Diagnostic accuracy of an artificial intelligence algorithm versus radiologists for fracture detection on cervical spine CT

OBJECTIVES

To compare diagnostic accuracy of a deep learning artificial intelligence (AI) for cervical spine (C-spine) fracture detection on CT to attending radiologists and assess which undetected fractures were injuries in need of stabilising therapy (IST).

METHODS

This single-centre, retrospective diagnostic accuracy study included consecutive patients (age ≥18 years; 2007-2014) screened for C-spine fractures with CT. To validate ground truth, one radiologist and three neurosurgeons independently examined scans positive for fracture. Negative scans were followed up until 2022 through patient files and two radiologists reviewed negative scans that were flagged positive by AI. The neurosurgeons determined which fractures were ISTs. Diagnostic accuracy of AI and attending radiologists (index tests) were compared using McNemar.

RESULTS

Of the 2368 scans (median age, 48, interquartile range 30-65; 1441 men) analysed, 221 (9.3%) scans contained C-spine fractures with 133 IST. AI detected 158/221 scans with fractures (sensitivity 71.5%, 95% CI 65.5-77.4%) and 2118/2147 scans without fractures (specificity 98.6%, 95% CI 98.2-99.1). In comparison, attending radiologists detected 195/221 scans with fractures (sensitivity 88.2%, 95% CI 84.0-92.5%, p < 0.001) and 2130/2147 scans without fracture (specificity 99.2%, 95% CI 98.8-99.6, p = 0.07). Of the fractures undetected by AI 30/63 were ISTs versus 4/26 for radiologists. AI detected 22/26 fractures undetected by the radiologists, including 3/4 undetected ISTs.

CONCLUSION

Compared to attending radiologists, the artificial intelligence has a lower sensitivity and a higher miss rate of fractures in need of stabilising therapy; however, it detected most fractures undetected by the radiologists, including fractures in need of stabilising therapy. Clinical relevance statement The artificial intelligence algorithm missed more cervical spine fractures on CT than attending radiologists, but detected 84.6% of fractures undetected by radiologists, including fractures in need of stabilising therapy.

KEY POINTS

The impact of artificial intelligence for cervical spine fracture detection on CT on fracture management is unknown. The algorithm detected less fractures than attending radiologists, but detected most fractures undetected by the radiologists including almost all in need of stabilising therapy. The artificial intelligence algorithm shows potential as a concurrent reader.

Endovascular treatment versus no endovascular treatment after 6-24 h in patients with ischaemic stroke and collateral flow on CT angiography (MR CLEAN-LATE) in the Netherlands: a multicentre, open-label, blinded-endpoint, randomised, controlled, phase 3 trial

BACKGROUND

Endovascular treatment for anterior circulation ischaemic stroke is effective and safe within a 6 h window. MR CLEAN-LATE aimed to assess efficacy and safety of endovascular treatment for patients treated in the late window (6-24 h from symptom onset or last seen well) selected on the basis of the presence of collateral flow on CT angiography (CTA).

METHODS

MR CLEAN-LATE was a multicentre, open-label, blinded-endpoint, randomised, controlled, phase 3 trial done in 18 stroke intervention centres in the Netherlands. Patients aged 18 years or older with ischaemic stroke, presenting in the late window with an anterior circulation large-vessel occlusion and collateral flow on CTA, and a neurological deficit score of at least 2 on the National Institutes of Health Stroke Scale were included. Patients who were eligible for late-window endovascular treatment were treated according to national guidelines (based on clinical and perfusion imaging criteria derived from the DAWN and DEFUSE-3 trials) and excluded from MR CLEAN-LATE enrolment. Patients were randomly assigned (1:1) to receive endovascular treatment or no endovascular treatment (control), in addition to best medical treatment. Randomisation was web based, with block sizes ranging from eight to 20, and stratified by centre. The primary outcome was the modified Rankin Scale (mRS) score at 90 days after randomisation. Safety outcomes included all-cause mortality at 90 days after randomisation and symptomatic intracranial haemorrhage. All randomly assigned patients who provided deferred consent or died before consent could be obtained comprised the modified intention-to-treat population, in which the primary and safety outcomes were assessed. Analyses were adjusted for predefined confounders. Treatment effect was estimated with ordinal logistic regression and reported as an adjusted common odds ratio (OR) with a 95% CI. This trial was registered with the ISRCTN, ISRCTN19922220.

