently low clinical pre-test probability of PE.

MRI of the petromastoid canal in children.

PURPOSE: To characterize the MRI features of the petromastoid canal in children with sensorineural hearing loss (SNHL) and in normal infants. MATERIALS AND METHODS: High resolution MRI examinations of 564 children who were evaluated for SNHL and brain MRI examinations of 112 infants who had normal studies were studied independently by two reviewers. RESULTS: In SNHL group, visibility of the PMC decreased for right and left PMC (P < 0.001). The width of the right PMC significantly decreased as age increased (P < 0.0001). There was no relation between abnormalities of membranous labyrinth and cochlear nerve and PMC visibility in children with SNHL (p > 0.05). In the normal group, the PMC visibility decreased with increasing age (right P = 0.0001, left P = 0.001). In the normal group also, as age increased, the PMC width decreased for both PMCs (right, P = 0.0006; left, P = 0.03). CONCLUSION: The PMC is more frequently visualized in young children. Its visibility and width are not associated with abnormalities of membranous labyrinth and cochlear nerves.

Is there an association between radiologist turnaround time of emergency department abdominal CT studies and radiologic report quality?

The aim of this study was to assess for an association between radiologists' turnaround time (TAT) and report quality for emergency department (ED) abdominopelvic CT examinations. Reports of 60 consecutive ED abdominopelvic CT studies from five abdominal radiologists (300 total reports) were included. An ED radiologist, abdominal radiologist, and ED physician independently evaluated satisfaction with report content (1-10 scale), satisfaction with report clarity (1-10 scale), and extent to which the report advanced the patient on a previously published clinical spectrum scale (1-5 scale). TAT (time between completion of imaging and completion of the final report) and report quality were compared between radiologists using unpaired t tests; associations between TAT and report quality scores for individual radiologists were assessed using Pearson's correlation coefficients. The five radiologists' mean TAT varied from 35 to 53 min. There were significant differences in report content in half of comparisons between radiologists by observer 1 (p ≤ 0.032) and in a minority of comparisons by observer 2 (p ≤ 0.047), in report clarity in majority of comparisons by observer 1 (p ≤ 0.031) and in a minority of comparisons by observer 2 (p ≤ 0.010), and in impact on patient care in a minority of comparisons for all observers (p ≤ 0.047). There were weak positive correlations between TAT and report content and clarity for three radiologists for observer 1 (r = 0.270-0.362) and no correlation between TAT and any report quality measure for remaining combinations of the five radiologists and three observers (r = -0.197 to +0.181). While both TAT and report quality vary between radiologists, these two factors were not associated for individual radiologists.

Imaging of necrotizing fasciitis.

Necrotizing Fasciitis (NF) is a rare, but highly lethal, rapidly progressive necrotic infection of the soft-tissue fascia. More common in immunocompromised patients, NF typically affects the extremities, perineal area, genital area (Fournier gangrene), and less commonly the torso. Although classically a clinical diagnosis, imaging is a powerful adjunct to facilitate early diagnosis in equivocal cases. The key features of NF of radiography, ultrasound, CT, and MRI are reviewed, as are the inherent limitations of NF diagnosis for each modality. With a high level of sensitivity, specificity, and soft tissue contrast, MRI remains the gold standard imaging method for evaluation of NF.

Recommendations for additional imaging on emergency department CT examinations: comparison of emergency- and organ-based subspecialty radiologists.

PURPOSE: Past studies have identified a high frequency of recommendations for additional imaging (RAI) for computed tomography (CT) studies performed in an emergency department (ED), thereby potentially contributing to increased imaging utilization and costs. The purpose of this study was to compare rates of RAI within the ED setting between ED-based and organ-based subspecialty radiologists. METHODS: We identified 600 ED CT studies, comprising 200 head, chest, and abdominal CT studies, split equally between cases reviewed by ED-based and organ-based radiologists. Frequency of RAI for the three examinations was compared between these subspecialty groups. RESULTS: Frequencies of RAI were 21.5 %, 13.5 %, and 5.5 % for CT examinations of the chest, abdomen, and brain, respectively. There was a significantly higher frequency of RAI for chest CT studies interpreted by chest radiologists than by ED radiologists (28.0 % vs. 15.0 %, respectively, p = 0.036), largely due to a higher rate of RAI for incidentally detected lung nodules and masses as well as other pulmonary parenchymal abnormalities by chest radiologists. There was no significant difference in RAI on brain or abdominal CT studies between the two groups (p = 0.426-1.0). However, on abdominal studies, only ED-based radiologists provided RAI for abnormalities of the bowel or uterus, while only organ-based radiologists provided RAI for pancreatic abnormalities. Only 25.6 % of RAI were subsequently performed at our institution. CONCLUSION: For chest CT studies performed at the authors' institution, differences in management of incidental pulmonary nodules contributed to a significantly higher frequency of RAI by chest radiologists than by ED-based radiologists. Further investigation of the impact of these differences on cost and patient outcomes is warranted.

