Effects of age and sex on the distribution and symmetry of lumbar spinal and neural foraminal stenosis: a natural language processing analysis of 43,255 lumbar MRI reports.

PURPOSE: The purpose of this study is to investigate relationship of patient age and sex to patterns of degenerative spinal stenosis on lumbar MRI (LMRI), rated as moderate or greater by a spine radiologist, using natural language processing (NLP) tools. METHODS: In this retrospective, IRB-approved study, LMRI reports acquired from 2007 to 2017 at a single institution were parsed with a rules-based natural language processing (NLP) algorithm for free-text descriptors of spinal canal stenosis (SCS) and neural foraminal stenosis (NFS) at each of six spinal levels (T12-S1) and categorized according to a 6-point grading scale. Demographic differences in the anatomic distribution of moderate (grade 3) or greater SCS and NFS were calculated by sex, and age and within-group differences for NFS symmetry (left vs. right) were calculated as odds ratios. RESULTS: Forty-three thousand two hundred fifty-five LMRI reports (34,947 unique patients, mean age = 54.7; sex = 54.9% women) interpreted by 152 radiologists were studied. Prevalence of significant SCS and NFS increased caudally from T12-L1 to L4-5 though less at L5-S1. NFS was asymmetrically more prevalent on the left at L2-L3 and L5-S1 (p < 0.001). SCS and NFS were more prevalent in men and SCS increased with age at all levels, but the effect size of age was largest at T12-L3. Younger patients (< 50 years) had relatively higher NFS prevalence at L5-S1. CONCLUSION: NLP can identify patterns of lumbar spine degeneration through analysis of a large corpus of radiologist interpretations. Demographic differences in stenosis prevalence shed light on the natural history and pathogenesis of LSDD.

The Composite Severity Score for Lumbar Spine MRI: a Metric of Cumulative Degenerative Disease Predicts Time Spent on Interpretation and Reporting.

Conventional measures of radiologist efficiency, such as the relative value unit, fail to account for variations in the complexity and difficulty of a given study. For lumbar spine MRI (LMRI), an ideal performance metric should account for the global severity of lumbar degenerative disease (LSDD) which may influence reporting time (RT), thereby affecting clinical productivity. This study aims to derive a global LSDD metric and estimate its effect on RT. A 10-year archive of LMRI reports comprising 13,388 exams was reviewed. Objective reporting timestamps were used to calculate RT. A natural language processing (NLP) tool was used to extract radiologist-assigned stenosis severity using a 6-point scale (0 = "normal" to 5 = "severe") at each lumbar level. The composite severity score (CSS) was calculated as the sum of each of 18 stenosis grades. The predictive values of CSS, sex, age, radiologist identity, and referring service on RT were examined with multiple regression models. The NLP tool accurately classified LSDD in 94.8% of cases in a validation set. The CSS increased with patient age and differed between men and women. In a univariable model, CSS was a significant predictor of mean RT (R2 = 0.38, p < 0.001) and independent predictor of mean RT (p < 0.001) controlling for patient sex, patient age, service location, and interpreting radiologist. The predictive strength of CSS was stronger for the low CSS range (CSS = 0-25, R2 = 0.83, p < 0.001) compared to higher CSS values (CSS > 25, R2 = 0.15, p = 0.05). Individual radiologist study volume was negatively correlated with mean RT (Pearson's R = - 0.35, p < 0.001). The composite severity score predicts radiologist reporting efficiency in LMRI, providing a quantitative measure of case complexity which may be useful for workflow planning and performance evaluation.

Wide Range of Clinical Outcomes in Patients with Gliomatosis Cerebri Growth Pattern: A Clinical, Radiographic, and Histopathologic Study.