FINDINGS

Between Feb 2, 2018, and Jan 27, 2022, 535 patients were randomly assigned, and 502 (94%) patients provided deferred consent or died before consent was obtained (255 in the endovascular treatment group and 247 in the control group; 261 [52%] females). The median mRS score at 90 days was lower in the endovascular treatment group than in the control group (3 [IQR 2-5] vs 4 [2-6]), and we observed a shift towards better outcomes on the mRS for the endovascular treatment group (adjusted common OR 1·67 [95% CI 1·20-2·32]). All-cause mortality did not differ significantly between groups (62 [24%] of 255 patients vs 74 [30%] of 247 patients; adjusted OR 0·72 [95% CI 0·44-1·18]). Symptomatic intracranial haemorrhage occurred more often in the endovascular treatment group than in the control group (17 [7%] vs four [2%]; adjusted OR 4·59 [95% CI 1·49-14·10]).

INTERPRETATION

In this study, endovascular treatment was efficacious and safe for patients with ischaemic stroke caused by an anterior circulation large-vessel occlusion who presented 6-24 h from onset or last seen well, and who were selected on the basis of the presence of collateral flow on CTA. Selection of patients for endovascular treatment in the late window could be primarily based on the presence of collateral flow.

FUNDING

Collaboration for New Treatments of Acute Stroke consortium, Dutch Heart Foundation, Stryker, Medtronic, Cerenovus, Top Sector Life Sciences & Health, and the Netherlands Brain Foundation.

Association between thrombus composition and stroke etiology in the MR CLEAN Registry biobank

PURPOSE

The composition of thrombi retrieved during endovascular thrombectomy (EVT) in acute ischemic stroke (AIS) due to large vessel occlusion (LVO) may differ depending on their origin. In this study, we investigated the association between thrombus composition and stroke etiology in a large population of patients from the Dutch MR CLEAN Registry treated with EVT in daily clinical practice.

METHODS

The thrombi of 332 patients with AIS were histologically analyzed for red blood cells (RBC), fibrin/platelets (F/P), and white blood cells (leukocytes) using a machine learning algorithm. Stroke etiology was assessed using the Trial of Org 10,172 in acute stroke treatment (TOAST) classification.

RESULTS

The thrombi of cardioembolic origin contained less RBC and more F/P than those of non-cardioembolic origin (25.8% vs 41.2% RBC [p = 0.003] and 67.1% vs 54.5% F/P [p = 0.004]). The likelihood of a non-cardioembolic source of stroke increased with increasing thrombus RBC content (OR 1.02; [95% CI 1.00-1.06] for each percent increase) and decreased with a higher F/P content (OR 1.02; [95% CI 1.00-1.06]). Thrombus composition in patients with a cardioembolic origin and undetermined origin was similar.

CONCLUSION

Thrombus composition is significantly associated with stroke etiology, with an increase in RBC and a decrease in F/P raising the odds for a non-cardioembolic cause. No difference between composition of cardioembolic thrombi and of undetermined origin was seen. This emphasizes the need for more extensive monitoring for arrhythmias and/or extended cardiac analysis in case of an undetermined origin.

Estimation of treatment effects in observational stroke care data: comparison of statistical approaches

Introduction

Various statistical approaches can be used to deal with unmeasured confounding when estimating treatment effects in observational studies, each with its own pros and cons. This study aimed to compare treatment effects as estimated by different statistical approaches for two interventions in observational stroke care data.

