Inter- and Intra-Observer Agreement When Using a Diagnostic Labeling Scheme for Annotating Findings on Chest X-rays-An Early Step in the Development of a Deep Learning-Based Decision Support System.

Consistent annotation of data is a prerequisite for the successful training and testing of artificial intelligence-based decision support systems in radiology. This can be obtained by standardizing terminology when annotating diagnostic images. The purpose of this study was to evaluate the annotation consistency among radiologists when using a novel diagnostic labeling scheme for chest X-rays. Six radiologists with experience ranging from one to sixteen years, annotated a set of 100 fully anonymized chest X-rays. The blinded radiologists annotated on two separate occasions. Statistical analyses were done using Randolph's kappa and PABAK, and the proportions of specific agreements were calculated. Fair-to-excellent agreement was found for all labels among the annotators (Randolph's Kappa, 0.40-0.99). The PABAK ranged from 0.12 to 1 for the two-reader inter-rater agreement and 0.26 to 1 for the intra-rater agreement. Descriptive and broad labels achieved the highest proportion of positive agreement in both the inter- and intra-reader analyses. Annotating findings with specific, interpretive labels were found to be difficult for less experienced radiologists. Annotating images with descriptive labels may increase agreement between radiologists with different experience levels compared to annotation with interpretive labels.

Head or Neck First? Speed and Rates of Reperfusion in Thrombectomy for Tandem Large Vessel Occlusion Strokes.

BACKGROUND: We aim to evaluate the speed and rates of reperfusion in tandem large vessel occlusion acute stroke patients undergoing upfront cervical lesion treatment (Neck-First: angioplasty and/or stent before thrombectomy) as compared to direct intracranial occlusion therapy (Head-First) in a large international multicenter cohort. METHODS: The Thrombectomy In TANdem Lesions (TITAN) collaboration pooled individual data of prospectively collected thrombectomy international databases for all consecutive anterior circulation tandem patients who underwent emergent thrombectomy. The co-primary outcome measures were rates of successful reperfusion (modified Thrombolysis in Cerebral Infarction 2b/3) and time from groin puncture to successful reperfusion. RESULTS: In total, 289 patients with tandem atherosclerotic etiology were included in the analysis (182 Neck-First and 107 Head-First patients). Except for differences in the Alberta Stroke Program Early CT Score (ASPECTS; median 8 [range 7-10] Neck-First vs. 7 [range 6-8] Head-First; p < 0.001) and cervical internal carotid artery (ICA) lesion severity (complete occlusion in 35% of the Neck-First vs. 57% of the Head-First patients; p < 0.001), patient characteristics were well balanced. After adjustments, there was no difference in successful reperfusion rates between the study groups (odds ratio associated with Neck-First: 1.18 [95% confidence interval, 0.60-2.17]). The time to successful reperfusion from groin puncture was significantly shorter in the Head-First group after adjustments (median 56 min [range 39-90] vs. 70 [range 50-102]; p = 0.001). No significant differences in the rates of full reperfusion, symptomatic hemorrhage, 90-day independence, or mortality were observed. Sensitivity analysis excluding patients with complete cervical ICA occlusion yielded similar results. CONCLUSIONS: The upfront approach of the intracranial lesion in patients with tandem large vessel occlusion strokes leads to similar reperfusion rates but faster reperfusion as compared to initial cervical revascularization followed by mechanical thrombectomy. Controlled studies are warranted.

Impact of Antiplatelet Therapy During Endovascular Therapy for Tandem Occlusions: A Collaborative Pooled Analysis.

