Explore the vast range of titles available.

Similarity measures play a pivotal role in automatic techniques designed to analyse large volumes of textual data. Conventional approaches, treating texts as paradigmatic examples of unstructured data, tend to overlook their structural nuances, leading to a loss of valuable information. In this paper, we propose a novel orthographic similarity measure tailored for the semi-structured analysis of texts. We explore a graph-based representation for texts, where the graph’s structure is shaped by a hierarchical decomposition of textual discourse units. Employing the concept of edit distances, our orthographic similarity measure is computed hierarchically across all components in this textual graph, integrating precomputed similarity values among lower-level nodes. The relevance and applicability of the presented approach are illustrated by a real-world example, featuring texts that exhibit intricate interconnections among their components. The resulting similarity scores, between all different structural levels of the graph, allow for a deeper understanding of the (structural) interconnections among texts and enhances the explainability of similarity measures as well as the tools using them.

Upper gastrointestinal (GI) bleeding due to duodenal invasion is a very unusual presentation revealing the initial diagnosis of hepatocellular carcinoma (HCC), especially in patients without cirrhosis. No clear recommendations are available in this setting. A 68-year-old man was admitted to the emergency department with melena. The esophagogastroduodenoscopy (EGD) revealed an oozing hemorrhagic ulcer of the duodenal bulb (Forrest I b) secondary to an invasive, undetermined bulky liver mass that was biopsied. The histopathological examination confirmed an HCC. The patient was started on chemotherapy (Gemcitabine and Oxaliplatin) with good initial response. Nevertheless, after eight months of treatment, there was a recurrence of the ulcer bleeding and a disease progression was identified. Selective transarterial embolization (TAE) was used to control the duodenal bleeding, permitting the patient to receive immunotherapy with a long-lasting control of the disease. Our case report suggests that selective TAE is a therapeutic option that can be used to stop GI bleeding due to invasive HCC in order to allow oncological treatment.

BackgroundThe added-value of systematic biopsy (SB) in patients undergoing magnetic resonance imaging (MRI)-targeted biopsy (TB) remains unclear and the spatial distribution of positive cores relative to the MRI lesion has been poorly studied. The aim of this study was to determine the utility of perilesional biopsy in detecting clinically significant prostate cancer (csPCa).MethodsWe enrolled 505 consecutive patients that underwent SB and TB for suspicious MRI lesions (PI-RADS score 3-5) at Jules Bordet Institute between June 2016 and January 2022. Patient-specific tridimensional prostate maps were reviewed to determine the distance between systematic cores containing csPCa and the MRI index lesion. Primary outcomes were the cancer detection rate (CDR) per patient and the cumulative cancer distribution rate of positive cores for each 5 mm interval from the MRI index lesion. The secondary outcome was the identification of risk groups for the presence of csPCa beyond a 10 mm margin using the chi-square automated interaction detector (CHAID) machine learning algorithm.ResultsOverall, the CDR for csPCa of TB, SB, and combined method were 32%, 25%, and 37%, respectively. While combined method detected more csPCa compared to TB (37% vs. 32%, p < 0.001), no difference was found when TB was associated with perilesional sampling within 10 mm (37% vs. 35%, p = 0.2). The cumulative cancer distribution rate for csPCa reached 86% for the 10 mm margin. The CHAID algorithm identified three risk groups: (1) PI-RADS3 (“low-risk”), (2) PI-RADS4 or PI-RADS5 and PSA density <0.15 ng/ml (“intermediate-risk”), and (3) PI-RADS 5 and PSA density ≥0.15 ng/ml (“high-risk”). The risk of missing csPCa was 2%, 8%, and 29% for low-, intermediate- and high-risk groups, respectively. Avoiding biopsies beyond a 10 mm margin prevented the detection of 19% of non-csPCa.ConclusionsPerilesional biopsy template using a 10 mm margin seems a reasonable alternative to the combined method with a comparable detection of csPCa. Our risk stratification may further enhance the selection of patients.

PurposeThere is currently no consensus regarding the optimal number of multiparametric magnetic resonance imaging (MRI)-targeted biopsy (TB) cores and their spatial distribution within the MRI lesion. We aim to determine the number of TB cores and location needed to adequately detect csPCa.MethodsWe conducted a retrospective cohort study of 505 consecutive patients undergoing TB for positive MRI lesions defined by a PI-RADS score ≥ 3 between June 2016 and January 2022. Cores chronology and locations were prospectively recorded. The co-primary outcomes were the first core to detect clinically significant prostate cancer (csPCa) and the first highest ISUP grade group. The incremental benefit of each additional core was evaluated. Analysis was then performed by distinguishing central (cTB) and peripheral (pTB) within the MRI lesion.ResultsOverall, csPCa was detected in 37% of patients. To reach a csPCa detection rate of 95%, a 3-core strategy was required, except for patients with PI-RADS 5 lesions and those with PSA density ≥ 0.2 ng/ml/cc who benefited from a fourth TB core. At multivariable analysis, only a PSA density ≥ 0.2 ng/ml/cc was an independent predictive factor of having the highest ISUP grade group on the fourth TB cores (p = 0.03). No significant difference in the cancer detection rate was found between cTB and pTB (p = 0.9). Omitting pTB would miss 18% of all csPCa.ConclusionA 3-core strategy should be considered for TB to optimize csPCa detection with additional cores needed for PI-RADS 5 lesions and high PSA density. Biopsy cores from both central and peripheral zones are required.