A "High-Reliability Organization" Approach to Improve Trauma Imaging Performance.

PURPOSE: CT is the gold standard for triaging critically ill patients, including in trauma, and its use has increased over time. CT turnaround times (TATs) are frequently targeted for improvement. As opposed to linear reductionist processes such as Lean and Six Sigma, a high-reliability organization (HRO) approach focuses on culture and teams to enable rapid problem solving. The authors evaluated the HRO model to rapidly generate, trial, select, and implement improvement interventions to improve trauma patient CT performance. METHODS: All trauma patients presenting to a single institution's emergency department during a 5-month period were included. Project periods included 2-month preintervention, 1-month wash-in, and 2-month postintervention. Each initial trauma CT encounter during the wash-in and postintervention periods created job briefs in which the radiologist ensured that all involved had the pertinent clinical information and agreed on the imaging needed, thereby creating a shared mental model as well as an opportunity to raise concerns and provide ideas for improvement. RESULTS: A total of 447 patients were included: 145 preintervention, 68 wash-in, and 234 postintervention. The seven selected interventions were trauma text alert; scripted CT technologist-radiologist communication; modification of CT acquisition, processing, sending, and interpretation; and trauma mobile phones. The seven selected interventions reduced trauma patient CT median TATs by 60% (78 vs 31 min, P < .001), demonstrating the effectiveness of an HRO approach to improvement. CONCLUSIONS: An HRO-based approach was rapid in generating, trialing, selecting, and implementing improvement interventions, and the interventions were effective at substantially decreasing trauma patient CT TATs.

Facial fractures: The "bottom-up" approach.

Given the demands of a busy high-volume trauma center, trauma radiologists are expected to evaluate an enormous number of images covering a multitude of facial bones in a short period of time in severely traumatized patients. Therefore, a comprehensive checklist, search pattern, and practical approach become indispensable for evaluation. Moreover, fracture complex classification conveys abundant information in a succinct shorthand fashion, which can be a large asset in a busy high-volume trauma center: reliably helping clinicians communicate urgent findings, make early treatment decisions, and effectively plan surgical approaches. Traditionally, radiologists' approach the CT axial dataset in top-down fashion: navigating their descent craniocaudal. However, a bottom-up approach may be advantageous, especially when it comes to facial fracture complex classification. Four key anatomic landmarks of the face, when evaluated sequentially in bottom-up fashion, are favorable to rapid single-sweep facial fracture characterization: the mandible, the pterygoid plates, the zygoma, and the bony orbits. That is, when done in succession: 1. Clearing the mandible rules out a panfacial smash fracture. 2. Clearing the pterygoid plates effectively rules out a Le Fort I, II, and III fracture. 3. Clearing the zygoma effectively rules out a zygomaticomaxillary complex (ZMC) type fracture. 4. Clearing the bony orbits effectively rules out a naso-orbital-ethmoid (NOE) fracture. Following this process of exclusion and elimination; as one ascends through the face, fracture characterization becomes more manageable and straightforward. Besides identifying all of the fractures and using the appropriate classification system, the radiologist also needs to recognize key clinically relevant soft tissue injuries that may be associated with facial fractures and thus should address these in the report.

Performance of RAPID noncontrast CT stroke platform in large vessel occlusion and intracranial hemorrhage detection.

