Background/Objectives: The aim of this study was to identify predictors of temporal window failure (TWF) in transcranial color-coded duplex sonography (TCCS) based on demographical and computed tomography (CT)-based parameters such as temporal bone thickness (TBT), and to define sex-specific thresholds for predicting TWF. Methods: We retrospectively analyzed a series of adult patients who underwent cranial CT and TCCS. Bitemporal TBT was measured in nine standardized regions on CT, and mean TBT per side was calculated. Temporal bone window (TBW) quality was graded with two semiquantitative TCCS scores assessing the visibility of the middle cerebral artery (MCA), contralateral temporal bone, mesencephalon, and ipsilateral sphenoid bone. Associations between TBT, sex, age, and TBW visibility were analyzed by correlation, chi-square tests, and logistic regression. Results: 200 patients (102 men, 98 women; mean age 68 ± 16 years) were enrolled. Mean TBT was 3.1 ± 0.7 mm (right) and 3.2 ± 0.7 mm (left). TBT correlated weakly with age (r = 0.15–0.18, p < 0.05) and was higher in women (p < 0.05). Age and sex influenced TBW visibility (p < 0.05) with small effect sizes. Increased TBT strongly predicted poor TBW (β ≈ –1.7, p < 0.001). Optimal TBT cut-offs predicting adequate TBW were 3.8 mm (men) and 3.3 mm (women), maximizing specificity (men: 0.95, women: 0.85) and negative predictive value (men: 0.87, women: 0.66). Conclusions: Advanced age and female sex were both associated with TWF. CT-assessed TBT represents a robust predictor of TCCS feasibility. Implementation of sex-specific TBT threshold values may facilitate patient pre-selection and improve procedural efficiency in neurosonographic diagnostics.
26 May 2023 in Proceedings of the ACM on Management of Data
Density-based clustering aims to find groups of similar objects (i.e., clusters) in a given dataset. Applications include, e.g., process mining and anomaly detection. It comes with two user parameters (ε, MinPts) that determine the clustering result, but are typically unknown in advance. Thus, users need to interactively test various settings until satisfying clusterings are found. However, existing solutions suffer from the following limitations: (a) Ineffective pruning of expensive neighborhood computations. (b) Approximate clustering, where objects are falsely labeled noise. (c) Restricted parameter tuning that is limited to ε whereas MinPts is constant, which reduces the explorable clusterings. (d) Inflexibility in terms of applicable data types and distance functions. We propose FINEX, a linear-space index that overcomes these limitations. Our index provides exact clusterings and can be queried with either of the two parameters. FINEX avoids neighborhood computations where possible and reduces the complexities of the remaining computations by leveraging fundamental properties of density-based clusters. Hence, our solution is efficient and flexible regarding data types and distance functions. Moreover, FINEX respects the original and straightforward notion of density-based clustering. In our experiments on 12 large real-world datasets from various domains, FINEX frequently outperforms state-of-the-art techniques for exact clustering by orders of magnitude.