BACKGROUND: The 2016 World Health Organization Classification of Central Nervous System Tumors categorizes gliomatosis cerebri growth pattern (GC) as a subgroup of diffuse infiltrating gliomas, defined by extent of brain involvement on magnetic resonance imaging (MRI). Clinical and radiographic features in GC patients are highly heterogeneous; however, prognosis has historically been considered poor. SUBJECTS, MATERIALS, AND METHODS: We performed a retrospective search for patients at our institution meeting radiographic criteria of primary, type I GC (defined as diffuse tumor infiltration without associated tumor mass and contrast enhancement on MRI) and analyzed their clinical, imaging, and histopathologic features. RESULTS: A total of 34 patients met radiographic criteria of primary, type I GC, and 33 had a confirmed histologic diagnosis of an infiltrating glial neoplasm. Age >47 years at diagnosis was associated with worse overall survival (OS) compared with age ≤47 years (hazard ratio [HR] 1.04, 95% confidence interval [CI] 1.01-1.07, p = .003). Patients with grade 2 tumors demonstrated a trend for improved OS compared with those with grade 3 tumors (HR 2.65, 95% CI 0.99-7.08, p = .051). Except for brainstem involvement, extent or location of radiographic involvement did not detectably affect clinical outcome. IDH mutation status identified a subgroup of GC patients with particularly long survival up to 25 years and was associated with longer time to progression (HR 4.81, 95% CI 0.99-23.47, p = .052). CONCLUSION: Patients with primary, type I GC do not uniformly carry a poor prognosis, even in the presence of widespread radiographic involvement. Consistent with other reports, IDH mutation status may identify patients with improved clinical outcome. Molecular characterization, rather than MRI features, may be most valuable for prognostication and management of GC patients. IMPLICATIONS FOR PRACTICE: Patients with gliomatosis cerebri growth pattern (GC) constitute a challenge to clinicians, given their wide range of clinical, histologic, and radiographic presentation, heterogeneous outcome patterns, and the lack of consensus on a standardized treatment approach. This study highlights that radiographic extent of disease-albeit category-defining-does not detectably influence survival and that IDH mutations may impact clinical outcome. Practicing oncologists should be aware that select GC patients may demonstrate exceptionally favorable survival times and prognosticate patients based on molecular markers, rather than imaging features alone.

Long-term outcomes of percutaneous lumbar facet synovial cyst rupture.

OBJECTIVE: To evaluate the therapeutic value, safety, and long-term clinical outcomes of percutaneous lumbar facet synovial cyst (LFSC) rupture. MATERIALS AND METHODS: Our study was institutional review board (IRB)-approved and Health Insurance Portability and Accountability Act (HIPAA)-compliant. The study group comprised 71 patients (44 women, mean age: 65 ± 17 years) who underwent CT- or fluoroscopy-guided percutaneous LFSC rupture. The technical success of LFSC rupture, the long-term clinical outcome, including repeat procedures or surgery, and imaging findings on MRI and CT were recorded. RESULTS: Seventy-nine LFSC ruptures were performed in 71 patients. CT guidance was used in 57 cases and fluoroscopy guidance in 22 cases. LFSC rupture was technically successful in 58 out of 79 cases (73 %). Mean injection volume for cyst rupture was 3.6 ± 2.2 mL and a combination of steroid and anesthetic was injected in all cases. Over a mean follow-up time of 44 months, 12 % of patients underwent repeat cyst rupture, and 46 % eventually underwent surgery, whereas the majority of patients (55 %) experienced symptomatic relief and did not undergo surgery. There was no significant association between a successful outcome and age, sex, level, or size of LFSC (p > 0.1). LFSCs with T2 hypointensity were more likely to require surgery (p = 0.02). There was one complication, a bacterial skin infection that completely resolved following antibiotic therapy. CONCLUSION: Percutaneous LFSC rupture is an effective and safe nonsurgical treatment option for LFSC. More than half of treated patients were able to avoid subsequent surgery. Therefore, percutaneous LFSC rupture should be considered before surgical intervention.

Rate of Contrast Extravasation on Computed Tomographic Angiography Predicts Hematoma Expansion and Mortality in Primary Intracerebral Hemorrhage.

