Search

The article in the journal "Thoracic & Cardiovascular Surgeon" discusses the histological characteristics and predictors of debris captured by the Sentinel cerebral embolic protection device during transcatheter aortic valve replacement. The study included 112 patients, with debris found in the filters of 96% of all patients. The most prevalent types of debris captured were endothelium/vascular wall, fibrin/thrombus, calcification, and foreign material. The study found differences in the size and number of captured debris among self- and balloon-expandable valves, with predilatation identified as a risk factor for embolic debris. [Extracted from the article]

This article in the journal "Thoracic & Cardiovascular Surgeon" examines the histopathologic findings of explanted MitraClip devices used in transcatheter edge-to-edge repair (TEER) of the mitral valve. The study analyzes the healing response and biocompatibility of the device on human mitral valve leaflets after failed TEER procedures. Results show that the device interaction with leaflet tissue leads to fibrous encapsulation, which adds structural stability to the repair but may reduce the likelihood of valve reconstruction, impacting treatment decisions, especially in low-risk and young patients. [Extracted from the article]

This article in the journal "Thoracic & Cardiovascular Surgeon" evaluates the performance of an open-source, offline-capable large language model in extracting data from unstructured electronic health records. The study found that the 405B model had a 100% accuracy rate, while the 70B and 8B models had slightly lower accuracy rates. The open-source model's performance was comparable to two closed-source models, suggesting its potential for use in processing health records. Further research is recommended to confirm these results with real health records. [Extracted from the article]

This article discusses the outcomes of complex surgery for Lusoria Artery in adults, a rare and surgically challenging disease. The study, conducted across multiple centers in Germany, Switzerland, and Belgium, analyzed patient data from 1993 to 2024. Results showed that surgical treatment for aberrant subclavian artery (aSA) pathology led to symptom improvement in most cases, with low mortality rates. The study highlights the importance of expertise in surgery centers for achieving successful outcomes in treating aSA. [Extracted from the article]

A 74-year-old female was found hemiplegic in a public restroom. After arrival in our stroke unit, a computed tomography (CT) was performed, and she was diagnosed with bilateral carotid artery and right vertebral artery occlusion due to an acute type A aortic dissection. The patient deteriorated quickly to a GCS of 3 and was brought to the operating room, where a salvage replacement of the ascending aorta and the proximal aortic arch was performed with unilateral antegrade cerebral perfusion. Three days later, a significant decrease in left ventricular function and increase in cardiac biomarkers were observed. Coronary CT displayed residual dissection of the aortic root, extending into the left main coronary artery. The patient underwent an intravascular ultrasound-guided stenting of the left main, resulting in total recovery of heart function. She was extubated on the fourth postoperative day, with no residual neurological impairment. This case report advocates for the proper management of patients with ATAAD with severe neurological impairment, emphasizing the importance of a robust multidisciplinary approach in their care., Competing Interests: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The author(s) declared that they were an editorial board member of Frontiers, at the time of submission. This had no impact on the peer review process and the final decision., (© 2025 Bidovec, Risteski, Stähli, Dzemali, Hofmann and Stehli.)

Objectives: We aimed to evaluate the performance of multiple large language models (LLMs) in data extraction from unstructured and semi-structured electronic health records., Methods: 50 synthetic medical notes in English, containing a structured and an unstructured part, were drafted and evaluated by domain experts, and subsequently used for LLM-prompting. 18 LLMs were evaluated against a baseline transformer-based model. Performance assessment comprised four entity extraction and five binary classification tasks with a total of 450 predictions for each LLM. LLM-response consistency assessment was performed over three same-prompt iterations., Results: Claude 3.0 Opus, Claude 3.0 Sonnet, Claude 2.0, GPT 4, Claude 2.1, Gemini Advanced, PaLM 2 chat-bison and Llama 3-70b exhibited an excellent overall accuracy >0.98 (0.995, 0.988, 0.988, 0.988, 0.986, 0.982, 0.982, and 0.982, respectively), significantly higher than the baseline RoBERTa model (0.742). Claude 2.0, Claude 2.1, Claude 3.0 Opus, PaLM 2 chat-bison, GPT 4, Claude 3.0 Sonnet and Llama 3-70b showed a marginally higher and Gemini Advanced a marginally lower multiple-run consistency than the baseline model RoBERTa (Krippendorff's alpha value 1, 0.998, 0.996, 0.996, 0.992, 0.991, 0.989, 0.988, and 0.985, respectively)., Discussion: Claude 3.0 Opus, Claude 3.0 Sonnet, Claude 2.0, GPT 4, Claude 2.1, Gemini Advanced, PaLM 2 chat bison and Llama 3-70b performed the best, exhibiting outstanding performance in both entity extraction and binary classification, with highly consistent responses over multiple same-prompt iterations. Their use could leverage data for research and unburden healthcare professionals. Real-data analyses are warranted to confirm their performance in a real-world setting., Conclusion: Claude 3.0 Opus, Claude 3.0 Sonnet, Claude 2.0, GPT 4, Claude 2.1, Gemini Advanced, PaLM 2 chat-bison and Llama 3-70b seem to be able to reliably extract data from unstructured and semi-structured electronic health records. Further analyses using real data are warranted to confirm their performance in a real-world setting., Competing Interests: Competing interests: None declared., (© Author(s) (or their employer(s)) 2025. Re-use permitted under CC BY-NC. No commercial re-use. See rights and permissions. Published by BMJ Group.)