Identifying Challenges and Opportunities in Human-AI Collaboration in Healthcare

Park, Sun Young; Kuo, Pei-Yi; Barbarin, Andrea; Kaziunas, Elizabeth; Chow, A. W.; Singh, Karandeep; Wilcox, Lauren; Lasecki, Walter S.

doi:10.1145/3311957.3359433

Cited by 49 publications

(7 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…That is, scientific research collaboration is essentially the collaboration of information [2]. The team carries out the division of labor and cooperation around the scientific research tasks [3,4] and jointly solves the problems of scientific research projects through the flow of information and knowledge among members [5]. Correspondingly, in the face of various risks in the internal and external environment of the project, how to promote the exchange and adoption of information among members is an important topic of risk management practice and theoretical research in scientific research projects.…”

Section: Introductionmentioning

confidence: 99%

The Impact of Risk Perception Difference of Members of a Scientific Research Project Team on Information Adoption: The Moderating Effect of Knowledge Inertia

et al. 2022

View full text Add to dashboard Cite

A scientific research project is always full of uncertainties and risks. In this condition, full exchange and complementarity of information resources among project team members are necessary and important to meet the information needs for project risk management and even affect the success of the project. The differentiated risk perception of members with various professional backgrounds can contribute to the communication and complementary of the necessary information within the team. However, too much difference in perceptions of project risks may cause members’ information conflicts, which may hinder the adoption of various information and do harm to the risk management. Considering the limited research on this “contradictory” relationship, especially for the scientific research project team, a special group, this study explores the impact of the risk perception differences of the scientific research project team members on information adoption behavior through the large sample empirical statistical method. The results show the hat risk perception difference of members positively affects the learning willingness and information adoption behavior but does not show a negative effect, and there is no inverted U-shaped relationship. Further, learning willingness plays a mediating role between both, while learning inertia and experience inertia positively and negatively moderate the positive effects of risk perception differences on learning willingness, respectively. From the cognitive perspective, this study further reveals the internal influence mechanism of risk perception difference of scientific research project team members on information behavior and provides a theoretical reference for improving the efficiency of information communication and optimizing collaborative team management.

show abstract

Section: Introductionmentioning

confidence: 99%

The Impact of Risk Perception Difference of Members of a Scientific Research Project Team on Information Adoption: The Moderating Effect of Knowledge Inertia

et al. 2022

View full text Add to dashboard Cite

show abstract

“…As an increasingly prominent context for discussing bias, machine learning systems exemplify a formulation of cultural bias rooted in individual prejudice and premised on its identification and removal (Park et al 2019). This treatment locates bias within a binary logic (bias/debias) wherein harmful assumptions and decisions get baked into an entity (a person, thing, or system) (Benjamin 2019, Eubanks 2018; Noble 2018).…”

Section: Textiles Technology and Liberatory Potentialmentioning

confidence: 99%

The Bias Cut: Toward a Technopoetics of Algorithmic Systems

Rosner

2022

Catalyst

View full text Add to dashboard Cite

This essay explores a material engagement with discourses of bias. At a time when the developers of algorithmic systems are exploring concepts of bias like never before, textile bias (or the skew of woven material) offers an alternative view into the scripts of computational engagement. To probe this potential, this essay engages a range of feminist and anti-racist interventions in performance arts, critical archival studies, and my own pedagogical collaborations. With these experiments, I ask, how might material bias inform ongoing analysis of cultural bias within machine learning systems? The experiments reveal interwoven dynamics of power, labor, and historicity with particular attention to complicity and change. Through angular encounters with bias, I explore the development of an emerging technopoetics of algorithmic systems.

show abstract

“…Of high relevance to this opportunity for improving HMT in healthcare is the prior work that has focused on AI methods and training data used to improve explainability. Researchers have pointed out many issues in using biased AI systems, e.g., diagnostic systems using datasets that are imbalanced with respect to race or other demographics [ 9 ]. Biased AI systems can diminish rather than augment human intelligence in collaborative decision teaming.…”

Section: Background and Related Workmentioning

confidence: 99%

“…Broad adoption in an industry well-grounded in human interaction, it should not be surprising that human machine teaming should be a dominant and priority topic for researchers in this field. Many of the existing AI-based autonomous health applications have already demonstrated an imperative to involve the human stakeholders who use these systems and are the most affected by them [ 9 ]. In contrast to the situation where autonomous systems were mainly automating routine human tasks in the past, machine collaboration implies that AI systems work jointly with humans like teammates to solve problems.…”

Section: Introductionmentioning

confidence: 99%

Providing Care: Intrinsic Human–Machine Teams and Data

Russell

Kumar

2022

Entropy

View full text Add to dashboard Cite

Despite the many successes of artificial intelligence in healthcare applications where human–machine teaming is an intrinsic characteristic of the environment, there is little work that proposes methods for adapting quantitative health data-features with human expertise insights. A method for incorporating qualitative expert perspectives in machine learning training data is proposed. The method implements an entropy-based consensus construct that minimizes the challenges of qualitative-scale data such that they can be combined with quantitative measures in a critical clinical event (CCE) vector. Specifically, the CCE vector minimizes the effects where (a) the sample size is too small, (b) the data may not be normally distributed, or (c) The data are from Likert scales, which are ordinal, so parametric statistics cannot be used. The incorporation of human perspectives in machine learning training data provides encoding of human considerations in the subsequent machine learning model. This encoding provides a basis for increasing explainability, understandability, and ultimately trust in AI-based clinical decision support system (CDSS), thereby improving human–machine teaming concerns. A discussion of applying the CCE vector in a CDSS regime and implications for machine learning are also presented.

show abstract

Identifying Challenges and Opportunities in Human-AI Collaboration in Healthcare

Cited by 49 publications

References 9 publications

The Impact of Risk Perception Difference of Members of a Scientific Research Project Team on Information Adoption: The Moderating Effect of Knowledge Inertia

The Impact of Risk Perception Difference of Members of a Scientific Research Project Team on Information Adoption: The Moderating Effect of Knowledge Inertia

The Bias Cut: Toward a Technopoetics of Algorithmic Systems

Providing Care: Intrinsic Human–Machine Teams and Data

Contact Info

Product

Resources

About