Machine-learning algorithm in acute stroke: real-world experience.

AIM: To assess the clinical performance of a commercially available machine learning (ML) algorithm in acute stroke. MATERIALS AND METHODS: CT and CT angiography (CTA) studies of 104 consecutive patients (43 females, age range 19-93, median age 62) performed for suspected acute stroke at a single tertiary institution with real-time ML software analysis (RAPID™ ASPECTS and CTA) were included. Studies were retrospectively reviewed independently by two neuroradiologists in a blinded manner. RESULTS: The cohort included 24 acute infarcts and 16 large vessel occlusions (LVO). RAPID™ ASPECTS interpretation demonstrated high sensitivity (87.5%) and NPV (87.5%) but very poor specificity (30.9%) and PPV (30.9%) for detection of acute ischaemic parenchymal changes. There was a high percentage of false positives (51.1%). In cases of proven LVO, RAPID™ ASPECTS showed good correlation with neuroradiologists' blinded independent interpretation, Pearson correlation coefficient = 0.96 (both readers), 0.63 (RAPID™ vs reader 1), 0.69 (RAPID™ vs reader 2). RAPID™ CTA interpretation demonstrated high sensitivity (92.3%), specificity (85.3%), and negative predictive (NPV) (98.5%) with moderate positive predictive value (PPV) (52.2%) for detection of LVO (N=13). False positives accounted for 12.5% of cases, of which 27.3% were attributed to arterial stenosis. CONCLUSION: RAPID™ CTA was robust and reliable in detection of LVO. Although demonstrating high sensitivity and NPV, RAPID™ ASPECTS interpretation was associated with a high number of false positives, which decreased clinicians' confidence in the algorithm. However, in cases of proven LVO, RAPID™ ASPECTS performed well and had good correlation with neuroradiologists' blinded interpretation.

MRI Evaluation of the Normal and Abnormal Endolymphatic Duct in the Pediatric Population: A Comparison with High-Resolution CT.

BACKGROUND AND PURPOSE: An enlarged vestibular aqueduct is the most commonly reported imaging abnormality in children with sensorineural hearing loss. MR imaging is often used to evaluate pediatric sensorineural hearing loss; however, there are no well-established size criteria on MR imaging to diagnose an enlarged endolymphatic duct. The first purpose of the study was to determine a range of normal endolymphatic duct sizes on MR imaging and compare it with that in high-resolution CT. The second purpose was to assess the sensitivity and specificity of MR imaging in diagnosing an enlarged endolymphatic duct in patients with an enlarged vestibular aqueduct on CT. MATERIALS AND METHODS: Endolymphatic duct midaperture measurements were analyzed in 52 patients with no history of sensorineural hearing loss. Comparison of CT and MR imaging was made in a second cohort of 41 patients with a normal midaperture width on CT. The sensitivity and specificity of MR imaging were then evaluated in a third cohort of 24 patients with a documented enlarged vestibular aqueduct on CT. RESULTS: In 94 ears, normal endolymphatic duct midaperture measurements ranged from 0 to 0.9 mm on MR imaging. A significant correlation (P <.001) and moderate agreement were found between CT and MR imaging in 81 ears with a normal vestibular aqueduct on CT. Twenty-four patients had bilateral (n = 14) or unilateral (n = 10) enlarged vestibular aqueducts on CT, and the sensitivity and specificity of MR imaging were 97% and 100%, respectively, for a diagnosis of an enlarged endolymphatic duct. CONCLUSIONS: MR imaging measurements of the normal endolymphatic duct are similar to those established for CT. MR imaging is a useful tool for the diagnosis of enlarged vestibular aqueduct.

COVID-19 Stroke Apical Lung Examination Study: A Diagnostic and Prognostic Imaging Biomarker in Suspected Acute Stroke.

