Combined molecular subtyping, grading, and segmentation of glioma using multi-task deep learning.

BACKGROUND: Accurate characterization of glioma is crucial for clinical decision making. A delineation of the tumor is also desirable in the initial decision stages but is time-consuming. Previously, deep learning methods have been developed that can either non-invasively predict the genetic or histological features of glioma, or that can automatically delineate the tumor, but not both tasks at the same time. Here, we present our method that can predict the molecular subtype and grade, while simultaneously providing a delineation of the tumor. METHODS: We developed a single multi-task convolutional neural network that uses the full 3D, structural, preoperative MRI scans to predict the IDH mutation status, the 1p/19q co-deletion status, and the grade of a tumor, while simultaneously segmenting the tumor. We trained our method using a patient cohort containing 1508 glioma patients from 16 institutes. We tested our method on an independent dataset of 240 patients from 13 different institutes. RESULTS: In the independent test set, we achieved an IDH-AUC of 0.90, an 1p/19q co-deletion AUC of 0.85, and a grade AUC of 0.81 (grade II/III/IV). For the tumor delineation, we achieved a mean whole tumor Dice score of 0.84. CONCLUSIONS: We developed a method that non-invasively predicts multiple, clinically relevant features of glioma. Evaluation in an independent dataset shows that the method achieves a high performance and that it generalizes well to the broader clinical population. This first-of-its-kind method opens the door to more generalizable, instead of hyper-specialized, AI methods.

Safety and efficacy of aspirin, unfractionated heparin, both, or neither during endovascular stroke treatment (MR CLEAN-MED): an open-label, multicentre, randomised controlled trial.

BACKGROUND: Aspirin and unfractionated heparin are often used during endovascular stroke treatment to improve reperfusion and outcomes. However, the effects and risks of anti-thrombotics for this indication are unknown. We therefore aimed to assess the safety and efficacy of intravenous aspirin, unfractionated heparin, both, or neither started during endovascular treatment in patients with ischaemic stroke. METHODS: We did an open-label, multicentre, randomised controlled trial with a 2 × 3 factorial design in 15 centres in the Netherlands. We enrolled adult patients (ie, ≥18 years) with ischaemic stroke due to an intracranial large-vessel occlusion in the anterior circulation in whom endovascular treatment could be initiated within 6 h of symptom onset. Eligible patients had a score of 2 or more on the National Institutes of Health Stroke Scale, and a CT or MRI ruling out intracranial haemorrhage. Randomisation was done using a web-based procedure with permuted blocks and stratified by centre. Patients were randomly assigned (1:1) to receive either periprocedural intravenous aspirin (300 mg bolus) or no aspirin, and randomly assigned (1:1:1) to receive moderate-dose unfractionated heparin (5000 IU bolus followed by 1250 IU/h for 6 h), low-dose unfractionated heparin (5000 IU bolus followed by 500 IU/h for 6 h), or no unfractionated heparin. The primary outcome was the score on the modified Rankin Scale at 90 days. Symptomatic intracranial haemorrhage was the main safety outcome. Analyses were based on intention to treat, and treatment effects were expressed as odds ratios (ORs) or common ORs, with adjustment for baseline prognostic factors. This trial is registered with the International Standard Randomised Controlled Trial Number, ISRCTN76741621. FINDINGS: Between Jan 22, 2018, and Jan 27, 2021, we randomly assigned 663 patients; of whom, 628 (95%) provided deferred consent or died before consent could be asked and were included in the modified intention-to-treat population. On Feb 4, 2021, after unblinding and analysis of the data, the trial steering committee permanently stopped patient recruitment and the trial was stopped for safety concerns. The risk of symptomatic intracranial haemorrhage was higher in patients allocated to receive aspirin than in those not receiving aspirin (43 [14%] of 310 vs 23 [7%] of 318; adjusted OR 1·95 [95% CI 1·13-3·35]) as well as in patients allocated to receive unfractionated heparin than in those not receiving unfractionated heparin (44 [13%] of 332 vs 22 [7%] of 296; 1·98 [1·14-3·46]). Both aspirin (adjusted common OR 0·91 [95% CI 0·69-1·21]) and unfractionated heparin (0·81 [0·61-1·08]) led to a non-significant shift towards worse modified Rankin Scale scores. INTERPRETATION: Periprocedural intravenous aspirin and unfractionated heparin during endovascular stroke treatment are both associated with an increased risk of symptomatic intracranial haemorrhage without evidence for a beneficial effect on functional outcome. FUNDING: The Collaboration for New Treatments of Acute Stroke consortium, the Brain Foundation Netherlands, the Ministry of Economic Affairs, Stryker, Medtronic, Cerenovus, and the Dutch Heart Foundation.

