Clinical trials are indispensable in developing new treatments, but they face obstacles in patient recruitment and retention, hindering the enrollment of necessary participants. To tackle these challenges, deep learning frameworks have been created to match patients to trials. These frameworks calculate the similarity between patients and clinical trial eligibility criteria, considering the discrepancy between inclusion and exclusion criteria. Recent studies have shown that these frameworks outperform earlier approaches. However, deep learning models may raise fairness issues in patient-trial matching when certain sensitive groups of individuals are underrepresented in clinical trials, leading to incomplete or inaccurate data and potential harm. To tackle the issue of fairness, this work proposes a fair patient-trial matching framework by generating a patient-criterion level fairness constraint. The proposed framework considers the inconsistency between the embedding of inclusion and exclusion criteria among patients of different sensitive groups. The experimental results on real-world patient-trial and patient-criterion matching tasks demonstrate that the proposed framework can successfully alleviate the predictions that tend to be biased.
more »
« less
[POSTER] Matching vs. Non-Matching Visuals and Shape for Embodied Virtual Healthcare Agents
Embodied virtual agents serving as patient simulators are widely
used in medical training scenarios, ranging from physical patients to
virtual patients presented via virtual and augmented reality technologies.
Physical-virtual patients are a hybrid solution that combines
the benefits of dynamic visuals integrated into a human-shaped physical
form that can also present other cues, such as pulse, breathing
sounds, and temperature. Sometimes in simulation the visuals and
shape do not match. We carried out a human-participant study employing
graduate nursing students in pediatric patient simulations
comprising conditions associated with matching/non-matching of
the visuals and shape.
more »
« less
- PAR ID:
- 10105863
- Date Published:
- Journal Name:
- IEEE Virtual Reality
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
more » « less
Background. Simulation has revolutionized teaching and learning. However, traditional manikins are limited in their ability to exhibit emotions, movements, and interactive eye gaze. As a result, students struggle with immersion and may be unable to authentically relate to the patient.
Intervention. We developed a new type of patient simulator called the Physical-Virtual Patients (PVP) which combines the physicality of manikins with the richness of dynamic visuals. The PVP uses spatial Augmented Reality to rear project dynamic imagery (e.g., facial expressions, ptosis, pupil reactions) on a semi-transparent physical shell. The shell occupies space and matches the dimensions of a human head.
Methods. We compared two groups of third semester nursing students (N=59) from a baccalaureate program using a between-participant design, one group interacting with a traditional high-fidelity manikin versus a more realistic PVP head. The learners had to perform a neurological assessment. We measured authenticity, urgency, and learning.
Results. Learners had a more realistic encounter with the PVP patient (p=0.046), they were more engaged with the PVP condition compared to the manikin in terms of authenticity of encounter and cognitive strategies. The PVP provoked a higher sense of urgency (p=0.002). There was increased learning for the PVP group compared to the manikin group on the pre and post-simulation scores (p=0.027).
Conclusion. The realism of the visuals in the PVP increases authenticity and engagement which results in a greater sense of urgency and overall learning.
-
more » « less
Abstract Cancer cell lines serve as model
in vitro systems for investigating therapeutic interventions. Recent advances in high‐throughput genomic profiling have enabled the systematic comparison between cell lines and patient tumor samples. The highly interconnected nature of biological data, however, presents a challenge when mapping patient tumors to cell lines. Standard clustering methods can be particularly susceptible to the high level of noise present in these datasets and only output clusters at one unknown scale of the data. In light of these challenges, we present NetCellMatch, a robust framework for network‐based matching of cell lines to patient tumors. NetCellMatch first constructs a global network across all cell line‐patient samples using their genomic similarity. Then, a multi‐scale community detection algorithm integrates information across topologically meaningful (clustering) scales to obtain Network‐Based Matching Scores (NBMS). NBMS are measures ofcluster robustness which map patient tumors to cell lines. We use NBMS to determine representative “avatar” cell lines for subgroups of patients. We apply NetCellMatch to reverse‐phase protein array data obtained from The Cancer Genome Atlas for patients and the MD Anderson Cell Line Project for cell lines. Along with avatar cell line identification, we evaluate connectivity patterns for breast, lung, and colon cancer and explore the proteomic profiles of avatars and their corresponding top matching patients. Our results demonstrate our framework's ability to identify both patient‐cell line matches and potential proteomic drivers of similarity. Our methods are general and can be easily adapted to other'omic datasets. -
This study investigated 37 third graders’ explanations of subtraction worked examples shown in number sentence or visual form (ten frame or number line) and their justifications for which visual and numerical worked examples corresponded to the same subtraction strategy. Results showed that third graders gave more detailed explanations in number sentence form than in visual form; whereas, they had higher accuracy in matching number sentences to visuals than vice versa. When matching, they were more likely to reason sufficiently when identifying processes represented in the worked examples as opposed to reasoning about the order of the numbers. When using worked examples, teachers should make use of visuals to help students focus on how the visuals represent the operations.more » « less
-
Conventional Intelligent Virtual Agents (IVAs) focus primarily on the visual and auditory channels for both the agent and the interacting human: the agent displays a visual appearance and speech as output, while processing the human’s verbal and non-verbal behavior as input. However, some interactions, particularly those between a patient and healthcare provider, inherently include tactile components.We introduce an Intelligent Physical-Virtual Agent (IPVA) head that occupies an appropriate physical volume; can be touched; and via human-in-the-loop control can change appearance, listen, speak, and react physiologically in response to human behavior. Compared to a traditional IVA, it provides a physical affordance, allowing for more realistic and compelling human-agent interactions. In a user study focusing on neurological assessment of a simulated patient showing stroke symptoms, we compared the IPVA head with a high-fidelity touch-aware mannequin that has a static appearance. Various measures of the human subjects indicated greater attention, affinity for, and presence with the IPVA patient, all factors that can improve healthcare training.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- The DOI is not currently available.