Patients and methods

We used prospectively collected data from the MR CLEAN registry including all patients (n = 3279) with ischemic stroke who underwent endovascular treatment (EVT) from 2014 to 2017 in 17 Dutch hospitals. Treatment effects of two interventions – i.e., receiving an intravenous thrombolytic (IVT) and undergoing general anesthesia (GA) before EVT – on good functional outcome (modified Rankin Scale ≤2) were estimated. We used three statistical regression-based approaches that vary in assumptions regarding the source of unmeasured confounding: individual-level (two subtypes), ecological, and instrumental variable analyses. In the latter, the preference for using the interventions in each hospital was used as an instrument.

Results

Use of IVT (range 66–87%) and GA (range 0–93%) varied substantially between hospitals. For IVT, the individual-level (OR ~ 1.33) resulted in significant positive effect estimates whereas in instrumental variable analysis no significant treatment effect was found (OR 1.11; 95% CI 0.58–1.56). The ecological analysis indicated no statistically significant different likelihood (β = − 0.002%; P = 0.99) of good functional outcome at hospitals using IVT 1% more frequently. For GA, we found non-significant opposite directions of points estimates the treatment effect in the individual-level (ORs ~ 0.60) versus the instrumental variable approach (OR = 1.04). The ecological analysis also resulted in a non-significant negative association (0.03% lower probability).

Discussion and conclusion

Both magnitude and direction of the estimated treatment effects for both interventions depend strongly on the statistical approach and thus on the source of (unmeasured) confounding. These issues should be understood concerning the specific characteristics of data, before applying an approach and interpreting the results. Instrumental variable analysis might be considered when unobserved confounding and practice variation is expected in observational multicenter studies.

Improvements in Endovascular Treatment for Acute Ischemic Stroke: A Longitudinal Study in the MR CLEAN Registry

We evaluated data from all patients in the Netherlands who underwent endovascular treatment for acute ischemic stroke in the past 3.5 years, to identify nationwide trends in time to treatment and procedural success, and assess their effect on clinical outcomes.

A Randomized Trial of Intravenous Alteplase before Endovascular Treatment for Stroke

BACKGROUND

The value of administering intravenous alteplase before endovascular treatment (EVT) for acute ischemic stroke has not been studied extensively, particularly in non-Asian populations.

METHODS

We performed an open-label, multicenter, randomized trial in Europe involving patients with stroke who presented directly to a hospital that was capable of providing EVT and who were eligible for intravenous alteplase and EVT. Patients were randomly assigned in a 1:1 ratio to receive EVT alone or intravenous alteplase followed by EVT (the standard of care). The primary end point was functional outcome on the modified Rankin scale (range, 0 [no disability] to 6 [death]) at 90 days. We assessed the superiority of EVT alone over alteplase plus EVT, as well as noninferiority by a margin of 0.8 for the lower boundary of the 95% confidence interval for the odds ratio of the two trial groups. Death from any cause and symptomatic intracerebral hemorrhage were the main safety end points.

RESULTS

The analysis included 539 patients. The median score on the modified Rankin scale at 90 days was 3 (interquartile range, 2 to 5) with EVT alone and 2 (interquartile range, 2 to 5) with alteplase plus EVT. The adjusted common odds ratio was 0.84 (95% confidence interval [CI], 0.62 to 1.15; P = 0.28), which showed neither superiority nor noninferiority of EVT alone. Mortality was 20.5% with EVT alone and 15.8% with alteplase plus EVT (adjusted odds ratio, 1.39; 95% CI, 0.84 to 2.30). Symptomatic intracerebral hemorrhage occurred in 5.9% and 5.3% of the patients in the respective groups (adjusted odds ratio, 1.30; 95% CI, 0.60 to 2.81).

CONCLUSIONS

In a randomized trial involving European patients, EVT alone was neither superior nor noninferior to intravenous alteplase followed by EVT with regard to disability outcome at 90 days after stroke. The incidence of symptomatic intracerebral hemorrhage was similar in the two groups. (Funded by the Collaboration for New Treatments of Acute Stroke consortium and others; MR CLEAN-NO IV ISRCTN number, ISRCTN80619088.).

Renal nerve stimulation: complete versus incomplete renal sympathetic denervation

PURPOSE

Blood pressure (BP) reduction after renal sympathetic denervation (RDN) is highly variable. Renal nerve stimulation (RNS) can localize sympathetic nerves. The RNS trial aimed to investigate the medium-term BP-lowering effects of the use of RNS during RDN, and explore if RNS can check the completeness of the denervation.