BACKGROUND AND PURPOSE: Diagnosis of coronavirus disease 2019 (COVID-19) relies on clinical features and reverse-transcriptase polymerase chain reaction testing, but the sensitivity is limited. Carotid CTA is a routine acute stroke investigation and includes the lung apices. We evaluated CTA as a potential COVID-19 diagnostic imaging biomarker. MATERIALS AND METHODS: This was a multicenter, retrospective study (n = 225) including CTAs of patients with suspected acute stroke from 3 hyperacute stroke units (March-April 2020). We evaluated the reliability and accuracy of candidate diagnostic imaging biomarkers. Demographics, clinical features, and risk factors for COVID-19 and stroke were analyzed using univariate and multivariate statistics. RESULTS: Apical ground-glass opacification was present in 22.2% (50/225) of patients. Ground-glass opacification had high interrater reliability (Fleiss κ = 0.81; 95% CI, 0.68-0.95) and, compared with reverse-transcriptase polymerase chain reaction, had good diagnostic performance (sensitivity, 75% [95% CI, 56-87]; specificity, 81% [95% CI, 71-88]; OR = 11.65 [95% CI, 4.14-32.78]; P < .001) on multivariate analysis. In contrast, all other contemporaneous demographic, clinical, and imaging features available at CTA were not diagnostic for COVID-19. The presence of apical ground-glass opacification was an independent predictor of increased 30-day mortality (18.0% versus 5.7%, P = .017; hazard ratio = 3.51; 95% CI, 1.42-8.66; P = .006). CONCLUSIONS: We identified a simple, reliable, and accurate COVID-19 diagnostic and prognostic imaging biomarker obtained from CTA lung apices: the presence or absence of ground-glass opacification. Our findings have important implications in the management of patients presenting with suspected stroke through early identification of COVID-19 and the subsequent limitation of disease transmission.

Machine learning and glioma imaging biomarkers.

AIM: To review how machine learning (ML) is applied to imaging biomarkers in neuro-oncology, in particular for diagnosis, prognosis, and treatment response monitoring. MATERIALS AND METHODS: The PubMed and MEDLINE databases were searched for articles published before September 2018 using relevant search terms. The search strategy focused on articles applying ML to high-grade glioma biomarkers for treatment response monitoring, prognosis, and prediction. RESULTS: Magnetic resonance imaging (MRI) is typically used throughout the patient pathway because routine structural imaging provides detailed anatomical and pathological information and advanced techniques provide additional physiological detail. Using carefully chosen image features, ML is frequently used to allow accurate classification in a variety of scenarios. Rather than being chosen by human selection, ML also enables image features to be identified by an algorithm. Much research is applied to determining molecular profiles, histological tumour grade, and prognosis using MRI images acquired at the time that patients first present with a brain tumour. Differentiating a treatment response from a post-treatment-related effect using imaging is clinically important and also an area of active study (described here in one of two Special Issue publications dedicated to the application of ML in glioma imaging). CONCLUSION: Although pioneering, most of the evidence is of a low level, having been obtained retrospectively and in single centres. Studies applying ML to build neuro-oncology monitoring biomarker models have yet to show an overall advantage over those using traditional statistical methods. Development and validation of ML models applied to neuro-oncology require large, well-annotated datasets, and therefore multidisciplinary and multi-centre collaborations are necessary.

Emendation of the Genus Auritidibacter Yassin et al. 2011 and Auritidibacter ignavus Yassin et al. 2011 based on features observed from Canadian and Swiss clinical isolates and whole-genome sequencing analysis.

Auritidibacter ignavus is a Gram-stain-positive bacillus derived from otorrhea. Four strains derived from ear discharges in Canada and Switzerland, with features consistent with but distinguishable from Auritidibacter ignavus IMMIB L-1656T (accession number FN554542) by 16S rRNA gene sequencing (97.5 % similarity), were thought to represent a novel species of the genus Auritidibacter. Auritidibacter ignavus DSM 45359T (=IMMIB L-1656T) was acquired to compare with Canadian and Swiss strains by whole-genome sequencing (WGS). Unexpectedly, those isolates were observed to be consistent with A. ignavus DSM 45359T by WGS (ANIb scores >98 %), MALDI-TOF (Bruker), cellular fatty acid analysis and biochemically (some differences were observed). A nearly full 16S rRNA gene sequence could not be readily prepared from A. ignavus DSM 45359T, even after multiple attempts. A 16S rRNA gene chimeric consensus sequence created from the genome assembly of A. ignavus DSM 45359T had only 97.5 % similarity to that of A. ignavus IMMIB L-1656T, implying that 16S rRNA sequence accession number FN554542 could not be replicated. We concluded that our isolates of members of the genus Auritidibacter were consistent with A. ignavus DSM 45359T, did not represent a novel species, and that the sequence corresponding to FN554542 was not reproducible. By WGS, A. ignavus DSM 45359T had genome of 2.53×106 bp with a DNA G+C content of 59.34%, while genomes of Canadian and Swiss isolates ranged from 2.47 to 2.59×106 bp with DNA G+C contents of 59.3-59.52 %. A. ignavus NML 100628 (=NCTC 14178=LMG 30897) did not demonstrate a rodcoccus cycle. Emendation of Auritidibacter ignavus was proposed based on these results.