Spatio-temporal deep learning for automatic detection of intracranial vessel perforation in digital subtraction angiography during endovascular thrombectomy.

Intracranial vessel perforation is a peri-procedural complication during endovascular therapy (EVT). Prompt recognition is important as its occurrence is strongly associated with unfavorable treatment outcomes. However, perforations can be hard to detect because they are rare, can be subtle, and the interventionalist is working under time pressure and focused on treatment of vessel occlusions. Automatic detection holds potential to improve rapid identification of intracranial vessel perforation. In this work, we present the first study on automated perforation detection and localization on X-ray digital subtraction angiography (DSA) image series. We adapt several state-of-the-art single-frame detectors and further propose temporal modules to learn the progressive dynamics of contrast extravasation. Application-tailored loss function and post-processing techniques are designed. We train and validate various automated methods using two national multi-center datasets (i.e., MR CLEAN Registry and MR CLEAN-NoIV Trial), and one international multi-trial dataset (i.e., the HERMES collaboration). With ten-fold cross-validation, the proposed methods achieve an area under the curve (AUC) of the receiver operating characteristic of 0.93 in terms of series level perforation classification. Perforation localization precision and recall reach 0.83 and 0.70 respectively. Furthermore, we demonstrate that the proposed automatic solutions perform at similar level as an expert radiologist.

autoTICI: Automatic Brain Tissue Reperfusion Scoring on 2D DSA Images of Acute Ischemic Stroke Patients.

The Thrombolysis in Cerebral Infarction (TICI) score is an important metric for reperfusion therapy assessment in acute ischemic stroke. It is commonly used as a technical outcome measure after endovascular treatment (EVT). Existing TICI scores are defined in coarse ordinal grades based on visual inspection, leading to inter- and intra-observer variation. In this work, we present autoTICI, an automatic and quantitative TICI scoring method. First, each digital subtraction angiography (DSA) acquisition is separated into four phases (non-contrast, arterial, parenchymal and venous phase) using a multi-path convolutional neural network (CNN), which exploits spatio-temporal features. The network also incorporates sequence level label dependencies in the form of a state-transition matrix. Next, a minimum intensity map (MINIP) is computed using the motion corrected arterial and parenchymal frames. On the MINIP image, vessel, perfusion and background pixels are segmented. Finally, we quantify the autoTICI score as the ratio of reperfused pixels after EVT. On a routinely acquired multi-center dataset, the proposed autoTICI shows good correlation with the extended TICI (eTICI) reference with an average area under the curve (AUC) score of 0.81. The AUC score is 0.90 with respect to the dichotomized eTICI. In terms of clinical outcome prediction, we demonstrate that autoTICI is overall comparable to eTICI.

MR CLEAN-NO IV: intravenous treatment followed by endovascular treatment versus direct endovascular treatment for acute ischemic stroke caused by a proximal intracranial occlusion-study protocol for a randomized clinical trial.

BACKGROUND: Endovascular treatment (EVT) has greatly improved the prognosis of acute ischemic stroke (AIS) patients with a proximal intracranial large vessel occlusion (LVO) of the anterior circulation. Currently, there is clinical equipoise concerning the added benefit of intravenous alteplase administration (IVT) prior to EVT. The aim of this study is to assess the efficacy and safety of omitting IVT before EVT in patients with AIS caused by an anterior circulation LVO. METHODS: MR CLEAN-NO IV is a multicenter randomized open-label clinical trial with blinded outcome assessment (PROBE design). Patients ≥ 18 years of age with a pre-stroke mRS < 3 with an LVO confirmed on CT angiography/MR angiography eligible for both IVT and EVT are randomized to receive either IVT (0.9 mg/kg) followed by EVT, or direct EVT in a 1:1 ratio. The primary objective is to assess superiority of direct EVT. Secondarily, non-inferiority of direct EVT compared to IVT before EVT will be explored. The primary outcome is the score on the modified Rankin Scale at 90 days. Ordinal regression with adjustment for prognostic variables will be used to estimate treatment effect. Secondary outcomes include reperfusion graded with the eTICI scale after EVT and stroke severity (National Institutes of Health Stroke Scale) at 24 h. Safety outcomes include intracranial hemorrhages scored according to the Heidelberg criteria. A total of 540 patients will be included. DISCUSSION: IVT prior to EVT might facilitate early reperfusion before EVT or improved reperfusion rates during EVT. Conversely, among other potential adverse effects, the increased risk of bleeding could nullify the beneficial effects of IVT. MR CLEAN-NO IV will provide insight into whether IVT is still of added value in patients eligible for EVT. TRIAL REGISTRATION: www.isrctn.com : ISRCTN80619088 . Registered on 31 October 2017.

