Cardiovascular Disease Prevention Recommendations From an Online Chat-Based AI Model

Gellert, George A.; Jaszczak, Jakub

doi:10.1001/jama.2023.8184

Cited by 2 publications

(2 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Regardless of the prediction class (AD/Unsure/CN), there does not appear to be a clear relationship between MMSE scores and the LLM chatbots' prediction performance for CN subjects (high MMSE scores are inherently limited in distribution). However, a significant subset of CN subjects with high MMSE scores (27)(28)(29)(30) tend to be misclassified as AD or Unsure by all three LLM chatbots, suggesting that high MMSE scores do not necessarily aid CN prediction; other factors, such as linguistic attributes or contextual cues, may offer a confounding impact on the LLM chatbots' prediction.…”

Section: Insights From Mmse Score Comparisonmentioning

confidence: 96%

“…Large language model (LLM) chatbots such as OpenAI's Generative Pretrained Transformer (GPT, versions 3.5 and 4) and Google's Bard, demonstrate impressive capabilities in many domains, including healthcare settings [25][26][27], to support early detection and clinical assessment. Here, we explore the utility of LLM chatbots (ChatGPT-3.5, ChatGPT-4, and Bard) for identifying AD in individuals, using textual transcriptions derived from spontaneous speech: a non-trivial assessment task that currently poses significant challenges for other state-of-the-art detection modalities and could benefit immediately from advanced artificial intelligence techniques directly applied "in-the-field".…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Performance Assessment of ChatGPT versus Bard in Detecting Alzheimer’s Dementia

B.T,

Chen

2024

Diagnostics

View full text Add to dashboard Cite

Large language models (LLMs) find increasing applications in many fields. Here, three LLM chatbots (ChatGPT-3.5, ChatGPT-4, and Bard) are assessed in their current form, as publicly available, for their ability to recognize Alzheimer’s dementia (AD) and Cognitively Normal (CN) individuals using textual input derived from spontaneous speech recordings. A zero-shot learning approach is used at two levels of independent queries, with the second query (chain-of-thought prompting) eliciting more detailed information than the first. Each LLM chatbot’s performance is evaluated on the prediction generated in terms of accuracy, sensitivity, specificity, precision, and F1 score. LLM chatbots generated a three-class outcome (“AD”, “CN”, or “Unsure”). When positively identifying AD, Bard produced the highest true-positives (89% recall) and highest F1 score (71%), but tended to misidentify CN as AD, with high confidence (low “Unsure” rates); for positively identifying CN, GPT-4 resulted in the highest true-negatives at 56% and highest F1 score (62%), adopting a diplomatic stance (moderate “Unsure” rates). Overall, the three LLM chatbots can identify AD vs. CN, surpassing chance-levels, but do not currently satisfy the requirements for clinical application.

show abstract

Section: Insights From Mmse Score Comparisonmentioning

confidence: 96%

Section: Introductionmentioning

confidence: 99%

Performance Assessment of ChatGPT versus Bard in Detecting Alzheimer’s Dementia

B.T,

Chen

2024

Diagnostics

View full text Add to dashboard Cite

show abstract

The Future of Digital Healthcare in Cardiac Prevention Programme in Indonesia: From Bedside to Boardroom

Dwiputra,

Ambari,

Desandri

et al. 2024

J Asian Pac Soc Cardiol

View full text Add to dashboard Cite

Digital technology has become integral to the practice of medicine, especially in response to the COVID-19 pandemic and its associated lockdowns and social distancing measures. It has allowed patients, including those undergoing cardiac rehabilitation, to access healthcare services regardless of their geographical location. However, the widespread integration of digital healthcare on a national scale presents numerous challenges. These obstacles encompass various aspects, including patient–physician relationships, legal policies, health information system quality, data privacy, cost–benefit analysis and the potential loss of essential human interaction. To address these hurdles, Indonesia has developed a framework and strategy for implementing digital healthcare. This review examines the experiences of digital healthcare leaders in different countries, considering their unique healthcare systems, cultural contexts and approaches to tackling these issues. Additionally, an analysis of Indonesia’s current healthcare system implementation is provided. By showcasing successful examples and considering critical factors, we hope to facilitate a successful implementation of digital healthcare in cardiac prevention while gathering valuable insights from these endeavours.

show abstract

Cardiovascular Disease Prevention Recommendations From an Online Chat-Based AI Model

Cited by 2 publications

References 8 publications

Performance Assessment of ChatGPT versus Bard in Detecting Alzheimer’s Dementia

Performance Assessment of ChatGPT versus Bard in Detecting Alzheimer’s Dementia

The Future of Digital Healthcare in Cardiac Prevention Programme in Indonesia: From Bedside to Boardroom

Contact Info

Product

Resources

About