TachAId-An interactive tool supporting the design of human-centered AI solutions.
Ontology highlight
ABSTRACT:
SUBMITTER: Bauroth M
PROVIDER: S-EPMC10963619 | biostudies-literature | 2024
REPOSITORIES: biostudies-literature
Publications
tachAId-An interactive tool supporting the design of human-centered AI solutions.
Bauroth Max M Rath-Manakidis Pavlos P Langholf Valentin V Wiskott Laurenz L Glasmachers Tobias T
Frontiers in artificial intelligence 20240312
In an era where Artificial Intelligence (AI) integration into business processes is crucial for maintaining competitiveness, there is a growing need for structured guidance on designing AI solutions that align with human needs. To this end, we present "technical assistance concerning human-centered AI development" (tachAId), an interactive advisory tool which comprehensively guides AI developers and decision makers in navigating the machine learning lifecycle with a focus on human-centered desig ...[more]
Similar Datasets
Project description:BackgroundmHealth (mobile health) systems have been deployed widely in low- and middle-income countries (LMICs) for health system strengthening, requiring considerable resource allocation. However, most solutions have not achieved scale or sustainability. Poor usability and failure to address perceived needs are among the principal reasons mHealth systems fail to achieve acceptance and adoption by health care workers. A human-centered design approach to improving mHealth system use requires an exploration of users' perceptions of mHealth systems, including the environmental, user-related, and technological aspects of a system. At present, there is a dearth of contextually intelligent tools available to mHealth developers that can guide such exploration before full-scale development and deployment.ObjectiveTo develop a tool to aid optimization of mHealth solutions in LMICs to facilitate human-centered design and, consequently, successful adoption.MethodsWe collated findings and themes from key qualitative studies on mHealth deployment in LMICs. We then used the Informatics Stack framework by Lehmann to label, sort, and collate findings and themes into a list of questions that explore the environment, users, artifacts, information governance, and interoperability of mHealth systems deployed in LMICs.ResultsWe developed the Vinyasa Tool to aid qualitative research about the need and usability of mHealth solutions in LMICs. The tool is a guide for focus group discussions and key informant interviews with community-based health care workers and primary care medical personnel who use or are expected to use proposed mHealth solutions. The tool consists of 71 questions organized in 11 sections that unpack and explore multiple aspects of mHealth systems from the perspectives of their users. These include the wider world and organization in which an mHealth solution is deployed; the roles, functions, workflow, and adoption behavior of a system's users; the security, privacy, and interoperability afforded by a system; and the artifacts of an information system-the data, information, knowledge, algorithms, and technology that constitute the system. The tool can be deployed in whole or in part, depending on the context of the study.ConclusionsThe Vinyasa Tool is the first such comprehensive qualitative research instrument incorporating questions contextualized to the LMIC setting. We expect it to find wide application among mHealth developers, health system administrators, and researchers developing and deploying mHealth tools for use by patients, providers, and administrators. The tool is expected to guide users toward human-centered design with the goal of improving relevance, usability, and, therefore, adoption.
| S-EPMC10580130 | biostudies-literature
Project description:Transparency in Machine Learning (ML), often also referred to as interpretability or explainability, attempts to reveal the working mechanisms of complex models. From a human-centered design perspective, transparency is not a property of the ML model but an affordance, i.e., a relationship between algorithm and users. Thus, prototyping and user evaluations are critical to attaining solutions that afford transparency. Following human-centered design principles in highly specialized and high stakes domains, such as medical image analysis, is challenging due to the limited access to end users and the knowledge imbalance between those users and ML designers. To investigate the state of transparent ML in medical image analysis, we conducted a systematic review of the literature from 2012 to 2021 in PubMed, EMBASE, and Compendex databases. We identified 2508 records and 68 articles met the inclusion criteria. Current techniques in transparent ML are dominated by computational feasibility and barely consider end users, e.g. clinical stakeholders. Despite the different roles and knowledge of ML developers and end users, no study reported formative user research to inform the design and development of transparent ML models. Only a few studies validated transparency claims through empirical user evaluations. These shortcomings put contemporary research on transparent ML at risk of being incomprehensible to users, and thus, clinically irrelevant. To alleviate these shortcomings in forthcoming research, we introduce the INTRPRT guideline, a design directive for transparent ML systems in medical image analysis. The INTRPRT guideline suggests human-centered design principles, recommending formative user research as the first step to understand user needs and domain requirements. Following these guidelines increases the likelihood that the algorithms afford transparency and enable stakeholders to capitalize on the benefits of transparent ML.
