Prediction of Obliteration After the Gamma Knife Radiosurgery of Arteriovenous Malformations Using Hand-Crafted Radiomics and Deep-Learning Methods.

INTRODUCTION: Brain arteriovenous malformations (bAVMs) are vascular abnormalities that can be treated with embolization or radiotherapy to prevent the risk of future rupture. In this study, we use hand-crafted radiomics and deep learning techniques to predict favorable vs. unfavorable outcomes following Gamma Knife radiosurgery (GKRS) of bAVMs and compare their prediction performances. METHODS: One hundred twenty-six patients seen at one academic medical center for GKRS obliteration of bAVMs over 15 years were retrospectively reviewed. Forty-two patients met the inclusion criteria. Favorable outcomes were defined as complete nidus obliteration demonstrated on cerebral angiogram and asymptomatic recovery. Unfavorable outcomes were defined as incomplete obliteration or complications relating to the AVM that developed after GKRS. Outcome predictions were made using a random forest model with hand-crafted radiomic features and a fine-tuned ResNet-34 convolutional neural network (CNN) model. The performance was evaluated by using a ten-fold cross-validation technique. RESULTS: The average accuracy and area-under-curve (AUC) values of the Random Forest Classifier (RFC) with radiomics features were 68.5 ±9.80% and 0.705 ±0.086, whereas those of the ResNet-34 model were 60.0 ±11.9% and 0.694 ±0.124. Four radiomics features used with RFC discriminated unfavorable response cases from favorable response cases with statistical significance. When cropped images were used with ResNet-34, the accuracy and AUC decreased to 59.3 ± 14.2% and 55.4 ±10.4%, respectively. CONCLUSIONS: A hand-crafted radiomics model and a pre-trained CNN model can be fine-tuned on pre-treatment MRI scans to predict clinical outcomes of AVM patients undergoing GKRS with equivalent prediction performance. The outcome predictions are promising but require further external validation on more patients.

Undifferentiated psychosis or schizophrenia associated with vermis-predominant cerebellar hypoplasia.

Schizophrenia (SCZ) is a well-studied neuropsychiatric condition that has been shown to have a high degree of genetic heritability. Still, little data on the specific genetic risk variants associated with the disease exists. Classification of the SCZ phenotype into SCZ-related endophenotypes is a promising methodology to parse out and elucidate the specific genetic risk variants for each. Here, we present a series of 17 previously reported individuals and a new proband with similar SCZ-related neuropsychiatric characteristics and shared brain imaging findings. Unsurprisingly, these individuals shared classic psychiatric features of SCZ. Interestingly, we also identified shared neuropsychiatric features in this series of individuals that had not been highlighted previously. A consistently decreased IQ, memory impairment, sleep and speech disturbances, and attention deficits were commonly reported findings. The brain imaging findings among these individuals also consistently showed posterior vermis predominant cerebellar hypoplasia (CBLH-V). Most individuals' diagnoses were initially described as Dandy-Walker malformation; however, our independent review of imaging suggests a more consistent pattern of posterior vermis predominant cerebellar hypoplasia rather than true Dandy-Walker malformation. While the specific genetic risk variants for this endophenotype are yet to be described, the aim of this paper is to present the shared neuropsychiatric features and consistent, symmetrical brain image findings which suggest that this subset of individuals comprises an endophenotype of SCZ with a high genetic solve rate.

Equipment Setup and Artifact Removal for Simultaneous Electroencephalogram and Functional Magnetic Resonance Imaging for Clinical Review in Epilepsy.

