PoCaP Corpus: A Multimodal Dataset for Smart Operating Room Speech Assistant Using Interventional Radiology Workflow Analysis

Demir, Kubilay Can; May, Matthias; Schmid, Axel; Uder, Michael; Breininger, Katharina; Weise, Tobias; Maier, Andreas; Yang, Seung Hee

doi:10.1007/978-3-031-16270-1_38

Cited by 3 publications

(1 citation statement)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The scarcity of medical data from VIR operating suites poses a significant obstacle to effective data collection. 38 Also, LLMs encounter challenges such as hallucinations and weak numerical reasoning. 36 39 When applied in patient care without due caution, these issues can lead to severe harm or even fatal consequences, underscoring the urgency of developing improved mitigation techniques.…”

Section: Challenges Ethics and Recommendations In Llm Implementation ...mentioning

confidence: 99%

Exploring the Potentials of Large Language Models in Vascular and Interventional Radiology: Opportunities and Challenges

Togunwa,

Ajibade,

Uche-Orji

et al. 2024

The Arab Journal of Interventional Radiology

View full text Add to dashboard Cite

The increasing integration of artificial intelligence (AI) in healthcare, particularly in vascular and interventional radiology (VIR), has opened avenues for enhanced efficiency and precision. This narrative review delves into the potential applications of large language models (LLMs) in VIR, with a focus on Chat Generative Pre-Trained Transformer (ChatGPT) and similar models. LLMs, designed for natural language processing, exhibit promising capabilities in clinical decision-making, workflow optimization, education, and patient-centered care. The discussion highlights LLMs' ability to analyze extensive medical literature, aiding radiologists in making informed decisions. Moreover, their role in improving clinical workflow, automating report generation, and intelligent patient scheduling is explored. This article also examines LLMs' impact on VIR education, presenting them as valuable tools for trainees. Additionally, the integration of LLMs into patient education processes is examined, highlighting their potential to enhance patient-centered care through simplified and accurate medical information dissemination. Despite these potentials, this paper discusses challenges and ethical considerations, including AI over-reliance, potential misinformation, and biases. The scarcity of comprehensive VIR datasets and the need for ongoing monitoring and interdisciplinary collaboration are also emphasized. Advocating for a balanced approach, the combination of LLMs with computer vision AI models addresses the inherently visual nature of VIR. Overall, while the widespread implementation of LLMs in VIR may be premature, their potential to improve various aspects of the discipline is undeniable. Recognizing challenges and ethical considerations, fostering collaboration, and adhering to ethical standards are essential for unlocking the full potential of LLMs in VIR, ushering in a new era of healthcare delivery and innovation.

show abstract

Section: Challenges Ethics and Recommendations In Llm Implementation ...mentioning

confidence: 99%

Exploring the Potentials of Large Language Models in Vascular and Interventional Radiology: Opportunities and Challenges

Togunwa,

Ajibade,

Uche-Orji

et al. 2024

The Arab Journal of Interventional Radiology

View full text Add to dashboard Cite

show abstract

Automatic Recognition of Speaker Labels Using CNN-SVM Scheme

Karthikeyan¹,

Kumar²,

Karthikeyan

2023

Lecture Notes in Networks and Systems

View full text Add to dashboard Cite

The effect of speech pathology on automatic speaker verification: a large-scale study

Tayebi Arasteh,

Weise,

Schuster

et al. 2023

Sci Rep

View full text Add to dashboard Cite

Navigating the challenges of data-driven speech processing, one of the primary hurdles is accessing reliable pathological speech data. While public datasets appear to offer solutions, they come with inherent risks of potential unintended exposure of patient health information via re-identification attacks. Using a comprehensive real-world pathological speech corpus, with over n$$=$$ = 3800 test subjects spanning various age groups and speech disorders, we employed a deep-learning-driven automatic speaker verification (ASV) approach. This resulted in a notable mean equal error rate (EER) of $$0.89 \pm 0.06 \%$$ 0.89 ± 0.06 % , outstripping traditional benchmarks. Our comprehensive assessments demonstrate that pathological speech overall faces heightened privacy breach risks compared to healthy speech. Specifically, adults with dysphonia are at heightened re-identification risks, whereas conditions like dysarthria yield results comparable to those of healthy speakers. Crucially, speech intelligibility does not influence the ASV system’s performance metrics. In pediatric cases, particularly those with cleft lip and palate, the recording environment plays a decisive role in re-identification. Merging data across pathological types led to a marked EER decrease, suggesting the potential benefits of pathological diversity in ASV, accompanied by a logarithmic boost in ASV effectiveness. In essence, this research sheds light on the dynamics between pathological speech and speaker verification, emphasizing its crucial role in safeguarding patient confidentiality in our increasingly digitized healthcare era.

show abstract

PoCaP Corpus: A Multimodal Dataset for Smart Operating Room Speech Assistant Using Interventional Radiology Workflow Analysis

Cited by 3 publications

References 12 publications

Exploring the Potentials of Large Language Models in Vascular and Interventional Radiology: Opportunities and Challenges

Exploring the Potentials of Large Language Models in Vascular and Interventional Radiology: Opportunities and Challenges

Automatic Recognition of Speaker Labels Using CNN-SVM Scheme

The effect of speech pathology on automatic speaker verification: a large-scale study

Contact Info

Product

Resources

About