Background: Noncontrast CT (NCCT) is used to evaluate for intracerebral hemorrhage (ICH) and ischemia in acute ischemic stroke (AIS). Large vessel occlusions (LVOs) are a major cause of AIS, but challenging to detect on NCCT. Aims: The purpose of this study is to evaluate an AI software called RAPID NCCT Stroke (RAPID, iSchemaView, Menlo Park, CA) for ICH and LVO detection compared to expert readers. Methods: In this IRB approved retrospective, multicenter study, stand-alone performance of the software was assessed based on the consensus of 3 neuroradiologists and sensitivity and specificity were determined. The platform's performance was then compared to interpretation by readers comprised of eight general radiologists (GR) and three neuroradiologists (NR) in detecting ICH and hyperdense vessel sign (HVS) indicating LVO. Results: A total of 244 cases were included. Of the 244, 115 were LVOs and 26 were ICHs. One hundred three cases did not have LVO nor ICH. Stand-alone performance of the software demonstrated sensitivities and specificities of 96.2 and 99.5% for ICH and 63.5 and 95.1% for LVO detection. Compared to all 11 readers and eight GR readers only respectively, the software demonstrated superiority, achieving significantly higher sensitivities (63.5% versus 43.6%, p < 0.0001 and 63.5% versus 40.9%, p = 0.001). Conclusion: The RAPID NCCT Stroke platform demonstrates superior performance to radiologists for detecting LVO from a NCCT. Use of this software platform could lead to earlier LVO detection and expedited transfer of these patients to a thrombectomy capable center.

Histogram analysis of whole-lesion enhancement in differentiating clear cell from papillary subtype of renal cell cancer.

PURPOSE: To compare histogram analysis of voxel-based whole-lesion (WL) enhancement to qualitative assessment and region-of-interest (ROI)-based enhancement analysis in discriminating the renal cell cancer (RCC) subtype clear cell RCC (ccRCC) from papillary RCC (pRCC). MATERIALS AND METHODS: In this institutional review board-approved, HIPAA-compliant retrospective study, 73 patients underwent magnetic resonance (MR) imaging prior to surgery for RCC between January 2007 and January 2010. Three-dimensional fat-suppressed T1-weighted gradient-echo corticomedullary phase acquisitions, obtained before and after contrast agent administration, were transferred to a workstation at which automated registration followed by semiautomated segmentation of the RCC was performed. Percent enhancement was computed on a per-voxel basis: (SI(post) - SI(pre))/SI(pre) .100, where SI(pre) and SI(post) indicate signal intensity before and after contrast enhancement, respectively. The WL quantitative parameters of mean, median, and third quartile enhancement and histogram distribution parameters kurtosis and skewness were computed for each lesion. WL enhancement parameters were compared with ROI-based analysis and qualitative assessment with regards to diagnostic accuracy and interreader agreement in differentiating ccRCC from pRCC. RESULTS: There were 19 pRCCs and 55 ccRCCs at pathologic examination. ccRCC had significantly higher WL mean, median, and third quartile enhancement compared with pRCC and hade significantly lower kurtosis and skewness (all P < .001). Third quartile enhancement had the highest accuracy (94.6%; area under the curve, 0.980) in discriminating ccRCC from pRCC, which was significantly higher than the accuracy of qualitative assessment (86.0%; P = .04) but not significantly higher than that of ROI enhancement (89.2%; P = .52). WL enhancement parameters had higher interreader agreement (κ = 0.91-1.0) compared with ROI enhancement or qualitative assessment (κ = 0.83 and 0.7, respectively) in discriminating ccRCC from pRCC. CONCLUSION: WL enhancement histogram analysis is feasible and can potentially be used to differentiate ccRCC from pRCC with high accuracy. SUPPLEMENTAL MATERIAL: http://radiology.rsna.org/lookup/suppl/doi:10.1148/radiol.12111281/-/DC1.

Radiology Imaging Volume Changes During Discrete COVID-19 Pandemic Waves: Implications for the Delta Variant of Coronavirus and Future Pandemics.