| S-EPMC9581990 | biostudies-literature
Project description:As part of the Special Issue topic "Human-Centered AI at Work: Common Ground in Theories and Methods," we present a perspective article that looks at human-AI teamwork from a team-centered AI perspective, i. e., we highlight important design aspects that the technology needs to fulfill in order to be accepted by humans and to be fully utilized in the role of a team member in teamwork. Drawing from the model of an idealized teamwork process, we discuss the teamwork requirements for successful human-AI teaming in interdependent and complex work domains, including e.g., responsiveness, situation awareness, and flexible decision-making. We emphasize the need for team-centered AI that aligns goals, communication, and decision making with humans, and outline the requirements for such team-centered AI from a technical perspective, such as cognitive competence, reinforcement learning, and semantic communication. In doing so, we highlight the challenges and open questions associated with its implementation that need to be solved in order to enable effective human-AI teaming.
| S-EPMC10565103 | biostudies-literature
Project description:The success of smart homes is fundamentally dependent on their adoption and use by people in the context of daily life. This study investigated ways to control and adapt the technology to fulfill user daily needs, which are the active drivers of smart housing technology adoption. A framework of smart home services was developed by focusing on the practicability of each variable from the perspective of supporting user experience. By developing scenarios based on previous studies, we identified residents' behaviors and intentions regarding smart home technology and its use. Their issues were identified through the health problems and daily activities identified in the scenarios, and customized smart home services were developed for each resident based on technical solutions, space requirements, and design solutions. The main strength of this study is the adoption of user-centered methods to build a rich picture of the issues that affect households and the community related to the design, installation, and use of technology.
| S-EPMC7100148 | biostudies-literature
Project description:Given the steady increase of emergency department (ED) visits related to opioid overdoses, this study aims to determine the design and usability of an ED-centered mHealth patient-to-peer referral prototype tool that allows patients to refer peers to comprehensive HIV/HCV and opioid misuse prevention services. Two iterative focus group discussion (FDG) sessions and one use-case session were conducted. Eligible participants who were ≥18 years, had a history of injection drug use (IDU), and had utilized the ED in the past year were recruited through the distribution of flyers at the study institution, including the study ED. Human-centered design process was completed by using participant feedback on perceived utility, usability/accessibility, tool design, and clarity/readability to fine-tune prototype version and drive subsequent discussion sessions. Sixteen consented individuals participated in at least one of the sessions. Feedback revealed that participants favored the inclusion of the webpage link on the referral card as means to bypass QR code if needed, more descriptions highlighting the exact services offered, and the fact that no personal information was required to complete the referral process. The prototype underwent several adjustments between user-centered FDG sessions, which ultimately ended in including features such as an online webpage with educational videos, SMS text-message communication system, and QR code usage into the final patient-to-peer referral tool prototype. The findings of this study suggest a human-centered designed patient-to-peer referral tool could be a feasible approach to linking community members at risk of IDU to HIV/HCV and opioid use-related preventive services from ED patients.
| S-EPMC9761213 | biostudies-literature
Project description:BackgroundThe SARS CoV-2 (COVID-19) pandemic catalyzed a dramatic shift in healthcare delivery, with telemedicine emerging as a common mode of care provision. While pre-pandemic telemedicine services were more commonly used for preventive visits and had better adherence among younger and more affluent demographics, the landscape of telehealth in the post-pandemic period has shifted significantly to include surgical visits and publicly-insured patient populations. Without specific insights from patients and clinicians to guide this transition, telehealth delivery risks exacerbating disparities in access, experience and outcomes for medically underserved populations.MethodsWe utilized a human-centered design (HCD) approach to gain insights into patient and clinician perspectives on telehealth delivery at a surgical outpatient clinic in an urban safety-net hospital and level 1 trauma center. During the Inspiration phase of HCD, we conducted 19 in-depth interviews with patients and surgical clinicians, and applied a combined thematic analysis and design synthesis approach to identify key insight statements representing actionable tensions across cohorts. During the Ideation phase of HCD, we held a structured brainstorming session to identify solutions and facilitated a discussion with surgical faculty to co-design and refine a prototype.ResultsInterview analysis revealed 12 main themes, which were then reorganized into 5 core insights across both groups: "In-person appointments can be resource intensive for patients, making their attendance costly in more ways than one"; "When sacrificing connection for convenience, telehealth exacerbates discrimination felt by historically marginalized patients"; "Personal interactions are crucial for establishing new relationships and repairing mistrust between patients and clinicians"; "Visual cues and non-verbal communication are essential for personalized and effective surgical care"; "Patients and clinicians value the human infrastructure built into the in-person visit experience." Brainstorming participants generated ideas from the first insight statement. Subsequent prototyping and co-design sessions led to the development of a screening prototype allowing both clinic staff and patients to book telehealth appropriate appointments.ConclusionsThis study offers a HCD approach to developing insights and tailoring health service interventions to the local contexts for safety-net providers. By understanding the unique needs and preferences of underserved populations, we can develop telehealth interventions that increase adoption and ensure equitable access to care.