MATERIAL AND METHODS

Forty-four treatment-resistant hypertensive patients were included in the prospective, single-center RNS trial. The primary study endpoint was change in 24-h BP at 6- to 12-month follow-up after RDN. The secondary study endpoints were the acute procedural RNS-induced BP response before and after RDN; number of antihypertensive drugs at follow-up; and the correlation between the RNS-induced BP increase before versus after RDN (delta [Δ] RNS-induced BP).

RESULTS

Before RDN, the RNS-induced systolic BP rise was 43(±21) mmHg, and decreased to 9(±12) mmHg after RDN (p < 0.001). Mean 24-h systolic/diastolic BP decreased from 147(±12)/82(±11) mmHg at baseline to 135(±11)/76(±10) mmHg (p < 0.001/<0.001) at follow-up (10 [6-12] months), with 1 antihypertensive drug less compared to baseline. The Δ RNS-induced BP and the 24-h BP decrease at follow-up were correlated for systolic (R = 0.44, p = 0.004) and diastolic (R = 0.48, p = 0.003) BP. Patients with ≤0 mmHg residual RNS-induced BP response after RDN had a significant lower mean 24-h systolic BP at follow-up compared to the patients with >0 mmHg residual RNS-induced BP response (126 ± 4 mmHg versus 135 ± 10 mmHg, p = 0.04). 83% of the patients with ≤0 mmHg residual RNS-induced BP response had normal 24-h BP at follow-up, compared to 33% in the patients with >0 mmHg residual RNS-induced BP response (p = 0.023).

CONCLUSION

The use of RNS during RDN leads to clinically significant and sustained lowering of 24-h BP with fewer antihypertensive drugs at follow-up. RNS-induced BP changes were correlated with 24-h BP changes at follow-up. Moreover, patients with complete denervation had significant lower BP compared to the patients with incomplete denervation.

Endovascular Treatment for Acute Ischemic Stroke in Children: Experience From the MR CLEAN Registry

BACKGROUND AND PURPOSE

Multiple trials have shown the efficacy and safety of endovascular therapy (EVT) of acute ischemic stroke in adults. Trials in children are lacking and only case reports and case series exist. However, the long-term outcome of children with acute ischemic stroke can be devastating with significant mortality and morbidity. In this study, we describe the safety and efficacy of EVT in children with anterior circulation acute ischemic stroke who were included in the MR CLEAN Registry (Multicenter Randomized Clinical Trial of Endovascular Treatment for Acute Ischemic Stroke in the Netherlands).

METHODS

Patients under the age of 18 years who were treated with EVT for acute ischemic stroke between March 2014 and July 2017 were retrospectively reviewed up to 6 months after EVT. Nine children, aged 13 months to 16 years (median 14 years, interquartile range, 3-15 years), underwent EVT. Stroke cause was thromboembolism in children with end-stage heart failure on left ventricular assist device (4 of these 9 cases). Median time from onset to imaging was 133 minutes. Four children received intravenous alteplase before EVT, with median onset to needle time of 165 minutes. In all but one patient, EVT was technically successful. No major periprocedural complications occurred.

RESULTS

At 24 hours after EVT, 3 children completely recovered and 4 children showed partial recovery (median National Institutes of Health Stroke Scale, 3.5), whereas 2 patients on left ventricular assist device died within the first week due to the occurrence of multiple strokes. One patient on left ventricular assist device developed a fatal massive intracranial hemorrhage and another child died due to left ventricular assist device-related complications. Among the 5 stroke survivors, all had a favorable outcome (modified Rankin Scale score, 0-2) at 6 months follow-up.

CONCLUSIONS

EVT of children with acute ischemic stroke seems safe and feasible. However, these findings should be interpreted with caution as more and larger studies are needed to clarify the trade-off between risks and benefits of this treatment.