Simultaneous electroencephalogram and functional magnetic resonance imaging (EEG-fMRI) is a unique combined technique that provides synergy in the understanding and localization of seizure onset in epilepsy. However, reported experimental protocols for EEG-fMRI recordings fail to address details about conducting such procedures on epilepsy patients. In addition, these protocols are limited solely to research settings. To fill the gap between patient monitoring in an epilepsy monitoring unit (EMU) and conducting research with an epilepsy patient, we introduce a unique EEG-fMRI recording protocol of epilepsy during the interictal period. The use of an MR conditional electrode set, which can also be used in the EMU for a simultaneous scalp EEG and video recording, allows an easy transition of EEG recordings from the EMU to the scanning room for concurrent EEG-fMRI recordings. Details on the recording procedures using this specific MR conditional electrode set are provided. In addition, the study explains step-by-step EEG processing procedures to remove the imaging artifacts, which can then be used for clinical review. This experimental protocol promotes an amendment to the conventional EEG-fMRI recording for enhanced applicability in both clinical (i.e., EMU) and research settings. Furthermore, this protocol provides the potential to expand this modality to postictal EEG-fMRI recordings in the clinical setting.

COO and MYC/BCL2 status do not predict outcome among patients with stage I/II DLBCL: a retrospective multicenter study.

In advanced-stage diffuse large B-cell lymphoma (DLBCL), the presence of an activated B-cell phenotype or a non-germinal center (GCB) phenotype, coexpression of MYC and BCL2 by immunohistochemistry, and the cooccurrence of MYC and BCL2 or BCL6 rearrangements are associated with inferior outcomes. It is unclear whether these variables remain prognostic in stage I/II patients. In this retrospective study, we evaluated the prognostic impact of cell of origin (COO), as well as dual-expressor (DE) status and molecular double-hit (DH) status, in stage I/II DLBCL by positron emission tomography with computed tomography (PET-CT). A total of 211 patients treated with R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone)-like regimens, with or without radiotherapy, was included. The median follow-up in the entire cohort was 4 years (range, 0.4-9.4), with estimated 4-year progression-free survival (PFS) and overall survival (OS) rates of 85% (95% confidence interval [CI], 79-89) and 88% (95% CI, 83-92), respectively. By univariable analysis, DE (PFS: hazard ratio [HR], 1.27; 95% CI, 0.58-2.81, P = .55 and OS: HR, 1.40; 95% CI, 0.60-3.30; P = .44), DH (PFS: HR, 1.21; 95% CI, 0.27-5.31; P = .80 and OS: HR, 0.61; 95% CI, 0.08-4.73; P = .64), and non-GCB status (PFS: HR, 1.59; 95% CI, 0.83-3.03; P = .16 and OS: HR, 1.80; 95% CI, 0.89-3.67; P = .10) were associated with poorer outcomes. In patients with PET-CT-defined stage I/II DLBCL treated with R-CHOP-like therapy, with or without radiation, COO and DE and DH status were not significantly associated with inferior PFS or OS.

Status epilepticus alert reduces time to administration of second-line antiseizure medications.

BACKGROUND: Status epilepticus (SE) is a neurologic emergency with high morbidity and mortality. Delays in SE treatment are common in clinical practice and can be associated with poorer outcomes. Our goal was to determine whether the implementation of an SE alert protocol improves time to administration of a second-line antiseizure medication (ASM) in hospitalized adults. METHODS: We developed and implemented an inpatient SE alert system. A quasiexperimental cohort study was performed. We analyzed all patients aged 18-85 years who were managed at the University of Kentucky Medical Center using the SE alert protocol between March 2015 and June 2017 (n = 19). Controls were the first 20 consecutive patients treated for SE over the same time period, but who were managed with usual care (i.e., without SE alert protocol). RESULTS: Time to administration of a second-line ASM was shorter with the use of the SE alert system (22.21 ± 3.44 minutes) compared to usual care (58.30 ± 6.72 minutes; p < 0.0001). CONCLUSION: Implementation of an SE alert system led to a marked improvement in time to administration of a second-line ASM. CLASSIFICATION OF EVIDENCE: This study provides Class III evidence that for adult inpatients treated for SE, implementation of an SE alert protocol reduces time to administration of second-line ASM.