Background and Purpose- Antiplatelet agents could be used in the setting of endovascular therapy for tandem occlusions to reduce the risk of de novo intracranial embolic migration, reocclusion of the extracranial internal carotid artery lesion, or in-stent thrombosis in case of carotid stent placement but have to be balanced with the intracerebral hemorrhagic transformation risk. In this study, we aim to investigate the impact of acute antiplatelet therapy administration on outcomes during endovascular therapy for anterior circulation tandem occlusions. Methods- This is a retrospective analysis of a collaborative pooled analysis of 11 prospective databases from the multicenter observational TITAN registry (Thrombectomy in Tandem Lesions). Patients were divided into groups based on the number of antiplatelet administered during endovascular therapy. The primary outcome was favorable outcome, defined as a modified Rankin Scale score of 0 to 2 at 90 days. Results- This study included a total of 369 patients; 145 (39.3%) did not receive any antiplatelet agent and 224 (60.7%) received at least 1 antiplatelet agent during the procedure. Rate of favorable outcome was nonsignificantly higher in patients treated with antiplatelet therapy (58.3%) compared with those treated without antiplatelet (46.0%; adjusted odds ratio, 1.38 [95% CI, 0.78-2.43]; P=0.26). Rate of 90-day mortality was significantly lower in patients treated with antiplatelet therapy (11.2% versus 18.7%; adjusted odds ratio, 0.47 [95% CI, 0.22-0.98]; P=0.042), without increasing the risk of any intracerebral hemorrhage. Successful reperfusion (modified Thrombolysis in Cerebral Ischemia score 2b-3) rate was significantly better in the antiplatelet therapy group (83.9% versus 71.0%; adjusted odds ratio, 1.89 [95% CI, 1.01-3.64]; P=0.045). Conclusions- Administration of antiplatelet therapy during endovascular therapy for anterior circulation tandem occlusions was safe and was associated with a lower 90-day mortality. Optimal antiplatelet therapy remains to be assessed, especially when emergent carotid artery stenting is performed. Further randomized controlled trials are needed.

Periprocedural Heparin During Endovascular Treatment of Tandem Lesions in Patients with Acute Ischemic Stroke: A Propensity Score Analysis from TITAN Registry.

BACKGROUND AND PURPOSE: Data on safety and efficacy of periprocedural use of heparin are limited during treatment of acute ischemic stroke patients with anterior circulation tandem occlusion. This study aimed to investigate the impact of heparin use during endovascular therapy of anterior circulation tandem occlusions on the functional and safety outcomes. METHODS: A retrospective analysis of the multicenter observational TITAN registry was performed. Patients with anterior circulation tandem occlusion and treated with endovascular therapy (EVT) were included, with or without extracranial carotid intervention. We divided patients into two groups based on periprocedural heparin use (heparin vs. non-heparin). The dose of intravenous unfractionated heparin ranged from 1500 to 2500 I.U. Primary study endpoint was 90-day Modified Rankin Scale (mRS). Secondary study endpoint included angiographic and safety endpoints such as hemorrhagic complications. A propensity-score-matched analysis was performed. RESULTS: Among 369 patients, heparin was used in 68 patients (18.4%). In the propensity-score-matched cohort, favorable outcome (mRS 0-2) occurred in 51.3% in heparin group and 58.0% in non-heparin group (matched OR, 0.76; 95% CI, 0.32-1.78; P = 0.52). Similar result was found in propensity-score-adjusted cohort (adjusted OR, 0.72; 95%CI, 0.39-1.32; P = 0.28). Likewise, there was no difference in the rate of successful reperfusion (mTICI 2b-3) (propensity-score-adjusted OR, 1.03; 95%CI, 0.50-2.09; P = 0.93) neither in safety endpoints between the two groups. CONCLUSIONS: Periprocedural heparin use during EVT of anterior circulation tandem occlusions was not associated with better functional, angiographic or safety outcomes. These findings are applicable for low doses of heparin, and further studies are warranted.

Effect of extracranial lesion severity on outcome of endovascular thrombectomy in patients with anterior circulation tandem occlusion: analysis of the TITAN registry.

INTRODUCTION: Endovascular treatment (EVT) for tandem occlusion (TO) of the anterior circulation is complex but effective. The effect of extracranial internal carotid artery (EICA) lesion severity on the outcomes of EVT is unknown. In this study we investigated the effect of EICA lesion severity on the outcomes of tandem occlusion EVT. METHODS: A multicenter retrospective TITAN (Thrombectomy In TANdem lesions) study that included 18 international endovascular capable centers was performed. Patients who received EVT for atherosclerotic TO with or without EICA lesion intervention were included. Patients were divided into two groups based on the EICA lesion severity (high-grade stenosis (≥90% North American Symptomatic Carotid Endarterectomy Trial) vs complete occlusion). Outcome measures included the 90-day clinical outcome (modified Rankin Scale score (mRS)), angiographic reperfusion (modified Thrombolysis In Cerebral Ischemia (mTICI) at the end of the procedure), procedural complications, and intracranial hemorrhage at 24 hours follow-up. RESULTS: A total of 305 patients were included in the study, of whom 135 had complete EICA occlusion and 170 had severe EICA stenosis. The EICA occlusion group had shorter mean onset-to-groin time (259±120 min vs 305±202 min; p=0.037), more patients with diabetes, and fewer with hyperlipidemia. With respect to the outcome, mTICI 2b-3 reperfusion was lower in the EICA occlusion group (70% vs 81%; p=0.03). The favorable outcome (90-day mRS 0-2), intracerebral hemorrhage and procedural complications were similar in both groups. CONCLUSION: Atherosclerotic occlusion of the EICA in acute tandem strokes was associated with a lower rate of mTICI 2b-3 reperfusion but similar functional and safety outcomes when compared with high-grade EICA stenosis.