Topographical Mapping of 436 Newly Diagnosed IDH Wildtype Glioblastoma With vs. Without MGMT Promoter Methylation.

Introduction: O6 -methylguanine-methyltransferase (MGMT) promoter methylation and isocitrate dehydrogenase (IDH) mutation status are important prognostic factors for patients with glioblastoma. There are conflicting reports about a differential topographical distribution of glioblastoma with vs. without MGMT promoter methylation, possibly caused by molecular heterogeneity in glioblastoma populations. We initiated this study to re-evaluate the topographical distribution of glioblastoma with vs. without MGMT promoter methylation in light of the updated WHO 2016 classification. Methods: Preoperative T2-weighted/FLAIR and postcontrast T1-weighted MRI scans of patients aged 18 year or older with IDH wildtype glioblastoma were collected. Tumors were semi-automatically segmented, and the topographical distribution between glioblastoma with vs. without MGMT promoter methylation was visualized using frequency heatmaps. Then, voxel-wise differences were analyzed using permutation testing with Threshold Free Cluster Enhancement. Results: Four hundred thirty-six IDH wildtype glioblastoma patients were included; 211 with and 225 without MGMT promoter methylation. Visual examination suggested that when compared with MGMT unmethylated glioblastoma, MGMT methylated glioblastoma were more frequently located near bifrontal and left occipital periventricular area and less frequently near the right occipital periventricular area. Statistical analyses, however, showed no significant difference in topographical distribution between MGMT methylated vs. MGMT unmethylated glioblastoma. Conclusions: This study re-evaluated the topographical distribution of MGMT promoter methylation in 436 newly diagnosed IDH wildtype glioblastoma, which is the largest homogenous IDH wildtype glioblastoma population to date. There was no statistically significant difference in anatomical localization between MGMT methylated vs. unmethylated IDH wildtype glioblastoma.

Predicting the 1p/19q Codeletion Status of Presumed Low-Grade Glioma with an Externally Validated Machine Learning Algorithm.

PURPOSE: Patients with 1p/19q codeleted low-grade glioma (LGG) have longer overall survival and better treatment response than patients with 1p/19q intact tumors. Therefore, it is relevant to know the 1p/19q status. To investigate whether the 1p/19q status can be assessed prior to tumor resection, we developed a machine learning algorithm to predict the 1p/19q status of presumed LGG based on preoperative MRI. EXPERIMENTAL DESIGN: Preoperative brain MR images from 284 patients who had undergone biopsy or resection of presumed LGG were used to train a support vector machine algorithm. The algorithm was trained on the basis of features extracted from post-contrast T1-weighted and T2-weighted MR images and on patients' age and sex. The performance of the algorithm compared with tissue diagnosis was assessed on an external validation dataset of MR images from 129 patients with LGG from The Cancer Imaging Archive (TCIA). Four clinical experts also predicted the 1p/19q status of the TCIA MR images. RESULTS: The algorithm achieved an AUC of 0.72 in the external validation dataset. The algorithm had a higher predictive performance than the average of the neurosurgeons (AUC 0.52) but lower than that of the neuroradiologists (AUC of 0.81). There was a wide variability between clinical experts (AUC 0.45-0.83). CONCLUSIONS: Our results suggest that our algorithm can noninvasively predict the 1p/19q status of presumed LGG with a performance that on average outperformed the oncological neurosurgeons. Evaluation on an independent dataset indicates that our algorithm is robust and generalizable.

Stereotactic radiosurgery for brain AVMs: role of interobserver variation in target definition on digital subtraction angiography.

PURPOSE: We evaluated the extent of interobserver variation in contouring arteriovenous malformations (AVMs) on digital subtraction angiography (DSA) with respect to volume, spatial localization, and dosimetry and correlated our findings with the clinical outcome. METHODS AND MATERIALS: Thirty-one patients who had undergone radiosurgery for brain AVMs were studied. Six clinicians independently contoured the nidus on the original DSA. As a measure of variation, the ratio between the volumes of agreement and the corresponding encompassing volumes, as well as the absolute positional shift between the individual target volumes were derived. Using the original treatment plan, the dosimetric coverage of the individually contoured volumes with standard collimators was compared with a similar plan using dynamic conformal arcs. RESULTS: The mean contoured nidus volume was 3.6 +/- 5.6 cm3. The mean agreement ratio was 0.45 +/- 0.18 for all possible pairs of observers. The mean absolute positional shift between individually contoured volumes was 2.8 +/- 2.6 mm. These differences were more marked in previously treated groups and tended to be more pronounced in those with treatment failure. The mean coverage of the individual volumes by the 80% prescription isodose was 88.1% +/- 3.2% using conventional collimators and 78.9% +/- 4.4% using dynamic conformal arcs (p = 0.001). CONCLUSION: Substantial interobserver variations exist when contouring brain AVMs on DSA for the purpose of radiosurgical planning. Such variations may result in underdosage to the AVM and, thereby, contribute to treatment failure. The consequences of contouring variations may increase with the use of more conformal radiosurgical techniques.

