1-20 of 2671
Journal Article
“Everything is electronic health record-driven”: the role of the electronic health record in the emergency department diagnostic process
JAMIA Open, Volume 8, Issue 2, April 2025, ooaf029, https://doi.org/10.1093/jamiaopen/ooaf029
Published: 23 April 2025
Image
Alignment of identified themes related to EHR use in the ED diagnostic care...
Published: 23 April 2025
Figure 1.
Alignment of identified themes related to EHR use in the ED diagnostic care process on the ED-adapted NASEM diagnostic safety framework. This figure aligns the themes identified in this study with the a conceptual framework showing the process of ED care, from left to right: patient experiences &
Journal Article
Correction to: Leveraging deep learning to detect stance in Spanish tweets on COVID-19 vaccination
JAMIA Open, Volume 8, Issue 2, April 2025, ooaf028, https://doi.org/10.1093/jamiaopen/ooaf028
Published: 15 April 2025
Image
Model-derived mortality probability in (A) ARDS and (B) sepsis cohorts. The...
Published: 10 April 2025
Figure 2.
Model-derived mortality probability in (A) ARDS and (B) sepsis cohorts. The green line represents patients who were eventually discharged alive from ICU, while the orange line represents patients who died in ICU. Probability of mortalities is aggregated over repeated hours since admission to show th
Journal Article
A deep learning model for clinical outcome prediction using longitudinal inpatient electronic health records
JAMIA Open, Volume 8, Issue 2, April 2025, ooaf026, https://doi.org/10.1093/jamiaopen/ooaf026
Published: 10 April 2025
Image
The TECO algorithm design. This figure demonstrates a 24-hour mortality pre...
Published: 10 April 2025
Figure 1.
The TECO algorithm design. This figure demonstrates a 24-hour mortality prediction example, utilizing data from the preceding 96 hours to predict the binary outcome (death vs non-death). Time-dependent, ICU monitoring variables were aligned and concatenated with baseline variables, then embedded int
Image
Feature importance analysis in the ARDS cohort using (A) random forest and ...
Published: 10 April 2025
Figure 3.
Feature importance analysis in the ARDS cohort using (A) random forest and (B) TECO. Random forest feature importance was assessed using permutation importance, with each feature permuted 20 times. Transformer-based, Encounter-level Clinical Outcome feature importance was estimated via expected grad
Image
The prompt refinement process. Sankey diagram illustrating the iterative...
Published: 09 April 2025
Figure 1.
The prompt refinement process. Sankey diagram illustrating the iterative process of prompt development and refinement used in this study. The diagram traces the progression from the initial set of 8 draft prompts through subsequent revisions and additions, culminating in the final 11 prompts used
Image
Correlation between LLM model type and prompt series, the average score for...
Published: 09 April 2025
Figure 2.
Correlation between LLM model type and prompt series, the average score for individual metrics and overall score. Heatmap visualization depicting the performance of each large language model (LLM) across the 8 evaluative criteria used in this study. Scores are stratified by prompt series, with da
Journal Article
A proof-of-concept study for patient use of open notes with large language models
JAMIA Open, Volume 8, Issue 2, April 2025, ooaf021, https://doi.org/10.1093/jamiaopen/ooaf021
Published: 09 April 2025
Image
Average score per metric by the clinician and patient rater. Box plots i...
Published: 09 April 2025
Figure 3.
Average score per metric by the clinician and patient rater. Box plots illustrating the average scores assigned by the clinician and patient rater across the 8 evaluation metrics. Each box represents the distribution of scores for a given metric. The scores highlight variability and similarities
Image
Study 3 outcomes: (A) grade-level readability scores by PLSA type and (B) m...
Published: 03 April 2025
Figure 5.
Study 3 outcomes: (A) grade-level readability scores by PLSA type and (B) medical writer (top)- and patient/patient advocate (bottom)-assessed qualitative readability metrics by PLSA type. ARI = automated readability index; BAI = bespoke artificial intelligence; CRAS = comprehensive readability asse
Journal Article
Using artificial intelligence to expedite and enhance plain language summary abstract writing of scientific content
JAMIA Open, Volume 8, Issue 2, April 2025, ooaf023, https://doi.org/10.1093/jamiaopen/ooaf023
Published: 03 April 2025
Image
Study 1 outcomes: (A) mean grade-level readability and (B) reading time by ...
Published: 03 April 2025
Figure 2.
Study 1 outcomes: (A) mean grade-level readability and (B) reading time by PLSA type. ANOVA = analysis of variance; ARI = automated readability index; BAI = bespoke artificial intelligence; MW = medical writer; OSA = original scientific abstract; PLSA = plain language summary abstract; SD = standard
Image
Study 1 outcome: PLSA accessibility as measured by US OECD PIAAC population...
Published: 03 April 2025
Figure 3.
Study 1 outcome: PLSA accessibility as measured by US OECD PIAAC population literacy data ( n = 3892). BAI = bespoke artificial intelligence; MW = medical writer; OECD = organization for economic co-operation and development; OSA = original scientific abstract; PIAAC = programme for the internation
Image
Study designs. AI = artificial intelligence; ARI = automated readability in...
Published: 03 April 2025
Figure 1.
Study designs. AI = artificial intelligence; ARI = automated readability index; BAI = bespoke artificial intelligence; CRAS = comprehensive readability assessment scale; MW = medical writer; NBAI = non-bespoke artificial intelligence; OECD = organization for economic co-operation and development; OS
Image
Study 2 outcomes: (A) time to complete task by condition, (B) effort to com...
Published: 03 April 2025
Figure 4.
Study 2 outcomes: (A) time to complete task by condition, (B) effort to complete task by condition, (C) SME-assessed accuracy by PLSA type, and (D) PCP clarity assessment by PLSA type. AI = artificial intelligence; ANOVA = analysis of variance; BAI = bespoke artificial intelligence; MW = medical wri
Journal Article
Computer-assisted prescription of erythropoiesis-stimulating agents in patients undergoing maintenance hemodialysis: a randomized control trial for artificial intelligence model selection
JAMIA Open, Volume 8, Issue 2, April 2025, ooaf020, https://doi.org/10.1093/jamiaopen/ooaf020
Published: 27 March 2025
Image
Study flow diagram. Flowchart depicting enrollment of 124 patients, rand...
Published: 27 March 2025
Figure 2.
Study flow diagram. Flowchart depicting enrollment of 124 patients, randomized based on age, gender, and HD vintage into Control (Dr) and Intervention (AI) groups with 62 patients each. In the Control group, 8 withdrew: 5 due to blood transfusions, 1 due to death, and 2 for other reasons, leaving
Image
(A) Primary outcome: AI versus Dr on the mean absolute difference between H...
Published: 27 March 2025
Figure 3.
(A) Primary outcome: AI versus Dr on the mean absolute difference between Hb and 11 g/dL by various models. The margin is set at 0.25 g/dL, assuming that the mean absolute difference between Hb and 11 g/dL is 0.75-0.8 g/dL in the dataset, allowing 0.20-0.25 g/dL as the tolerable space to keep Hb bet