Hemorrhagic Transformation After Thrombectomy for Tandem Occlusions.

Background and Purpose- Carotid artery stenting in tandem large vessel occlusion strokes is usually avoided because of the intracranial hemorrhagic risks induced by antiplatelet therapy during thrombectomy interventions. This study aimed to evaluate the incidence of hemorrhagic transformation following thrombectomy in large vessel occlusion strokes patients with atherosclerotic cervical carotid occlusion, associated factors, and clinical relevance. Methods- The TITAN (Thrombectomy in Tandem Lesions) collaboration pooled individual data of prospectively collected multicentric thrombectomy databases for consecutive anterior circulation tandem large vessel occlusion strokes patients who underwent thrombectomy. Hemorrhagic infarction (HI) and parenchymal hematoma (PH) were assessed within 24 hours. Results- Among 289 patients with atherosclerotic cause, 66 (24.7%) patients developed HI and 38 (14.2%) PH. Intracranial carotid occlusion, diabetes mellitus, absence of prior intravenous thrombolysis, and complete extracranial carotid occlusion were independent predictors of HI. Similar predictors were found for PH with addition of higher baseline National Institutes of Health Stroke Scale and Alberta Stroke Program Early CT Score <7. No detrimental effect of HI on 90-day clinical outcome was found. The occurrence of PH was associated with increased mortality rates (adjusted odds ratio, 2.63; 95% CI, 1.05-6.59; P=0.039) and had no detrimental effect on 90-day modified Rankin Scale 0 to 2 (adjusted odds ratio, 0.52; 95% CI, 0.20-1.28; P=0.25). Conclusions- Incidence of PH after tandem large vessel occlusion strokes thrombectomy is equivalent to those reported in the literature data for isolated occlusions. Similar predictors were found for PH and HI within 24 hours, whereas acute carotid artery stenting and antiplatelet therapy were not, suggesting an aggressive endovascular treatment of tandem occlusions.

[Christmas article: The best bubbles - radiology as sommelier].

INTRODUCTION: The aim of the study was to assess bubble characteristics in three different sparkling wines, two different sparkling waters, one standard beer and one demineralised water for ironing with CT, MRI and ultrasound (US). Furthermore, the correlation between personal taste and the measured bubble intensity should be assessed. METHODS: The study was carried out as a prospective study. All liquid samples were scanned in three rounds, and the outcomes were measured blinded. Afterwards four healthy volunteers from a department of radiology tasted all samples: three sparkling wines with different ages, two tiers of sparkling waters, one pilsner beer and demineralised water for ironing. The main outcome measures were: bubble size, intensity and decay, evaluated with MRI and US and Hounsfield units (HU) from CT. Foam, bubble intensity, smell and colour were evaluated from a custom-made questionnaire. RESULTS: The decay of bubbles in sparkling wine and beer was slower than in sparkling water as seen on US and MRI, and the intensity as well as the bubble size diminished over time. CT was unable to assess bubble characteristics but showed increasing HU values with increasing sugar content. Measurements on US and MRI were correlated with more factors from the taste evaluation than CT was. No signs of intoxication were observed in the assessors. CONCLUSIONS: US and MRI could be useful tools for evaluating sparkling wine objectively in blaseology. CT was of insufficient spatial resolution to assess individual bubbles.

Carotid Stenting With Antithrombotic Agents and Intracranial Thrombectomy Leads to the Highest Recanalization Rate in Patients With Acute Stroke With Tandem Lesions.