BACKGROUND AND PURPOSE: In primary intracerebral hemorrhage, the presence of contrast extravasation after computed tomographic angiography (CTA), termed the spot sign, predicts hematoma expansion and mortality. Because the biological underpinnings of the spot sign are not fully understood, we investigated whether the rate of contrast extravasation, which may reflect the rate of bleeding, predicts expansion and mortality beyond the simple presence of the spot sign. METHODS: Consecutive intracerebral hemorrhage patients with first-pass CTA followed by a 90-second delayed postcontrast CT (delayed CTA) were included. CTAs were reviewed for spot sign presence by 2 blinded readers. Spot sign volumes on first-pass and delayed CTA and intracerebral hemorrhage volumes were measured using semiautomated software. Extravasation rates were calculated and tested for association with hematoma expansion and mortality using uni- and multivariable logistic regressions. RESULTS: One hundred and sixty-two patients were included, 48 (30%) of whom had ≥1 spot sign. Median spot sign volume was 0.04 mL on first-pass CTA and 0.4 mL on delayed CTA. Median extravasation rate was 0.23 mL/min overall and 0.30 mL/min among expanders versus 0.07 mL/min in nonexpanders. Extravasation rates were also significantly higher in patients who died in hospital: 0.27 mL/min versus 0.04 mL/min. In multivariable analysis, the extravasation rate was independently associated with in-hospital mortality (odds ratio, 1.09 [95% confidence interval, 1.04-1.18], P=0.004), 90-day mortality (odds ratio, 1.15 [95% confidence interval, 1.08-1.27]; P=0.0004), and hematoma expansion (odds ratio, 1.03 [95% confidence interval, 1.01-1.08]; P=0.047). CONCLUSIONS: Contrast extravasation rate, or spot sign growth, further refines the ability to predict hematoma expansion and mortality. Our results support the hypothesis that the spot sign directly measures active bleeding in acute intracerebral hemorrhage.

Spot sign on 90-second delayed computed tomography angiography improves sensitivity for hematoma expansion and mortality: prospective study.

BACKGROUND AND PURPOSE: The computed tomography angiography (CTA) spot sign is a validated biomarker for poor outcome and hematoma expansion in intracerebral hemorrhage. The spot sign has proven to be a dynamic entity, with multimodal imaging proving to be of additional value. We investigated whether the addition of a 90-second delayed CTA acquisition would capture additional intracerebral hemorrhage patients with the spot sign and increase the sensitivity of the spot sign. METHODS: We prospectively enrolled consecutive intracerebral hemorrhage patients undergoing first pass and 90-second delayed CTA for 18 months at a single academic center. Univariate and multivariate logistic regression were performed to assess clinical and neuroimaging covariates for relationship with hematoma expansion and mortality. RESULTS: Sensitivity of the spot sign for hematoma expansion on first pass CTA was 55%, which increased to 64% if the spot sign was present on either CTA acquisition. In multivariate analysis the spot sign presence was associated with significant hematoma expansion: odds ratio, 17.7 (95% confidence interval, 3.7-84.2; P=0.0004), 8.3 (95% confidence interval, 2.0-33.4; P=0.004), and 12.0 (95% confidence interval, 2.9-50.5; P=0.0008) if present on first pass, delayed, or either CTA acquisition, respectively. Spot sign presence on either acquisitions was also significant for mortality. CONCLUSIONS: We demonstrate improved sensitivity for predicting hematoma expansion and poor outcome by adding a 90-second delayed CTA, which may enhance selection of patients who may benefit from hemostatic therapy.

Emergency department CT screening of patients with nontraumatic neurological symptoms referred to the posterior fossa: comparison of thin versus thick slice images.

Evaluation of the posterior fossa (PF) on 5-mm-thick helical CT images (current default) has improved diagnostic accuracy compared to 5-mm sequential CT images; however, 5-mm-thick images may not be ideal for PF pathology due to volume averaging of rapid changes in anatomy in the Z-direction. Therefore, we sought to determine if routine review of 1.25-mm-thin helical CT images has superior accuracy in screening for nontraumatic PF pathology. MRI proof of diagnosis was obtained within 6 h of helical CT acquisition for 90 consecutive ED patients with, and 88 without, posterior fossa lesions. Helical CT images were post-processed at 1.25 and 5-mm-axial slice thickness. Two neuroradiologists blinded to the clinical/MRI findings reviewed both image sets. Interobserver agreement and accuracy were rated using Kappa statistics and ROC analysis, respectively. Of the 90/178 (51 %) who were MR positive, 60/90 (66 %) had stroke and 30/90 (33 %) had other etiologies. There was excellent interobserver agreement (κ > 0.97) for both thick and thin slice assessments. The accuracy, sensitivity, and specificity for 1.25-mm images were 65, 44, and 84 %, respectively, and for 5-mm images were 67, 45, and 85 %, respectively. The diagnostic accuracy was not significantly different (p > 0.5). In this cohort of patients with nontraumatic neurological symptoms referred to the posterior fossa, 1.25-mm-thin slice CT reformatted images do not have superior accuracy compared to 5-mm-thick images. This information has implications on optimizing resource utilizations and efficiency in a busy emergency room. Review of 1.25-mm-thin images may help diagnostic accuracy only when review of 5-mm-thick images as current default is inconclusive.

