Graph-Based Fusion of Imaging and Non-Imaging Data for Disease Trajectory Prediction

This study proposes a graph convolutional neural networks (GCN) architecture for fusion of radiological imaging and non-imaging tabular electronic health records (EHR) for the purpose of clinical event prediction. We focused on a cohort of hospitalized patients with positive RT-PCR test for COVID-19 and developed GCN based models to predict three dependent clinical events (discharge from hospital, admission into ICU, and mortality) using demographics, billing codes for procedures and diagnoses and chest X-rays. We hypothesized that the two-fold learning opportunity provided by the GCN is ideal for fusion of imaging information and tabular data as node and edge features, respectively. Our experiments indicate the validity of our hypothesis where GCN based predictive models outperform single modality and traditional fusion models. We compared the proposed models against two variations of imaging-based models, including DenseNet-121 architecture with learnable classification layers and Random Forest classifiers using disease severity score estimated by pre-trained convolutional neural network. GCN based model outperforms both imaging-only methods. We also validated our models on an external dataset where GCN showed valuable generalization capabilities. We noticed that edge-formation function can be adapted even after training the GCN model without limiting application scope of the model. Our models take advantage of this fact for generalization to external data. © 2023 IEEE.

A Systematic Review of 'Fair' AI Model Development for Image Classification and Prediction

Purpose: The new challenge in Artificial Intelligence (AI) is to understand the limitations of models to reduce potential harm. Particularly, unknown disparities based on demographic factors could encrypt currently existing inequalities worsening patient care for some groups. Method(s): Following PRISMA guidelines, we present a systematic review of 'fair' deep learning modeling techniques for natural and medical image applications which were published between year 2011 to 2021. Our search used Covidence review management software and incorporates articles from PubMed, IEEE, and ACM search engines and three reviewers independently review the manuscripts. Result(s): Inter-rater agreement was 0.89 and conflicts were resolved by obtaining consensus between three reviewers. Our search initially retrieved 692 studies but after careful screening, our review included 22 manuscripts that carried four prevailing themes;'fair' training dataset generation (4/22), representation learning (10/22), model disparity across institutions (5/22) and model fairness with respect to patient demographics (3/22). We benchmark the current literature regarding fairness in AI-based image analysis and highlighted the existing challenges. We observe that often discussion regarding fairness are limited to analyzing existing bias without further establishing methodologies to overcome model disparities. Conclusion(s): Based on the current research trends, exploration of adversarial learning for demographic/camera/institution agnostic models is an important direction to minimize disparity gaps for imaging. Privacy preserving approaches also present encouraging performance for both natural and medical image domain. Copyright © 2022, Taiwanese Society of Biomedical Engineering.

The impact of hepatic steatosis on COVID-19 related outcomes

Background: Metabolic factors, such as diabetes, portend an increased risk of adverse outcomes in patients with COVID-19 We evaluated the association of radiographic hepatic steatosis with length of stay (LOS), need for intensive care unit (ICU), and mortality in patients admitted with COVID-19 Methods: We retrospectively identified a cohort of patients with COVID-19 admitted to two US medical centers from March 1, 2020 to June 22, 2020 All patients with abdominal imaging (computed tomography, magnetic resonance imaging, or ultrasound) within the last year were eligible Patient demographics, body mass index (BMI), comorbidities, baseline and admission laboratory values were collected Statistical comparisons were made using Mann- Whitney U-test for continuous measures and Chi-square and Fisher's exact test for categorical measures Multivariable linear and Adjusted covariates included age, sex, hemoglobin A1c, admission ALT, admission albumin, type of imaging, andpresence of chronic liver disease and heart disease Results: A total of 319 eligible patients (50% women and mean age of 63.7) were identified. 14% of patients had radiographic steatosis. There were no significant differences among age, race, BMI, or presence of diabetes or heart disease between those with steatosis and those without However, patients with steatosis had more chronic liver disease (40% vs 23%, p=0 02) and higher baseline ALT levels (28 IU/L vs 17 IU/L, p=0 002), but not more cirrhosis (12 8% vs 15 6%, p=0 63) The median LOS in those with steatosis was 6 1 days (4- 16), while 9 days (4-18) for those without steatosis (p=0 6) The need for ICU (24 4% vs 32 1%, p=0 3) and mortality rate (6.7% vs 16.5%, p=0.12) were not significantly different. On multivariable analysis, there was no difference in hospital LOS in those with steatosis compared to those without (b: -16 97, 95% CI: -35-1 05, P-value=0 06) Additionally, there was no difference in need for ICU care (OR: 0 19, 95% CI: 0 01-2 44, P-value=0 19) or mortality rates (OR: 1 4, 95% CI: 0 07-26 9, P-value=0 8) between the two groups on multivariable analysis Conclusion: Among patients with COVID-19 admitted to the hospital, those with radiographic hepatic steatosis do not have worse outcomes as measured by LOS, need for ICU, or in-hospital mortality Larger samples with more precise staging of liver disease may be required to identify subtle differences, if present.