OBJECTIVES: The aim of this study was to identify the optimal endovascular approach in patients with acute stroke with tandem lesions. BACKGROUND: At present, there is no consensus about the ideal technical strategy for the endovascular treatment of patients with acute ischemic stroke with tandem lesions of the extracranial internal carotid artery (ICA) and intracranial cerebral arteries. METHODS: This was an international, multicenter registry with a total of 482 patients with acute ischemic stroke and tandem lesions. Patients were treated by intracranial thrombectomy as well as 1 of the following 4 strategies: 1) acute carotid artery stenting of the extracranial ICA with antithrombotic agents; 2) acute carotid artery stenting of the extracranial ICA without antithrombotic agents; 3) balloon angioplasty of the extracranial ICA; and 4) intracranial thrombectomy alone. The main outcome endpoints of the study were the degree of recanalization and the 90-day clinical outcome. The safety endpoints were symptomatic intracerebral hemorrhage and all causes of mortality at 90 days. RESULTS: Using univariate analysis, the rates of successful reperfusion (modified Thrombolysis in Cerebral Infarction grades 2B and 3) and favorable clinical outcome after 90 days were significantly higher after acute carotid stenting with antithrombotic therapy and thrombectomy compared with the group with thrombectomy alone. After adjusting for confounding variables, acute stenting with antithrombotic therapy was independently associated with successful recanalization (odds ratio: 2.4; 95% confidence interval: 1.25 to 4.59; p = 0.008). The rates of symptomatic intracerebral hemorrhage and 90-day mortality were comparable among all 4 treatment groups. CONCLUSIONS: Acute stenting of the extracranial ICA with antithrombotic therapy in combination with intracranial thrombectomy is associated with higher recanalization rates in treatment of patients with acute stroke with tandem lesions.

Thrombectomy in Acute Stroke With Tandem Occlusions From Dissection Versus Atherosclerotic Cause.

BACKGROUND AND PURPOSE: Tandem steno-occlusive lesions were poorly represented in randomized trials and represent a major challenge for endovascular thrombectomy in acute anterior circulation strokes. The impact of the cervical carotid lesion cause (ie, atherosclerotic versus dissection) on outcome of tandem patients endovascularly treated remains to be assessed. METHODS: We retrospectively analyzed individual data of prospectively collected consecutive tandem patients treated with endovascular thrombectomy. The primary outcome was favorable outcome at 90 days (modified Rankin Scale score of 0-2). Secondary efficacy outcomes included successful reperfusion (modified Thrombolysis in Cerebrovascular Infarction scores of 2b-3), time to reperfusion, and safety outcomes encompassed procedural complications, symptomatic intracerebral hemorrhage, and 90-day mortality. RESULTS: Among the 295 included patients, 65 had cervical carotid dissection and 230 had cervical carotid atherosclerotic cause. The rate of favorable outcome was 56.3% in the dissection group versus 47.6% in the atherosclerotic arm (center-, age-, and admission National Institutes of Health Stroke Scale-adjusted odds ratio, 1.08; 95% confidence interval, 0.50-2.30; P=0.85). No significant differences were observed in secondary outcomes. The rates of successful reperfusion, symptomatic intracerebral hemorrhage, and 90-day mortality were 78.5% versus 74.5% (P=0.13), 4.6% versus 5.2% (P=1.0), and 7.8% versus 15.3% (P=0.94) in the dissection versus atherosclerotic groups, respectively. The median procedural time was 76 minutes (interquartile range, 52-95 minutes) in the dissection group and 67 minutes (interquartile range, 45-98 minutes) in the atherosclerotic group (P=0.24). CONCLUSIONS: We found no differences in the outcomes of patients with anterior circulation tandem atherosclerotic and dissection lesions treated with endovascular thrombectomy. Further studies are warranted.

Carotid artery stenting versus no stenting assisting thrombectomy for acute ischaemic stroke: protocol for a systematic review of randomised clinical trials with meta-analyses and trial sequential analyses.

BACKGROUND: In patients with intracranial large vessel arterial occlusion, ipsilateral extracranial carotid artery occlusions or near-occlusions pose a significant hurdle in endovascular management of acute ischaemic stroke. Stenting of the carotid lesion may be beneficial in this situation to provide a stable access for introducing catheters through the carotid lesion into the intracranial vasculature and the target occlusion. Furthermore, carotid stenting may ensure ample blood flow for wash-out of clot material and reperfusion of the ischaemic penumbral tissue. However, antiplatelet therapy administered to prevent stent thrombosis and sudden increase in blood flow after reopening of the carotid lesion may increase the risk for intracranial haemorrhagic complications. This review aims to assess the benefits and harms of carotid stenting vs. no stenting assisting thrombectomy for acute ischaemic stroke. METHODS: International and regional electronic databases will be searched to identify eligible randomised clinical trials. To identify further published, unpublished, or on-going and planned trials searches of Google Scholar, Worldwide Food and Drug Administrations, Worldwide Medicines Agencies, company homepages, reference lists, conference proceedings, and the Science Citation Index cited reference search index will be conducted. Manufacturers of relevant interventional equipment, authors, colleagues, and researchers active in the field will be contacted. No language restrictions will be applied to these searches. Randomised clinical trials will be included for assessing benefits and harms and quasi-randomised studies, and observational studies will be included for assessing harms of the intervention. Meta-analyses will be performed according to the recommendations of the Cochrane Handbook for Systematic Reviews of Interventions, and Trial Sequential Analyses will be conducted to control the risk of random errors and prevent premature statements of superiority of the experimental or control intervention or premature statement of futility. The quality of the evidence will be evaluated with the Grading of Recommendations Assessment, Development, and Evaluation. DISCUSSION: This systematic review of carotid stenting in endovascular management of acute ischaemic stroke in patients with concomitant extracranial carotid lesions and intracranial embolism will assess benefits and harms of this intervention and assesses whether carotid stenting should be encouraged or avoided in acute ischaemic stroke and identify targets for further research. SYSTEMATIC REVIEW REGISTRATION: PROSPERO CRD42016033346.