Not without Context-A Multiple Methods Study on Evaluation and Correction of Automated Brain Tumor Segmentations by Experts.

RATIONALE AND OBJECTIVES: Brain tumor segmentations are integral to the clinical management of patients with glioblastoma, the deadliest primary brain tumor in adults. The manual delineation of tumors is time-consuming and highly provider-dependent. These two problems must be addressed by introducing automated, deep-learning-based segmentation tools. This study aimed to identify criteria experts use to evaluate the quality of automatically generated segmentations and their thought processes as they correct them. MATERIALS AND METHODS: Multiple methods were used to develop a detailed understanding of the complex factors that shape experts' perception of segmentation quality and their thought processes in correcting proposed segmentations. Data from a questionnaire and semistructured interview with neuro-oncologists and neuroradiologists were collected between August and December 2021 and analyzed using a combined deductive and inductive approach. RESULTS: Brain tumors are highly complex and ambiguous segmentation targets. Therefore, physicians rely heavily on the given context related to the patient and clinical context in evaluating the quality and need to correct brain tumor segmentation. Most importantly, the intended clinical application determines the segmentation quality criteria and editing decisions. Physicians' personal beliefs and preferences about the capabilities of AI algorithms and whether questionable areas should not be included are additional criteria influencing the perception of segmentation quality and appearance of an edited segmentation. CONCLUSION: Our findings on experts' perceptions of segmentation quality will allow the design of improved frameworks for expert-centered evaluation of brain tumor segmentation models. In particular, the knowledge presented here can inspire the development of brain tumor-specific metrics for segmentation model training and evaluation.

Clot length distribution and predictors in anterior circulation stroke: implications for intra-arterial therapy.

BACKGROUND AND PURPOSE: Thin-section noncontrast computed tomography images can be used to measure hyperdense clot length in acute ischemic stroke. Clots≥8 mm have a very low probability of intravenous tissue-type plasminogen activator recanalization and hence may benefit from a bridging intra-arterial approach. To understand the prevalence of such clots, we sought to determine the distribution and predictors of clot lengths in consecutive anterior circulation proximal artery occlusions. METHODS: Of 623 consecutive patients with acute ischemic stroke, 53 met inclusion criteria: presentation<8 hours from onset; intracranial internal carotid artery-terminus or proximal-middle cerebral artery occlusion; admission thin-slice noncontrast computed tomography (≤2.5 mm); and no intravenous tissue-type plasminogen activator pretreatment. For each patient, hyperdense clot length was measured and recorded along with additional relevant imaging and clinical data. RESULTS: Mean age was 70 years, and mean time to computed tomography was 213 minutes. Median baseline National Institutes of Health Stroke Scale was 16.5. Occlusions were located in the internal carotid artery-terminus (34% [18 of 53]), middle cerebral artery M1 (49% [26 of 53]) and M2 segments (17% [9 of 53]). Hyperdense thrombus was visible in 96%, with mean and median clot lengths (mm) of 18.5 (±14.2) and 16.1 (7.6-25.2), respectively. Occlusion location was the strongest predictor of clot length (multivariate, P=0.02). Clot length was ≥8 mm in 94%, 73%, and 22% of internal carotid artery-terminus, M1, and M2 occlusions, respectively. CONCLUSIONS: The majority of anterior circulation proximal occlusions are ≥8 mm long, helping to explain the low published rates of intravenous tissue-type plasminogen activator recanalization. Internal carotid artery-terminus occlusion is an excellent marker for clot length≥8 mm; vessel-imaging status alone may be sufficient. Thin-section noncontrast computed tomography seems useful for patients with middle cerebral artery occlusion because of the wide variability of clot lengths.