Blood Pressure During Endovascular Treatment Under Conscious Sedation or Local Anesthesia

Objective

To evaluate the role of blood pressure (BP) as mediator of the effect of conscious sedation (CS) compared to local anesthesia (LA) on functional outcome after endovascular treatment (EVT).

Methods

Patients treated in the Multicenter Randomized Clinical Trial of Endovascular Treatment for Acute Ischemic Stroke in the Netherlands (MR CLEAN) Registry centers with CS or LA as preferred anesthetic approach during EVT for ischemic stroke were analyzed. First, we evaluated the effect of CS on area under the threshold (AUT), relative difference between baseline and lowest procedural mean arterial pressure (∆LMAP), and procedural BP trend, compared to LA. Second, we assessed the association between BP and functional outcome (modified Rankin Scale [mRS]) with multivariable regression. Lastly, we evaluated whether BP explained the effect of CS on mRS.

Results

In 440 patients with available BP data, patients treated under CS (n = 262) had larger AUTs (median 228 vs 23 mm Hg*min), larger ∆LMAP (median 16% vs 6%), and a more negative BP trend (−0.22 vs −0.08 mm Hg/min) compared to LA (n = 178). Larger ∆LMAP and AUTs were associated with worse mRS (adjusted common odds ratio [acOR] per 10% drop 0.87, 95% confidence interval [CI] 0.78–0.97, and acOR per 300 mm Hg*min 0.89, 95% CI 0.82–0.97). Patients treated under CS had worse mRS compared to LA (acOR 0.59, 95% CI 0.40–0.87) and this association remained when adjusting for ∆LMAP and AUT (acOR 0.62, 95% CI 0.42–0.92).

Conclusions

Large BP drops are associated with worse functional outcome. However, BP drops do not explain the worse outcomes in the CS group.

Treatment of atrial fibrillation in patients with enhanced sympathetic tone by pulmonary vein isolation or pulmonary vein isolation and renal artery denervation: clinical background and study design : The ASAF trial: ablation of sympathetic atrial fibrillation

BACKGROUND

Hypertension is an important, modifiable risk factor for the development of atrial fibrillation (AF). Even after pulmonary vein isolation (PVI), 20-40% experience recurrent AF. Animal studies have shown that renal denervation (RDN) reduces AF inducibility. One clinical study with important limitations suggested that RDN additional to PVI could reduce recurrent AF.

OBJECTIVE

The goal of this multicenter randomized controlled study is to investigate whether RDN added to PVI reduces AF recurrence.

METHODS

The main end point is the time until first AF recurrence according to EHRA guidelines after a blanking period of 3 months. Assuming a 12-month accrual period and 12 months of follow-up, a power of 0.80, a two-sided alpha of 0.05 and an expected drop-out of 10% per group, 69 patients per group are required. We plan to randomize a total of 138 hypertensive patients with AF and signs of sympathetic overdrive in a 1:1 fashion. Patients should use at least two antihypertensive drugs. Sympathetic overdrive includes obesity, exercise-induced excessive blood pressure (BP) increase, significant white coat hypertension, hospital admission or fever induced AF, tachycardia induced AF and diabetes mellitus. The interventional group will undergo PVI + RDN and the control group will undergo PVI.

RESULTS

Patients will have follow-up for 1 year, and continuous loop monitoring is advocated.

CONCLUSION

This randomized, controlled study will elucidate if RDN on top of PVI reduces AF recurrence.

Time to Reperfusion and Treatment Effect for Acute Ischemic Stroke: A Randomized Clinical Trial

IMPORTANCE

Intra-arterial treatment (IAT) for acute ischemic stroke caused by intracranial arterial occlusion leads to improved functional outcome in patients treated within 6 hours after onset. The influence of treatment delay on treatment effect is not yet known.

OBJECTIVE

To evaluate the influence of time from stroke onset to the start of treatment and from stroke onset to reperfusion on the effect of IAT.