Thrombectomy assisted by carotid stenting in acute ischemic stroke management: benefits and harms.

Extracranial carotid artery occlusion or high-grade stenosis with concomitant intracranial embolism causes severe ischemic stroke and shows poor response rates to intravenous thrombolysis (IVT). Endovascular therapy (EVT) utilizing thrombectomy assisted by carotid stenting was long considered risky because of procedural complexities and necessity of potent platelet inhibition-in particular following IVT. This study assesses the benefits and harms of thrombectomy assisted by carotid stenting and identifies factors associated with clinical outcome and procedural complications. Retrospective single-center analysis of 47 consecutive stroke patients with carotid occlusion or high-grade stenosis and concomitant intracranial embolus treated between September 2011 and December 2014. Benefits included early improvement of stroke severity (NIHSS ≥ 10) or complete remission within 72 h and favorable long-term outcome (mRS ≤ 2). Harms included complications during and following EVT. Mean age was 64.3 years (standard deviation ±12.5), 40 (85%) patients received IVT initially. Median NIHSS was 16 (inter-quartile range 14-19). Mean time from stroke onset to recanalization was 311 min (standard deviation ±78.0). Early clinical improvement was detected in 22 (46%) patients. Favorable outcome at 3 months occurred in 32 (68%) patients. Expedited patient management was associated with favorable clinical outcome. Two (4%) patients experienced symptomatic hemorrhage. Eight (17%) patients experienced stent thrombosis. Four (9%) patients died. Thrombectomy assisted by carotid stenting seems beneficial and reasonably safe with a promising rate of favorable outcome. Nevertheless, adverse events and complications call for additional clinical investigations prior to recommendation as clinical standard. Expeditious patient management is central to achieve good clinical outcome.

Carotid stent-assisted thrombectomy in acute ischemic stroke.

Acute carotid occlusion or near-occlusion with concomitant intracranial embolism cause severe acute ischemic strokes in patients. These concomitant occlusions have suggested poor response to intravenous thrombolysis and complicate endovascular treatment. Nevertheless, endovascular stent-assisted thrombectomy may improve outcome in patients but the treatment is not without concerns. Required antiplatelet therapy to prevent stent thrombosis may increase the rate of intracranial hemorrhage, especially after recent thrombolysis. Furthermore, technical difficulties in access of the intracranial vasculature may cause adverse events, even in the hands of experienced interventionalists. These concerns currently defy the treatment in being recommended for general use and only on a compassionate basis. However, recent patient series have suggested reasonable safety and efficacy for carotid stent-assisted thrombectomy.

Facial recognition and laser surface scan: a pilot study.

Surface scanning of the face of a suspect is presented as a way to better match the facial features with those of a perpetrator from CCTV footage. We performed a simple pilot study where we obtained facial surface scans of volunteers and then in blind trials tried to match these scans with 2D photographs of the faces of the volunteers. Fifteen male volunteers were surface scanned using a Polhemus FastSCAN Cobra Handheld Laser Scanner. Three photographs were taken of each volunteer's face in full frontal, profile and from above at an angle of 45 degrees and also 45 degrees laterally. Via special software (MIMICS and Photoshop) the surface scans were matched with the photographs in blind trials. The matches were graded as: a good fit; possible fit; and no fit. All the surface scans and photos were matched correctly, although one surface scan could be matched with two angled photographs, meaning that the discriminatory value was 86.7%. We also tested the surface scanner in terms of reliability in establishing point measures on skulls, and compared with physical measurements performed by calipers. The variation was on average 1 mm for five cranial measures. We suggest how surface scanning might be applied in forensic facial identification.