Virtual monochromatic reconstruction of dual-energy unenhanced head CT at 65-75 keV maximizes image quality compared with conventional polychromatic CT.

PURPOSE: To determine the virtual monochromatic imaging (VMI) energy levels that maximize brain parenchymal image quality in dual-energy unenhanced head computed tomography (CT) and to assess the improvement with this technique compared with conventional polychromatic scanning. MATERIALS AND METHODS: Institutional review board approval was obtained with no informed consent required for this HIPAA-compliant retrospective analysis. Twenty-five consecutive unenhanced head CT scans were acquired with a 64-section dual-energy scanner with fast tube voltage switching (80-140 kVp). Scans were retrospectively reconstructed at VMI energy levels from 40 to 140 keV in 5-keV increments and were analyzed by using four quality indexes: gray matter (GM) signal-to-noise ratio (SNR), white matter (WM) SNR, GM-WM contrast-to-noise ratio (CNR), and posterior fossa artifact index (PFAI). Optimal mean values for each parameter were compared with those from 50 consecutive scans obtained with the same scanner in 120-kVp single-energy mode. Repeated-measures analysis of variance and Dunnett post hoc t test were then used to determine significance. RESULTS: Maximal GM SNR, WM SNR, and GM-WM CNR values were observed at 65 keV, and minimal PFAI was observed at 75 keV. These values were significantly better than those of conventional polychromatic CT (P < .01); quality index improvement ratios (corrected for radiation dose) ranged from 17% to 50%. CONCLUSION: Virtual monochromatic reconstruction of dual-energy unenhanced head CT scans at 65-75 keV (optimal energy levels) maximizes image quality compared with scans obtained with conventional polychromatic CT.

Using a web-based image quality assurance reporting system to improve image quality.

OBJECTIVE: The purpose of this study is to show the impact of a web-based image quality assurance reporting system on the rates of three common image quality errors at our institution. MATERIALS AND METHODS: A web-based image quality assurance reporting system was developed and used beginning in April 2009. Image quality endpoints were assessed immediately before deployment (period 1), approximately 18 months after deployment of a prototype reporting system (period 2), and approximately 12 months after deployment of a subsequent upgraded department-wide reporting system (period 3). A total of 3067 axillary shoulder radiographs were reviewed for correct orientation, 355 shoulder CT scans were reviewed for correct reformatting of coronal and sagittal images, and 346 sacral MRI scans were reviewed for correct acquisition plane of axial images. Error rates for each review period were calculated and compared using the Fisher exact test. RESULTS: Error rates of axillary shoulder radiograph orientation were 35.9%, 7.2%, and 10.0%, respectively, for the three review periods. The decrease in error rate between periods 1 and 2 was statistically significant (p < 0.0001). Error rates of shoulder CT reformats were 9.8%, 2.7%, and 5.8%, respectively, for the three review periods. The decrease in error rate between periods 1 and 2 was statistically significant (p = 0.03). Error rates for sacral MRI axial sequences were 96.5%, 32.5%, and 3.4%, respectively, for the three review periods. The decrease in error rates between periods 1 and 2 and between periods 2 and 3 was statistically significant (p < 0.0001). CONCLUSION: A web-based system for reporting image quality errors may be effective for improving image quality.

Role of cardiac and extracranial vascular CT in the evaluation/management of cerebral ischemia and stroke.

Collectively, cardiac and large artery sources are responsible for the largest proportion of acute ischemic stroke. Technological advancements in computed tomography (CT) continue to improve evaluation of these patients. The literature was reviewed for the potential role and impact of these innovations in evaluation and management of these patients. In conclusion, incorporation of early cardiac and extracranial vascular CT angiography (CTA) in evaluation of patients with acute ischemic stroke may potentially improve patient management and outcome, while decreasing cost.

Head CT deep learning model is highly accurate for early infarct estimation.