Spontaneous regression of brain arteriovenous malformations--a clinical study and a systematic review of the literature.

OBJECTIVE AND IMPORTANCE: Complete spontaneous obliteration of a brain arteriovenous malformation (AVM) is a rare event, with 67 angiographically proven cases in the world literature. We present a new case and a systematic literature review to determine possible mechanisms underlying this unusual phenomenon. CLINICAL PRESENTATION: One patient with a brain AVM was referred for radiosurgical treatment. Shortly before treatment however, complete spontaneous regression occurred. This patient had experienced a hemorrhage in the months before referral. RESULTS: We found 38 articles in which 67 cases of complete and spontaneous regression of a brain AVM were presented. Male to female ratio was 1.2, with a mean age of 37 years (range 1-81). Regression occurred in 72% without new neurological events. Median size of the nidus was 2 cm (range 1-7). There was a single arterial feeder in 46 % and a single draining vein in 59%. CONCLUSION: Spontaneous regression of a brain AVM is the result of multiple interacting factors. Intracranial hemorrhage and the presence of a single draining vein seem to play a major role in this process.

Thalamic lesions in vascular dementia: low sensitivity of fluid-attenuated inversion recovery (FLAIR) imaging.

BACKGROUND AND PURPOSE: The criteria of the National Institute of Neurological Disorders and Stroke (NINDS)-Association Internationale pour la Recherche et l'Enseignement en Neurosciences (AIREN) include thalamic lesions for the diagnosis of vascular dementia (VaD). Although studies concerning VaD and brain aging advocate the use of fluid-attenuated inversion recovery (FLAIR) or T2-weighted images (T2-WI) to detect ischemic lesions, none compared the sensitivity of these sequences to depict thalamic lesions. METHODS: We performed a blinded review of T2-WI and FLAIR images in 73 patients fulfilling the radiological part of the NINDS-AIREN criteria (mean age, 71 years; range, 49 to 83 years). This sample was drawn from a large multicenter trial on VaD and was expected to have a high prevalence of thalamic lesions. In a side-by-side review, including T1-weighted images as well, lesions were classified according to presumed underlying pathology. RESULTS: The total number of thalamic lesions was 214. Two hundred eight (97%) were detected on T2-WI, but only 117 (55%) were detected on FLAIR (chi(2)=5.1; P<0.05). Although the mean size of lesions detected on T2-WI and not on FLAIR (4.4 mm) was significantly lower than the mean size of lesions detected on both sequences (6.7 mm) (P<0.001), 5 of the 29 lesions >10 mm on T2-WI were not visible on FLAIR. FLAIR detected only 81 (51%) of the 158 probable ischemic lesions and 30 (60%) of the 50 probable microbleeds. CONCLUSIONS: FLAIR should not be used as the only T2-weighted sequence to detect thalamic lesions in patients suspected of having VaD.

Spinal-cord MRI in multiple sclerosis.

The potential of MRI of the spinal cord as a diagnostic tool in MS has recently gained much interest. Dual echo spin echo MRI is most sensitive for the detection of spinal-cord abnormalities, which range from multiple focal lesions to confluent areas of high signal intensity. In some patients, commonly those with primary progressive disease, diffuse areas of slightly increased signal intensity are found. Disappointingly, the relation between MRI findings and clinical disability is weak. Studies relating MRI findings with histopathology have revealed substantial axonal loss in the spinal cords of patients with MS, whether focal lesions are present of not. Further, diffuse cord atrophy is found in advanced MS, which may reflect axonal loss. In the diagnostic setting, spinal-cord imaging is valuable. First, asymptomatic spinal-cord lesions are very rare in disorders other than MS. For example, in a patient with equivocal brain findings such as an elderly patient with vascular-ischaemic lesions, a normal spinal-cord examination can help rule out MS. Second, presence of asymptomatic spinal lesions may help confirm a diagnosis of MS when few or no brain lesions are present.