DESIGN, SETTING, AND PARTICIPANTS

The Multicenter Randomized Clinical Trial of Endovascular Treatment of Acute Ischemic Stroke in the Netherlands (MR CLEAN) was a multicenter, randomized clinical open-label trial of IAT vs no IAT in 500 patients. The time to the start of treatment was defined as the time from onset of symptoms to groin puncture (TOG). The time from onset of treatment to reperfusion (TOR) was defined as the time to reopening the vessel occlusion or the end of the procedure in cases for which reperfusion was not achieved. Data were collected from December 3, 2010, to June 3, 2014, and analyzed (intention to treat) from July 1, 2014, to September 19, 2015.

MAIN OUTCOMES AND MEASURES

Main outcome was the modified Rankin Scale (mRS) score for functional outcome (range, 0 [no symptoms] to 6 [death]). Multiple ordinal logistic regression analysis estimated the effect of treatment and tested for the interaction of time to randomization, TOG, and TOR with treatment. The effect of treatment as a risk difference on reaching independence (mRS score, 0-2) was computed as a function of TOG and TOR. Calculations were adjusted for age, National Institutes of Health Stroke Scale score, previous stroke, atrial fibrillation, diabetes mellitus, and intracranial arterial terminus occlusion.

RESULTS

Among 500 patients (58% male; median age, 67 years), the median TOG was 260 (interquartile range [IQR], 210-311) minutes; median TOR, 340 (IQR, 274-395) minutes. An interaction between TOR and treatment (P = .04) existed, but not between TOG and treatment (P = .26). The adjusted risk difference (95% CI) was 25.9% (8.3%-44.4%) when reperfusion was reached at 3 hours, 18.8% (6.6%-32.6%) at 4 hours, and 6.7% (0.4%-14.5%) at 6 hours.

CONCLUSION AND RELEVANCE

For every hour of reperfusion delay, the initially large benefit of IAT decreases; the absolute risk difference for a good outcome is reduced by 6% per hour of delay. Patients with acute ischemic stroke require immediate diagnostic workup and IAT in case of intracranial arterial vessel occlusion.

TRIAL REGISTRATION

trialregister.nl Identifier: NTR1804.

A randomized trial of intraarterial treatment for acute ischemic stroke

BACKGROUND

In patients with acute ischemic stroke caused by a proximal intracranial arterial occlusion, intraarterial treatment is highly effective for emergency revascularization. However, proof of a beneficial effect on functional outcome is lacking.

METHODS

We randomly assigned eligible patients to either intraarterial treatment plus usual care or usual care alone. Eligible patients had a proximal arterial occlusion in the anterior cerebral circulation that was confirmed on vessel imaging and that could be treated intraarterially within 6 hours after symptom onset. The primary outcome was the modified Rankin scale score at 90 days; this categorical scale measures functional outcome, with scores ranging from 0 (no symptoms) to 6 (death). The treatment effect was estimated with ordinal logistic regression as a common odds ratio, adjusted for prespecified prognostic factors. The adjusted common odds ratio measured the likelihood that intraarterial treatment would lead to lower modified Rankin scores, as compared with usual care alone (shift analysis).

RESULTS

We enrolled 500 patients at 16 medical centers in The Netherlands (233 assigned to intraarterial treatment and 267 to usual care alone). The mean age was 65 years (range, 23 to 96), and 445 patients (89.0%) were treated with intravenous alteplase before randomization. Retrievable stents were used in 190 of the 233 patients (81.5%) assigned to intraarterial treatment. The adjusted common odds ratio was 1.67 (95% confidence interval [CI], 1.21 to 2.30). There was an absolute difference of 13.5 percentage points (95% CI, 5.9 to 21.2) in the rate of functional independence (modified Rankin score, 0 to 2) in favor of the intervention (32.6% vs. 19.1%). There were no significant differences in mortality or the occurrence of symptomatic intracerebral hemorrhage.

CONCLUSIONS

In patients with acute ischemic stroke caused by a proximal intracranial occlusion of the anterior circulation, intraarterial treatment administered within 6 hours after stroke onset was effective and safe. (Funded by the Dutch Heart Foundation and others; MR CLEAN Netherlands Trial Registry number, NTR1804, and Current Controlled Trials number, ISRCTN10888758.).