Outcome Study of the Pipeline Embolization Device with Shield Technology in Unruptured Aneurysms (PEDSU).

BACKGROUND AND PURPOSE: The recently introduced Pipeline Flex Embolization Device with Shield Technology (Pipeline Shield) is the third generation of Pipeline flow-diverter devices. It has a new stent-surface modification, which reduces thrombogenicity. We aimed to evaluate clinical and radiographic (safety and efficacy) outcomes of the Pipeline Shield. MATERIALS AND METHODS: The 30-day and 1-year mortality and morbidity rates and the 6- and 18-month radiographic aneurysm occlusion outcomes for procedures performed between March 2016 and January 2018 were analyzed. 3D-TOF-MRA was used for follow-up. RESULTS: Forty-four attempted Pipeline Shield procedures were performed for 41 patients with 44 target aneurysms (total of 52 aneurysms treated). A total of 88.5% of devices were inserted in the anterior circulation, and 11.5%, in the posterior circulation; 49/52 (94.2%) aneurysms were saccular; and 1/52 (1.9%) was fusiform. One (1.9%) aneurysm was an iatrogenic pseudoaneurysm, and 1 (1.9%) was a dissecting aneurysm. Seventy-one percent (35/49) of the saccular aneurysms were wide-neck (neck, >4 mm), 34.6% (18/52) were large (≥10 mm), and 3.8% (2/52) were giant (≥25 mm). The mean aneurysm sac maximal diameter was 9.0 mm, and the mean neck width was 5.0 mm. The cumulative mortality and morbidity rates were 2.3% and 6.8% at 1 year, respectively. The adequate occlusion rate was 78.8% at 6 months and 90.3% at 18 months. CONCLUSIONS: In this pragmatic and non-industry-sponsored study, the occlusion rates and safety outcomes were similar to those seen in previously published studies with flow-diverter devices and earlier generation Pipeline Embolization Devices.

The Fellowship of the Royal College of Radiologists (FRCR) examination: a review of the evidence.

The Fellowship of the Royal College of Radiologists (FRCR) examination is the professional qualification that is essential for career progression in clinical radiology within the UK. It is also important for career progression in many countries internationally. The FRCR has evolved and changed over the last decade. In this systematic review we appraise and summarise the available data relating to the FRCR emphasising the published evidence regarding the validity, reliability, and acceptability of this examination. Comparison is made to other equivalent medical examinations, as well as a more recently published commissioned external review of the FRCR examinations. The Clinical Radiology Part 2B (CR2B) examination in its pre-existing format is reliable, valid, and acceptable. Recommendations from the commissioned external review are based primarily on expert opinion, with a limited evidence base comprising data from a small sample acquired during a single examination sitting and without peer review. Unlike the CR2B examination, there is little evidence regarding assessment of the CR1 and CR2 examinations. Both the CR1 and CR2 examinations are currently in the process of undergoing major changes to their formats. Blueprinting items to the curriculum might improve acceptability. Other changes may improve transparency and reliability of these assessments. Our analysis and many aspects of the external review may provide pointers regarding how the upcoming data produced by the "automated" FRCR examinations can be further analysed to provide a more robust evaluation.

Evaluation of the Normal Cochlear Second Interscalar Ridge Angle and Depth on 3D T2-Weighted Images: A Tool for the Diagnosis of Scala Communis and Incomplete Partition Type II.

