arXiv subjects

Erik P. Duhaime

Publications and source records attributed to Erik P. Duhaime.

3 recordsLinked to original sources

arXiv2025

Explainable AI as a Double-Edged Sword in Dermatology: The Impact on Clinicians versus The Public

Artificial intelligence (AI) is increasingly permeating healthcare, from physician assistants to consumer applications. Since AI algorithm's opacity challenges human interaction, explainable AI (XAI) addresses this by providing AI decision-making insight, but evidence suggests XAI can paradoxically induce over-reliance or bias. We present results from two large-scale experiments (623 lay people; 153 primary care physicians, PCPs) combining a fairness-based diagnosis AI model and different XAI explanations to examine how XAI assistance, particularly multimodal large language models (LLMs), influences diagnostic performance. AI assistance balanced across skin tones improved accuracy and reduced diagnostic disparities. However, LLM explanations yielded divergent effects: lay users showed higher automation bias - accuracy boosted when AI was correct, reduced when AI erred - while experienced PCPs remained resilient, benefiting irrespective of AI accuracy. Presenting AI suggestions first also led to worse outcomes when the AI was incorrect for both groups. These findings highlight XAI's varying impact based on expertise and timing, underscoring LLMs as a "double-edged sword" in medical AI and informing future human-AI collaborative system design.

cs.HC↗

arXiv2023

Deep Learning for Detection and Localization of B-Lines in Lung Ultrasound

Lung ultrasound (LUS) is an important imaging modality used by emergency physicians to assess pulmonary congestion at the patient bedside. B-line artifacts in LUS videos are key findings associated with pulmonary congestion. Not only can the interpretation of LUS be challenging for novice operators, but visual quantification of B-lines remains subject to observer variability. In this work, we investigate the strengths and weaknesses of multiple deep learning approaches for automated B-line detection and localization in LUS videos. We curate and publish, BEDLUS, a new ultrasound dataset comprising 1,419 videos from 113 patients with a total of 15,755 expert-annotated B-lines. Based on this dataset, we present a benchmark of established deep learning methods applied to the task of B-line detection. To pave the way for interpretable quantification of B-lines, we propose a novel "single-point" approach to B-line localization using only the point of origin. Our results show that (a) the area under the receiver operating characteristic curve ranges from 0.864 to 0.955 for the benchmarked detection methods, (b) within this range, the best performance is achieved by models that leverage multiple successive frames as input, and (c) the proposed single-point approach for B-line localization reaches an F1-score of 0.65, performing on par with the inter-observer agreement. The dataset and developed methods can facilitate further biomedical research on automated interpretation of lung ultrasound with the potential to expand the clinical utility.

eess.IV↗

arXiv2019

Recruiting Hay to Find Needles: Recursive Incentives and Innovation in Social Networks

Finding innovative solutions to complex problems is often about finding people who have access to novel information and alternative viewpoints. Research has found that most people are connected to each other through just a few degrees of separation, but successful social search is often difficult because it depends on people using their weak ties to make connections to distant social networks. Recursive incentive schemes have shown promise for social search by motivating people to use their weak ties to find distant targets, such as specific people or even weather balloons placed at undisclosed locations. Here, we report on a case study of a similar recursive incentive scheme for finding innovative ideas. Specifically, we implemented a competition to reward individuals(s) who helped refer Grand Prize winner(s) in MIT's Climate CoLab, an open innovation platform for addressing global climate change. Using data on over 78,000 CoLab members and over 36,000 people from over 100 countries who engaged with the referral contest, we find that people who are referred using this method are more likely than others to submit proposals and to submit high quality proposals. Furthermore, we find suggestive evidence that among the contributors referred via the contest, those who had more than one degree of separation from a pre-existing CoLab member were more likely to submit high quality proposals. Thus, the results from this case study are consistent the theory that people from distant networks are more likely to provide innovative solutions to complex problems. More broadly, the results suggest that rewarding indirect intermediaries in addition to final finders may promote effective social network recruitment.

cs.SI↗