Rad-SpatialNet: A Frame-based Resource for Fine-Grained Spatial Relations in Radiology Reports.

This paper proposes a representation framework for encoding spatial language in radiology based on frame semantics. The framework is adopted from the existing SpatialNet representation in the general domain with the aim to generate more accurate representations of spatial language used by radiologists. We describe Rad-SpatialNet in detail along with illustrating the importance of incorporating domain knowledge in understanding the varied linguistic expressions involved in different radiological spatial relations. This work also constructs a corpus of 400 radiology reports of three examination types (chest X-rays, brain MRIs, and babygrams) annotated with fine-grained contextual information according to this schema. Spatial trigger expressions and elements corresponding to a spatial frame are annotated. We apply BERT-based models (BERTBASE and BERTLARGE) to first extract the trigger terms (lexical units for a spatial frame) and then to identify the related frame elements. The results of BERTLARGE are decent, with F1 of 77.89 for spatial trigger extraction and an overall F1 of 81.61 and 66.25 across all frame elements using gold and predicted spatial triggers respectively. This frame-based resource can be used to develop and evaluate more advanced natural language processing (NLP) methods for extracting fine-grained spatial information from radiology text in the future.

Landscape of Genomic Alterations in IDH Wild-Type Glioblastoma Identifies PI3K as a Favorable Prognostic Factor.

PURPOSE: IDH wild-type (WT) glioblastoma (GBM) is an aggressive tumor with poor survival despite current therapies. The aim of this study was to characterize its genomic profile and determine whether a particular molecular signature is associated with improved survival outcomes. PATIENTS AND METHODS: Tumor samples from 232 patients with IDH-WT GBM were sequenced, and the landscape of genomic alterations was fully delineated. Genomics data from The Cancer Genome Atlas (TCGA) cohort were analyzed for confirmation. Association of alterations with survival was evaluated in both univariable and multivariable approaches. RESULTS: The genomic landscape of IDH-WT GBM revealed a high frequency of CDKN2A/B loss, TERT promoter mutations, PTEN loss, EGFR alteration, and TP53 mutations. Novel variants or gene mutations, such as ARID1B and MLL2, were identified. To better understand synergistic effects and facilitate decision making for precision medicine, we identified 11 pairs of gene alterations that tended to co-occur or were mutually exclusive, which were confirmed in the TCGA cohort. Survival analysis showed that genomic alterations in TP53 were associated with worse overall survival (OS). However, alterations in PI3K class I genes were associated with significantly better OS (univariable analysis: P = .002; multivariable analysis: hazard ratio [HR], 0.5785; P = .00162) and longer progression-free survival (univariable analysis: P = .0043; multivariable analysis: HR, 0.6228; P = .00913). CONCLUSION: Genomic alterations in PI3K class I are a favorable prognostic factor in IDH-WT GBM. This new prognostic biomarker may facilitate risk stratification of patients, assist in clinical trial enrollment, and provide potential therapeutic targets.

Preflight, In-Flight, and Postflight Imaging of the Cervical and Lumbar Spine in Astronauts.

BACKGROUND: Back pain is a common complaint during spaceflight that is commonly attributed to intervertebral disc swelling in microgravity. Ultrasound (US) represents the only imaging modality on the International Space Station (ISS) to assess its etiology. The present study investigated: 1) The agreement and correlation of spinal US assessments as compared to results of pre- and postflight MRI studies; and 2) the trend in intervertebral disc characteristics over the course of spaceflight to ISS. METHODS: Seven ISS astronauts underwent pre- and postflight US examinations that included anterior disc height and anterior intervertebral angles with comparison to pre- and postflight MRI results. In-flight US images were analyzed for changes in disc height and angle. Statistical analysis included repeated measures ANOVA with Bonferroni post hoc analysis, Bland-Altman plots, and Pearson correlation. RESULTS: Bland-Altman plots revealed significant disagreement between disc heights and angles for MRI and US measurements while significant Pearson correlations were found in MRI and US measurements for lumbar disc height (r2 = 0.83) and angle (r2 = 0.89), but not for cervical disc height (r2 = 0.26) or angle (r2 = 0.02). Changes in anterior intervertebral disc angle-initially increases followed by decreases-were observed in the lumbar and cervical spine over the course of the long-duration mission. The cervical spine demonstrated a loss of total disc height during in-flight assessments (∼0.5 cm). DISCUSSION: Significant disagreement but significant correlation was noted between US and MRI measurements of disc height and angle. Consistency in imaging modality is important for trending measurements and more research related to US technique is required.Harrison MF, Garcia KM, Sargsyan AE, Ebert D, Riascos-Castaneda RF, Dulchavsky SA. Preflight, in-flight, and postflight imaging of the cervical and lumbar spine in astronauts. Aerosp Med Hum Perform. 2018; 89(1):32-40.