Search Articles

View query in Help articles search

Search Results (1 to 10 of 62 Results)

Download search results: CSV END BibTex RIS


Harnessing Moderate-Sized Language Models for Reliable Patient Data Deidentification in Emergency Department Records: Algorithm Development, Validation, and Implementation Study

Harnessing Moderate-Sized Language Models for Reliable Patient Data Deidentification in Emergency Department Records: Algorithm Development, Validation, and Implementation Study

We have selected 3 language models that share the following 2 characteristics: being open-source and of sufficiently small size for the production phase to be implemented on affordable PC-type systems. These are Llama 2 7 B, Mistral 7 B, and Mixtral. Llama 2 7 B is developed by Meta. Launched in 2023, this is a 7-billion-parameter model, which is claimed to exhibit a good balance between performance and efficiency. We also selected the Mistral 7 B model, introduced to the public in October 2023.

Océane Dorémus, Dylan Russon, Benjamin Contrand, Ariel Guerra-Adames, Marta Avalos-Fernandez, Cédric Gil-Jardiné, Emmanuel Lagarde

JMIR AI 2025;4:e57828

Reflections of Foster Youth Engaging in the Co-Design of Digital Mental Health Technology: Duoethnography Study

Reflections of Foster Youth Engaging in the Co-Design of Digital Mental Health Technology: Duoethnography Study

While the sample size of 4 FAB members contributing to the data collected for the TA is small, it reflects the number of young people with lived expertise in foster care who contributed to the development of Fostr Space and their reflections on the co-design process.

Ifunanya Ezimora, Tylia Lundberg, Dylan Miars, Jeruel Trujeque, Ashley Papias, Margareth V Del Cid, Johanna B Folk, Marina Tolou-Shams

JMIR Form Res 2025;9:e53231

Expanding a Health Technology Solution to Address Therapist Challenges in Implementing Homework With Adult Clients: Mixed Methods Study

Expanding a Health Technology Solution to Address Therapist Challenges in Implementing Homework With Adult Clients: Mixed Methods Study

Therapists were mostly mental health counselors (39/100, 39%), psychologists (28/100, 28%), and social workers (22/100, 22%) who worked in individual practice (55/100, 55%) and small clinic (33/100, 33%) settings and were primarily reimbursed via private (53/100, 53%) or public (32/100, 32%) insurance.

Brian E Bunnell, Kaitlyn R Schuler, Julia Ivanova, Lea Flynn, Janelle F Barrera, Jasmine Niazi, Dylan Turner, Brandon M Welch

JMIR Hum Factors 2024;11:e56567

Smartphone Screen Time Characteristics in People With Suicidal Thoughts: Retrospective Observational Data Analysis Study

Smartphone Screen Time Characteristics in People With Suicidal Thoughts: Retrospective Observational Data Analysis Study

However, the observed difference sizes were relatively small, and no statistical significance between i OSs was found. It is important to note that the absence of statistical significance does not necessarily imply the absence of a true effect. Nevertheless, given the substantial sample size, we deemed the results sufficient for our analysis.

Marta Karas, Debbie Huang, Zachary Clement, Alexander J Millner, Evan M Kleiman, Kate H Bentley, Kelly L Zuromski, Rebecca G Fortgang, Dylan DeMarco, Adam Haim, Abigail Donovan, Ralph J Buonopane, Suzanne A Bird, Jordan W Smoller, Matthew K Nock, Jukka-Pekka Onnela

JMIR Mhealth Uhealth 2024;12:e57439

Evaluation of a Musculoskeletal Digital Assessment Routing Tool (DART): Crossover Noninferiority Randomized Pilot Trial

Evaluation of a Musculoskeletal Digital Assessment Routing Tool (DART): Crossover Noninferiority Randomized Pilot Trial

A minimum sample size of 76 participants was chosen based on the estimated stepped rules of thumb from Whitehead et al [40] to demonstrate an extra small, standardized effect size (SD The Haydock Medical Centre is a well-established multidisciplinary primary care practice in the Northwest of England, with 50 staff and clinicians, serving over 15,000 patients.

Cabella Lowe, Ruth Sephton, William Marsh, Dylan Morrissey

JMIR Form Res 2024;8:e56715