Non-contrast head CT (NCCT) is extremely insensitive for early (< 3-6 h) acute infarct identification. We developed a deep learning model that detects and delineates suspected early acute infarcts on NCCT, using diffusion MRI as ground truth (3566 NCCT/MRI training patient pairs). The model substantially outperformed 3 expert neuroradiologists on a test set of 150 CT scans of patients who were potential candidates for thrombectomy (60 stroke-negative, 90 stroke-positive middle cerebral artery territory only infarcts), with sensitivity 96% (specificity 72%) for the model versus 61-66% (specificity 90-92%) for the experts; model infarct volume estimates also strongly correlated with those of diffusion MRI (r2 > 0.98). When this 150 CT test set was expanded to include a total of 364 CT scans with a more heterogeneous distribution of infarct locations (94 stroke-negative, 270 stroke-positive mixed territory infarcts), model sensitivity was 97%, specificity 99%, for detection of infarcts larger than the 70 mL volume threshold used for patient selection in several major randomized controlled trials of thrombectomy treatment.

Incorporating algorithmic uncertainty into a clinical machine deep learning algorithm for urgent head CTs.

Machine learning (ML) algorithms to detect critical findings on head CTs may expedite patient management. Most ML algorithms for diagnostic imaging analysis utilize dichotomous classifications to determine whether a specific abnormality is present. However, imaging findings may be indeterminate, and algorithmic inferences may have substantial uncertainty. We incorporated awareness of uncertainty into an ML algorithm that detects intracranial hemorrhage or other urgent intracranial abnormalities and evaluated prospectively identified, 1000 consecutive noncontrast head CTs assigned to Emergency Department Neuroradiology for interpretation. The algorithm classified the scans into high (IC+) and low (IC-) probabilities for intracranial hemorrhage or other urgent abnormalities. All other cases were designated as No Prediction (NP) by the algorithm. The positive predictive value for IC+ cases (N = 103) was 0.91 (CI: 0.84-0.96), and the negative predictive value for IC- cases (N = 729) was 0.94 (0.91-0.96). Admission, neurosurgical intervention, and 30-day mortality rates for IC+ was 75% (63-84), 35% (24-47), and 10% (4-20), compared to 43% (40-47), 4% (3-6), and 3% (2-5) for IC-. There were 168 NP cases, of which 32% had intracranial hemorrhage or other urgent abnormalities, 31% had artifacts and postoperative changes, and 29% had no abnormalities. An ML algorithm incorporating uncertainty classified most head CTs into clinically relevant groups with high predictive values and may help accelerate the management of patients with intracranial hemorrhage or other urgent intracranial abnormalities.

Development and clinical application of a deep learning model to identify acute infarct on magnetic resonance imaging.

Stroke is a leading cause of death and disability. The ability to quickly identify the presence of acute infarct and quantify the volume on magnetic resonance imaging (MRI) has important treatment implications. We developed a machine learning model that used the apparent diffusion coefficient and diffusion weighted imaging series. It was trained on 6,657 MRI studies from Massachusetts General Hospital (MGH; Boston, USA). All studies were labelled positive or negative for infarct (classification annotation) with 377 having the region of interest outlined (segmentation annotation). The different annotation types facilitated training on more studies while not requiring the extensive time to manually segment every study. We initially validated the model on studies sequestered from the training set. We then tested the model on studies from three clinical scenarios: consecutive stroke team activations for 6-months at MGH, consecutive stroke team activations for 6-months at a hospital that did not provide training data (Brigham and Women's Hospital [BWH]; Boston, USA), and an international site (Diagnósticos da América SA [DASA]; Brazil). The model results were compared to radiologist ground truth interpretations. The model performed better when trained on classification and segmentation annotations (area under the receiver operating curve [AUROC] 0.995 [95% CI 0.992-0.998] and median Dice coefficient for segmentation overlap of 0.797 [IQR 0.642-0.861]) compared to segmentation annotations alone (AUROC 0.982 [95% CI 0.972-0.990] and Dice coefficient 0.776 [IQR 0.584-0.857]). The model accurately identified infarcts for MGH stroke team activations (AUROC 0.964 [95% CI 0.943-0.982], 381 studies), BWH stroke team activations (AUROC 0.981 [95% CI 0.966-0.993], 247 studies), and at DASA (AUROC 0.998 [95% CI 0.993-1.000], 171 studies). The model accurately segmented infarcts with Pearson correlation comparing model output and ground truth volumes between 0.968 and 0.986 for the three scenarios. Acute infarct can be accurately detected and segmented on MRI in real-world clinical scenarios using a machine learning model.