| S-EPMC11817022 | biostudies-literature
Project description:Explainable Artificial Intelligence (XAI) aims to provide insights into the inner workings and the outputs of AI systems. Recently, there's been growing recognition that explainability is inherently human-centric, tied to how people perceive explanations. Despite this, there is no consensus in the research community on whether user evaluation is crucial in XAI, and if so, what exactly needs to be evaluated and how. This systematic literature review addresses this gap by providing a detailed overview of the current state of affairs in human-centered XAI evaluation. We reviewed 73 papers across various domains where XAI was evaluated with users. These studies assessed what makes an explanation "good" from a user's perspective, i.e., what makes an explanation meaningful to a user of an AI system. We identified 30 components of meaningful explanations that were evaluated in the reviewed papers and categorized them into a taxonomy of human-centered XAI evaluation, based on: (a) the contextualized quality of the explanation, (b) the contribution of the explanation to human-AI interaction, and (c) the contribution of the explanation to human-AI performance. Our analysis also revealed a lack of standardization in the methodologies applied in XAI user studies, with only 19 of the 73 papers applying an evaluation framework used by at least one other study in the sample. These inconsistencies hinder cross-study comparisons and broader insights. Our findings contribute to understanding what makes explanations meaningful to users and how to measure this, guiding the XAI community toward a more unified approach in human-centered explainability.
| S-EPMC11525002 | biostudies-literature
Project description:BackgroundPremature infants are at increased risk of kidney-related complications, including acute kidney injury (AKI) and chronic kidney disease (CKD). The risk of CKD in prematurely born infants is underrecognized by health care teams and caregivers. Understanding how to communicate the risk of CKD to caregivers is essential for longitudinal clinical follow-up and adherence.ObjectiveThis study aimed to determine family caregiver attitudes toward kidney health and risk communication during a neonatal intensive care admission. We also sought to understand caregiver preferences for the communication of information surrounding the risk of CKD in premature infants.MethodsWe augmented standard qualitative group sessions with human-centered design methods to assess parent preferences and clinician perspectives. Caregivers recruited had a prematurely born child who spent time in the neonatal intensive care unit at Riley Hospital for Children in Indianapolis, Indiana, and experienced AKI or another kidney complication, which put them at risk for future CKD. We used a variety of specific design methods in these sessions, including card sorting, projective methods, experience mapping, and constructive methods.ResultsA total of 7 clinicians and 8 caregivers participated in 3 group sessions. Caregivers and clinicians readily acknowledged barriers to and drivers of long-term kidney monitoring as well as opportunities for communication of the risk of long-term kidney disease. Caregivers' primary concerns were for both the type and depth of information conveyed as well as the time at which it was communicated. Participants emphasized the importance of collaboration between the hospital care team and the primary care provider. Participant input was synthesized into several prototype concepts and, ultimately, into a rough prototype of a website and an informational flyer.ConclusionsCaregivers of premature infants are open to communication about kidney health during their neonatal admission. The next phase of this work will translate caregivers' preferences into family-centered communication tools and test their efficacy in the neonatal intensive care unit.
| S-EPMC10366965 | biostudies-literature
Project description:Sensory and emotional experiences are essential for mental and physical well-being, especially within the realm of psychiatry. This article highlights recent advances in cognitive neuroscience, emphasizing the significance of pain recognition and empathic artificial intelligence (AI) in healthcare. We provide an overview of the recent development process in computational pain recognition and cognitive neuroscience regarding the mechanisms of pain and empathy. Through a comprehensive discussion, the article delves into critical questions such as the methodologies for AI in recognizing pain from diverse sources of information, the necessity for AI to exhibit empathic responses, and the associated advantages and obstacles linked with the development of empathic AI. Moreover, insights into the prospects and challenges are emphasized in relation to fostering artificial empathy. By delineating potential pathways for future research, the article aims to contribute to developing effective assistants equipped with empathic capabilities, thereby introducing safe and meaningful interactions between humans and AI, particularly in the context of mental health and psychiatry.
| S-EPMC11357883 | biostudies-literature
Project description:Building artificial intelligence (AI) that aligns with human values is an unsolved problem. Here we developed a human-in-the-loop research pipeline called Democratic AI, in which reinforcement learning is used to design a social mechanism that humans prefer by majority. A large group of humans played an online investment game that involved deciding whether to keep a monetary endowment or to share it with others for collective benefit. Shared revenue was returned to players under two different redistribution mechanisms, one designed by the AI and the other by humans. The AI discovered a mechanism that redressed initial wealth imbalance, sanctioned free riders and successfully won the majority vote. By optimizing for human preferences, Democratic AI offers a proof of concept for value-aligned policy innovation.
| S-EPMC9584820 | biostudies-literature