BACKGROUND AND PURPOSE: Cochlear malformations may be be subtle on imaging studies. The purpose of this study was to evaluate the angle and depth of the lateral second interscalar ridge or notch in ears without sensorineural hearing loss (normal ears) and compare them with ears that have a documented incomplete type II partition malformation. MATERIALS AND METHODS: The second interscalar ridge notch angle and depth were measured on MR imaging in normal ears by a single experienced neuroradiologist. The images of normal and incomplete partition II malformation ears were then randomly mixed for 2 novice evaluators to measure both the second interscalar ridge notch angle and depth in a blinded manner. For the mixed group, interobserver agreement was calculated, normal and abnormal ear measurements were compared, and receiver operating characteristic curves were generated. RESULTS: The 94 normal ears had a mean second interscalar ridge angle of 80.86° ± 11.4° and depth of 0.54 ± 0.14 mm with the 98th percentile for an angle of 101° and a depth of 0.3 mm. In the mixed group, agreement between the 2 readers was excellent, with significant differences for angle and depth found between normal and incomplete partition type II ears for angle and depth on average (P < .001). Receiver operating characteristic cutoffs for delineating normal from abnormal ears were similar for both readers (depth, 0.31/0.34 mm; angle, 114°/104°). CONCLUSIONS: A measured angle of >114° and a depth of the second interscalar ridge notch of ≤0.31 mm suggest the diagnosis of incomplete partition type II malformation and scala communis. These measurements can be accurately made by novice readers.

MR Imaging of the Cervical Spine in Nonaccidental Trauma: A Tertiary Institution Experience.

BACKGROUND AND PURPOSE: Cervical MR imaging has demonstrated a utility for detecting soft tissue injury in nonaccidental trauma. The purpose of this study was to identify the incidence and types of cervical spine injury on MR imaging in nonaccidental trauma and to correlate cervical spine injury with parenchymal injury on brain MR imaging and findings on head CT. MATERIALS AND METHODS: A retrospective review of children diagnosed with nonaccidental trauma in a tertiary referral pediatric hospital over 8 years was performed. Inclusion criteria were children younger than 5 years of age, a confirmed diagnosis of nonaccidental trauma, and cervical spine MR imaging within 1 week of presentation. Brain and cervical spine MR imaging, head CT, cervical radiographs, and skeletal surveys were reviewed. RESULTS: There were 89 patients included in this study (48 males; mean age, 9.1 months [range, 1-59 months]). Cervical spine injury on MR imaging was found in 61 patients (69%). Ligamentous injury was seen in 60 patients (67%), with interspinous ligaments being most commonly involved. Abnormal capsular fluid (atlanto-occipital and atlantoaxial) was present in 28 patients (32%). Cervical spine injury on MR imaging was significantly associated with parenchymal restricted diffusion on brain MR imaging and parenchymal injury on head CT (P = .0004 and P = .0104, respectively). Children with restricted diffusion on brain MR imaging were 6.22 (point estimate) times more likely to have cervical spine injury on MR imaging. CONCLUSIONS: There is a high incidence of cervical spine injury in pediatric nonaccidental trauma. Positive findings may affect management and suggest a traumatic etiology.

Spectrum of Clinical and Associated MR Imaging Findings in Children with Olfactory Anomalies.

BACKGROUND AND PURPOSE: The olfactory apparatus, consisting of the bulb and tract, is readily identifiable on MR imaging. Anomalous development of the olfactory apparatus may be the harbinger of anomalies of the secondary olfactory cortex and associated structures. We report a large single-site series of associated MR imaging findings in patients with olfactory anomalies. MATERIALS AND METHODS: A retrospective search of radiologic reports (2010 through 2014) was performed by using the keyword "olfactory"; MR imaging studies were reviewed for olfactory anomalies and intracranial and skull base malformations. Medical records were reviewed for clinical symptoms, neuroendocrine dysfunction, syndromic associations, and genetics. RESULTS: We identified 41 patients with olfactory anomalies (range, 0.03-18 years of age; M/F ratio, 19:22); olfactory anomalies were bilateral in 31 of 41 patients (76%) and absent olfactory bulbs and olfactory tracts were found in 56 of 82 (68%). Developmental delay was found in 24 (59%), and seizures, in 14 (34%). Pituitary dysfunction was present in 14 (34%), 8 had panhypopituitarism, and 2 had isolated hypogonadotropic hypogonadism. CNS anomalies, seen in 95% of patients, included hippocampal dysplasia in 26, cortical malformations in 15, malformed corpus callosum in 10, and optic pathway hypoplasia in 12. Infratentorial anomalies were seen in 15 (37%) patients and included an abnormal brain stem in 9 and an abnormal cerebellum in 3. Four patients had an abnormal membranous labyrinth. Genetic testing was performed in 23 (56%) and findings were abnormal in 11 (48%). CONCLUSIONS: Olfactory anomalies should prompt careful screening of the brain, skull base, and the pituitary gland for additional anomalies. Genetic testing should be considered.