PURPOSE: The aim of this study was to evaluate radiology imaging volumes at distinct time periods throughout the coronavirus disease 2019 (COVID-19) pandemic as a function of regional COVID-19 hospitalizations. METHODS: Radiology imaging volumes and statewide COVID-19 hospitalizations were collected, and four 28-day time periods throughout the COVID-19 pandemic of 2020 were analyzed: pre-COVID-19 in January, the "first wave" of COVID-19 hospitalizations in April, the "recovery" time period in the summer of 2020 with a relative nadir of COVID-19 hospitalizations, and the "third wave" of COVID-19 hospitalizations in November. Imaging studies were categorized as inpatient, outpatient, or emergency department on the basis of patient location at the time of acquisition. A Mann-Whitney U test was performed to compare daily imaging volumes during each discrete 28-day time period. RESULTS: Imaging volumes overall during the first wave of COVID-19 infections were 55% (11,098/20,011; P < .001) of pre-COVID-19 imaging volumes. Overall imaging volumes returned during the recovery time period to 99% (19,915/20,011; P = .725), and third-wave imaging volumes compared with the pre-COVID-19 period were significantly lower in the emergency department at 88.8% (7,951/8,955; P < .001), significantly higher for outpatients at 115.7% (8,818/7,621; P = .008), not significantly different for inpatients at 106% (3,650/3,435; P = .053), and overall unchanged when aggregated together at 102% (20,419/20,011; P = .629). CONCLUSIONS: Medical imaging rebounded after the first wave of COVID-19 hospitalizations, with relative stability of utilization over the ensuing phases of the pandemic. As widespread COVID-19 vaccination continues to occur, future surges in COVID-19 hospitalizations will likely have a negligible impact on imaging utilization.

Brainstorming Our Way to Improved Quality, Safety, and Resident Wellness in a Resource-Limited Emergency Department.

PURPOSE: To implement a more efficient standardized computed tomography (CT) protocoling system for emergency department (ED) patients in order to improve resident work satisfaction and wellness, and decrease lag time between ordering and protocoling a study. METHODS AND MATERIALS: Residents recorded lag times between time of order and time of protocol for 176 CT scans between November 2018 and January 2019. Pre- and postintervention resident surveys of 7 questions utilizing a 5-point Likert scale were used to assess the perceived efficiency and overall satisfaction with the protocolling system. CT technologists received a 2-step Standardized ED CT Protocoling Guidance Sheet for common indications and would consult the radiologist for any questions. RESULTS: Lag time between order and protocol averaged 17.8 minutes. Postintervention surveys demonstrated that residents were more satisfied with the new system (100% vs 6.1%), had an overall higher job satisfaction in the ED (91% vs 12.1%), thought the system was more efficient for a single study (100% vs 15.2%) and for an entire shift (100% vs 6.1%), volume of studies was maximized (91% vs 6.1%), and the workflow allowed residents to focus on interpreting studies and communicating findings (91% vs 3%). CONCLUSION: The implementation of an auto-protocolling system at our institution's ED took a system which was disruptive, inefficient, and unreliable, and eliminated both lag time and variation in time between ordering and protocoling, improving time to final report. It simultaneously decreased interruptions, allowing residents to focus on study interpretation, which increased resident work satisfaction, wellness, and educational benefit.

Differentiating shunt-responsive normal pressure hydrocephalus from Alzheimer disease and normal aging: pilot study using automated MRI brain tissue segmentation.

Evidence suggests that normal pressure hydrocephalus (NPH) is underdiagnosed in day to day radiologic practice, and differentiating NPH from cerebral atrophy due to other neurodegenerative diseases and normal aging remains a challenge. To better characterize NPH, we test the hypothesis that a prediction model based on automated MRI brain tissue segmentation can help differentiate shunt-responsive NPH patients from cerebral atrophy due to Alzheimer disease (AD) and normal aging. Brain segmentation into gray and white matter (GM, WM), and intracranial cerebrospinal fluid was derived from pre-shunt T1-weighted MRI of 15 shunt-responsive NPH patients (9 men, 72.6 ± 8.0 years-old), 17 AD patients (10 men, 72.1 ± 11.0 years-old) chosen as a representative of cerebral atrophy in this age group; and 18 matched healthy elderly controls (HC, 7 men, 69.7 ± 7.0 years old). A multinomial prediction model was generated based on brain tissue volume distributions. GM decrease of 33% relative to HC characterized AD (P < 0.005). High preoperative ventricular and near normal GM volumes characterized NPH. A multinomial regression model based on gender, GM and ventricular volume had 96.3% accuracy differentiating NPH from AD and HC. In conclusion, automated MRI brain tissue segmentation differentiates shunt-responsive NPH with high accuracy from atrophy due to AD and normal aging. This method may improve diagnosis of NPH and improve our ability to distinguish normal from pathologic aging.

Myelography: a primer.

Despite recent trends toward evaluation of back and neck pain with magnetic resonance imaging, myelography and postmyelography computed tomography continue to play an important role in the workup of many patients with spinal pathology. We present techniques for the safe and efficient performance of myelography, lumbar, and cervical puncture, which remain important skills to be mastered by radiology residents and fellows.