Using Keynote to present radiology images.

Numerous articles have offered instructions for working with advanced radiology images in Microsoft PowerPoint (Redmond, WA); however, no articles have detailed instructions to do the same on alternative presentation software. Apple Macintosh (Cupertino, CA) computers are gaining popularity with many radiologists, due in part to the availability of a powerful, free, open-source Digital Imaging and Communications in Medicine (DICOM) viewing and manipulating software OsiriX ( http://www.osirix-viewer.com ). Apple's own presentation software, Keynote, is particularly effective in dealing with medical images and cine clips. This article demonstrates how to use Apple's Keynote software to present radiology images and scrollable image stacks, without third-party add-on software. The article also illustrates how to compress media files and protect patient information in Keynote presentations. Lastly, it addresses the steps to converting between PowerPoint and Keynote file formats. Apple's Keynote software enables quick and efficient addition of multiple static images or scrollable image stacks, compression of media files, and removal of patient information. These functions can be accomplished by inexperienced users with no software modifications.

Lumbar MRI Reporting Efficiency for Trainees Over the Academic Year: An Opportunity for Improving Clinical Workflows in Academic Medical Centers.

PURPOSE: Reporting efficiency is commonly used to measure performance and quality in diagnostic imaging. For academic centers, balancing the clinical demand for efficient reporting and educational obligation to trainees remains a major challenge. The objective of this study was to quantify the effect of trainee education on reporting efficiency over the academic year (July to June) for a single diagnostic imaging examination type. METHODS: The authors reviewed a 10-year data set of lumbar spinal MRI reports and time-stamp data and compared change in mean reporting time for trainee versus attending radiologist-only reports. Odds ratios, linear regression, and correlation analysis were performed to evaluate relationships of mean and cumulative reporting times, volume, and study month. RESULTS: Mean reporting time for the trainee group peaked in July (287.2 ± 4.9 min). The largest month-to-month increase was from June to July (+46.3 min, P < .01) for the trainees, and July showed the largest deviation from annualized mean reporting time (odds ratio, 1.31; 95% confidence interval, 1.16-1.47; P < .001). Mean reporting time improved linearly over the course of the academic year from July to June for trainees (R2 = 0.59, P = .002), but this effect was absent in the attending radiologist-only group (P = .52). CONCLUSIONS: This study quantifies the effect of trainee education on reporting efficiency and models the operational "learning curve" of improved performance over the academic year. These data may inform staffing and workflow improvement efforts in academic radiology departments.

An explainable deep-learning algorithm for the detection of acute intracranial haemorrhage from small datasets.

Owing to improvements in image recognition via deep learning, machine-learning algorithms could eventually be applied to automated medical diagnoses that can guide clinical decision-making. However, these algorithms remain a 'black box' in terms of how they generate the predictions from the input data. Also, high-performance deep learning requires large, high-quality training datasets. Here, we report the development of an understandable deep-learning system that detects acute intracranial haemorrhage (ICH) and classifies five ICH subtypes from unenhanced head computed-tomography scans. By using a dataset of only 904 cases for algorithm training, the system achieved a performance similar to that of expert radiologists in two independent test datasets containing 200 cases (sensitivity of 98% and specificity of 95%) and 196 cases (sensitivity of 92% and specificity of 95%). The system includes an attention map and a prediction basis retrieved from training data to enhance explainability, and an iterative process that mimics the workflow of radiologists. Our approach to algorithm development can facilitate the development of deep-learning systems for a variety of clinical applications and accelerate their adoption into clinical practice.

Dual-Energy Computed Tomography Angiography of the Head and Neck and Related Applications.

Dual-energy computed tomography (DECT) has become an increasingly widespread and useful component of the neuroimaging armamentarium, offering automated bone removal, metallic artifact reduction, and improved characterization of iodinated contrast enhancement. The application of these techniques to CT neuroangiography enables a number of benefits including more efficient 3D post-processing, contrast dose reduction opportunities, successful differentiation of hemorrhage from contrast staining following thromboembolic recanalization therapy, improved detection of active contrast extravasation in the setting of intracranial hemorrhage, and more precise characterization of atheromatous steno-occlusive disease.