The Representation of Trust in Artificial Intelligence Healthcare Research

Kutza, Jan-Oliver; Hannemann, Niels; Hübner, Ursula; Babitsch, Birgit

doi:10.3233/shti230409

Cited by 4 publications

(3 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Patients' and doctors' confidence in using AI is also an important issue. Unfortunately this topic is very often neglected and there are few works that analyze this [23]. Studies show that only 37% of questioned people would be willing to trust AI in the healthcare system.…”

Section: The Use Of Ai -Ethics and The Lawmentioning

confidence: 99%

AI in Dermatology: Bridging the Gap Between Potential and Practice

Klarycki,

Tomczyk,

Podgórska

et al. 2024

J Educ Health Sport

View full text Add to dashboard Cite

Artificial intelligence has been rapidly penetrating every element of our lives for quite some time. However, its presence in health care has remained elusive. This is particularly apparent in the field of dermatology, where, given the characteristics of this discipline of medicine, it would seems that its presence should be abundant. Malignant skin lesions are still high in the statistics in terms of cancer mortality while being one of the easiest to treat when diagnosed early. There are many reasons why artificial intelligence is not used in daily practice as an aid for cancers detection. However the most important one is the ongoing insufficient quality of the algorithms, which, despite great results in laboratory settings, do not produce good enough outcomes in clinical settings. Other important reasons are that people still distrust and fear artificial intelligence and simply the legal lack of adaptation of countries to its lawful and safe use. Despite the work of scientists and legislators the road to seeing artificial intelligence as a helping tool for dermatologists on a daily basis is still very long and requires the attention of scientists and the whole medical community.

show abstract

Section: The Use Of Ai -Ethics and The Lawmentioning

confidence: 99%

AI in Dermatology: Bridging the Gap Between Potential and Practice

Klarycki,

Tomczyk,

Podgórska

et al. 2024

J Educ Health Sport

View full text Add to dashboard Cite

show abstract

“…Neser [3] distinguishes three forms of trust: system trust, interpersonal trust, and self-trust, which are also reported in the context of technologies. Although several factors affecting trust in digitalization and AI have been identified, the process of trust building has been insufficiently studied [4]. Therefore, the aim of this qualitative study 2 is to investigate 1) if and how (medical) experts have trust in digitalization or AI in healthcare, and 2) which factors contribute to trust building.…”

Section: Introductionmentioning

confidence: 99%

Trust in Digitalization and AI: Findings from a Qualitative Study on Healthcare Professionals in Germany

Babitsch,

Hannemann,

Kutza

et al. 2023

Telehealth Ecosystems in Practice

View full text Add to dashboard Cite

The acceptance and use of digital technologies depend on the trustworthiness attributed to them. Experts were interviewed about how they assign trust to digital technologies or AI (N=12). The data were analyzed applying the focused qualitative content analysis. All of the experts have experience with digital technologies, but only seven with AI. The majority of experts generally trust digital technologies, but only five experts expressed a general trust in AI. Similar reasons contributing to trust building were given for digital technologies and AI. The results show the complexity of the trust building process and the construct of trust itself. The development of explainable AI and professional training are prerequisites to support a critical and safe use of these technologies.

show abstract

“…The emergence of large language models (LLMs) such as GPT-4 (OpenAI) has led to their integration into search engines such as Bing (Microsoft) in order to enhance the search experience 2 . However, a major concern surrounding these LLMs is their inherent "black box" nature, as it is often unclear where these models obtain their information from 3 . Indeed, LLMs are known to "hallucinate," or generate responses that are unfaithful to the underlying training material 4 .…”

mentioning

confidence: 99%

Assessing the Accuracy and Reliability of AI-Generated Responses to Patient Questions Regarding Spine Surgery

Kasthuri,

Glueck,

Pham

et al. 2024

Journal of Bone and Joint Surgery

View full text Add to dashboard Cite

Background: In today’s digital age, patients increasingly rely on online search engines for medical information. The integration of large language models such as GPT-4 into search engines such as Bing raises concerns over the potential transmission of misinformation when patients search for information online regarding spine surgery. Methods: SearchResponse.io, a database that archives People Also Ask (PAA) data from Google, was utilized to determine the most popular patient questions regarding 4 specific spine surgery topics: anterior cervical discectomy and fusion, lumbar fusion, laminectomy, and spinal deformity. Bing’s responses to these questions, along with the cited sources, were recorded for analysis. Two fellowship-trained spine surgeons assessed the accuracy of the answers on a 6-point scale and the completeness of the answers on a 3-point scale. Inaccurate answers were re-queried 2 weeks later. Cited sources were categorized and evaluated against Journal of the American Medical Association (JAMA) benchmark criteria. Interrater reliability was measured with use of the kappa statistic. A linear regression analysis was utilized to explore the relationship between answer accuracy and the type of source, number of sources, and mean JAMA benchmark score. Results: Bing’s responses to 71 PAA questions were analyzed. The average completeness score was 2.03 (standard deviation [SD], 0.36), and the average accuracy score was 4.49 (SD, 1.10). Among the question topics, spinal deformity had the lowest mean completeness score. Re-querying the questions that initially had answers with low accuracy scores resulted in responses with improved accuracy. Among the cited sources, commercial sources were the most prevalent. The JAMA benchmark score across all sources averaged 2.63. Government sources had the highest mean benchmark score (3.30), whereas social media had the lowest (1.75). Conclusions: Bing’s answers were generally accurate and adequately complete, with incorrect responses rectified upon re-querying. The plurality of information was sourced from commercial websites. The type of source, number of sources, and mean JAMA benchmark score were not significantly correlated with answer accuracy. These findings underscore the importance of ongoing evaluation and improvement of large language models to ensure reliable and informative results for patients seeking information regarding spine surgery online amid the integration of these models in the search experience.

show abstract

The Representation of Trust in Artificial Intelligence Healthcare Research

Cited by 4 publications

References 10 publications

AI in Dermatology: Bridging the Gap Between Potential and Practice

AI in Dermatology: Bridging the Gap Between Potential and Practice

Trust in Digitalization and AI: Findings from a Qualitative Study on Healthcare Professionals in Germany

Assessing the Accuracy and Reliability of AI-Generated Responses to Patient Questions Regarding Spine Surgery

Contact Info

Product

Resources

About