Joanna Lin

Joanna Lin

Auto-generated profile, based on publicly available information

Joanna Lin

University of Calgary, Alberta, Canada (J.L., M.A., R.S., J.O., M.B.).

Research Focus & Metrics

Academic Career Length

1Year (2026-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 February 2026 in Digital Health
Verified
Journal Article
Open access
MSMansi Singhal
Mansi Singhal
JLJoanna Lin
Joanna Lin
MDMegan Delehanty
Megan Delehanty
AKAli Karimi
Ali Karimi
LRLeticia Rittner
Leticia Rittner
MBMariana Bento
Mariana Bento
Introduction: Alzheimer’s disease (AD) affects millions worldwide, and researchers heavily rely on datasets for diagnosis and treatment. Identifying relevant datasets is challenging due to data gaps and bias related to the demographics and geographic origin. Method: We investigated AD data gaps by identifying and manually curating publicly accessible AD datasets containing imaging and/or tabular data. We also extracted key information such as data availability, geographic location, and participant demographics. We used five Large Language Models (LLMs) to identify AD datasets, allowing us to explore potential datasets while also evaluating retrieval consistency across models. Result: We identified 24 publicly accessible AD datasets (open access or controlled access via registration). These datasets enabled us to emphasize three critical gaps: (1) variability in AD dataset retrieval, as observed through differences in LLM outputs, related to dataset visibility and accessibility; (2) geographical imbalance, with North America contributing 55.6% of datasets, US alone 66.7%, followed by Europe at 36.1%, and smaller shares from South America 11.1%, Asia 8.3%, and Africa 2.8%; and (3) demographic deficits, with the majority of datasets predominantly White, as 9 of 24 had over 80% White participants. Among the seven datasets that reported any Black participant representation, the proportion of Black participants ranged from 15.3% to 18.8%. Conclusion: These findings reveal significant disparities in the availability and retrieval of AD datasets, with most data concentrated in Western countries and critical gaps in demographic representation. LLMs show inconsistent retrieval, particularly for newer, smaller, or region-specific datasets, which may perpetuate existing biases.

12 April 2026
Verified
Conference Paper
Open access
IJIqra Jawaid
Iqra Jawaid
LVLucas Rodrigues Valença
Lucas Rodrigues Valença
AMAyah Metwali
Ayah Metwali
JLJoanna Lin
Joanna Lin
ZAZahra Abbasi
Zahra Abbasi
RSRonnie de Souza Santos
Ronnie de Souza Santos

1 November 2026 in Stroke
Verified
Journal Article
Closed access
JLJoanna Lin
Joanna Lin
MAMumu Aktar
Mumu Aktar
HBHakim Baazaoui
Hakim Baazaoui
Susanne WegenerSusanne Wegener
Susanne Wegener
RSRoberto Souza
Roberto Souza
MBMariana Bento
Mariana Bento
Artificial intelligence is increasingly integrated into acute stroke management, demonstrating strong performance in tasks such as large vessel occlusion detection, Alberta Stroke Program Early Computed Tomography Score scoring, and functional outcome prediction. However, concerns persist regarding both the black-box nature of these models and algorithmic bias that may exacerbate disparities in stroke incidence, treatment, and outcomes across racial and socioeconomic subgroups. This review synthesizes the current literature on explainable artificial intelligence and fairness in artificial intelligence applications for acute stroke management, identifies persistent challenges, and outlines recommendations for the development of equitable and trustworthy systems in stroke care. Recent studies have increasingly adopted post hoc explainability methods, though these are limited by approximation and misleading interpretations, especially since explanations are rarely formally tested. Explainability and fairness remain largely disconnected, with fairness evaluation remaining uncommon due to limited demographic metadata, regulatory constraints, and the absence of stroke-specific fairness criteria. Generalizability also remains a concern due to suboptimal data set partitioning strategies and inadequate reporting practices. Responsible artificial intelligence for acute stroke management requires unified evaluation frameworks that jointly assess explainability, fairness, and generalizability.