Patient-reported health data, especially patient-reported outcomes measures, are vital for improving clinical care but are often limited by memory bias, cognitive load, and inflexible questionnaires. Patients prefer conversational symptom reporting, highlighting the need for robust methods in symptom extraction and conversational intelligence. This study presents a weakly-supervised pipeline for training and evaluating medical Named Entity Recognition (NER) models across eight languages, with a focus on low-resource settings. A merged English medical corpus, annotated using the Stanza i2b2 model, was translated into German, Greek, Spanish, Italian, Portuguese, Polish, and Slovenian, preserving the entity annotations medical problems, diagnostic tests, and treatments. Data augmentation addressed the class imbalance, and the fine-tuned BERT-based models outperformed baselines consistently. The English model achieved the highest F1 score (80.07%), followed by German (78.70%), Spanish (77.61%), Portuguese (77.21%), Slovenian (75.72%), Italian (75.60%), Polish (75.56%), and Greek (69.10%). Compared to the existing baselines, our models demonstrated notable performance gains, particularly in English, Spanish, and Italian. This research underscores the feasibility and effectiveness of weakly-supervised multilingual approaches for medical entity extraction, contributing to improved information access in clinical narratives—especially in under-resourced languages.

Background: The aim of this study is to gain a comprehensive understanding of the latest advancements in breast cancer recurrence markers, with the aim of identifying minimally invasive or minimally intrusive markers as necessary approach for screening for breast cancer recurrence. Methods: We followed PRISMA guidelines, systematically searching Web of Science, Scopus, and PubMed from 2010 to December 2023 for secondary papers on breast cancer markers of recurrence. Keywords used to search the databases include but are not limited to: “breast cancer recurrence”, “markers”, “radiology”, “pathology”, “clinical features”. Studies focusing solely on outcomes after recurrence, such as survival or treatment response, were excluded to ensure the review targeted markers relevant to early prediction. The search was limited to English language. Selected papers underwent screening process according to inclusion/exclusion criteria, and data extraction included publication details, markers, marker modality, among others. Results: The number of papers considered for this review was 1,138. After two phases of screening process, a total number of 28 reviews were included in this scoping review. We have categorized markers into radiological, clinical, and histopathological types. Among the most relevant clinical markers correlated with breast cancer (BC) recurrence are clinical stage, carcinoembryogenic antigen (CEA), and cancer antigen 15.3 (CA 15.3). We have also identified that the following radiological markers are the most mentioned markers associated with recurrence: mammographic density (MD), tumor heterogeneity, most enhancing tumor volume (METV), radiomic features, and more. Furthermore, we identified nuclear grade, microenvironment heterogeneity, estrogen receptor (ER), androgen receptor (AR), human epidermal growth factor receptor 2 (HER2), Ki-67 antigen, as the most significant histopathological markers of breast cancer recurrence. Conclusion: This review identified promising markers for breast cancer recurrence in three categories: clinical, radiological and histopathological. General practitioners can leverage these insights for enhanced pre-screening, aiding in earlier detection and intervention, thus improving patient outcomes. Unclear cut-off values and disagreement on their use remain obstacles.

Patient-reported health data, especially Patient-Reported Outcomes Measures, are vital for improving clinical care but are often limited by memory bias, cognitive load, and inflexible questionnaires. Patients prefer conversational symptom reporting, highlighting the need for robust methods in symptom extraction and conversational intelligence. This study presents a weakly-supervised pipeline for training and evaluating medical Named Entity Recognition (NER) models across eight languages, with a focus on low-resource settings. A merged English medical corpus, annotated using the Stanza i2b2 model, was translated into German, Greek, Spanish, Italian, Portuguese, Polish, and Slovenian, preserving entity annotations (PROBLEM, TEST, TREATMENT). Data augmentation addressed class imbalance, and fine-tuned BERT-based models consistently outperformed baselines. The English model achieved the highest F1 score (80.07%), followed by German (78.70%), Spanish (77.61%), Portuguese (77.21%), Slovenian (75.72%), Italian (75.60%), Polish (75.56%), and Greek (69.10%). Compared to existing baselines, our models demonstrated notable performance gains, particularly in English, Spanish, and Italian. This research underscores the feasibility and effectiveness of weakly supervised, multilingual approaches for medical entity extraction, contributing to improved information access in clinical narratives—especially in under-resourced languages.