Search Articles

View query in Help articles search
e.g. John Smith or 0000-0001-6479-5330

Search Results (1 to 10 of 1721 Results)

Download search results: CSV END BibTex RIS

Improving Early Dementia Detection Among Diverse Older Adults With Cognitive Concerns With the 5-Cog Paradigm: Protocol for a Hybrid Effectiveness-Implementation Clinical Trial

Improving Early Dementia Detection Among Diverse Older Adults With Cognitive Concerns With the 5-Cog Paradigm: Protocol for a Hybrid Effectiveness-Implementation Clinical Trial

(D) IUH: negative 5-Cog result. At both sites, a paper token (Figure 4) is used as an additional feature to help ensure that care providers review the patients’ 5-Cog results. Patients are handed this token after they complete the 5-Cog battery and are asked to hand it to their care provider at their scheduled visit, generally within 30 minutes after the 5-Cog battery administration. Token to alert care provider for patient's 5-Cog participation.

Rachel Beth Rosansky Chalmer, Emmeline Ayers, Erica F Weiss, Nicole R Fowler, Andrew Telzak, Diana Summanwar, Jessica Zwerling, Cuiling Wang, Huiping Xu, Richard J Holden, Kevin Fiori, Dustin D French, Celeste Nsubayi, Asif Ansari, Paul Dexter, Anna Higbie, Pratibha Yadav, James M Walker, Harrshavasan Congivaram, Dristi Adhikari, Mairim Melecio-Vazquez, Malaz Boustani, Joe Verghese

JMIR Res Protoc 2025;14:e60471

Download Citation: END BibTex RIS

Download PDF Download XML
Article has an altmetric score of 1

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

Plot of recall by epoch for PII: (A) Location, (B) Telephone, (C) Name, (D) Date. PII: personal identifying information. BLEU-4 scores were calculated to assess whether the models modified the texts at the note level. During the deidentification process, medical texts remained almost unchanged as demonstrated by a consistently high BLEU-4 score (Figure 6) beyond epoch 5. Plot of BLEU score by epoch: clinical note as statistical unit.

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

JMIR AI 2025;4:e57828

Download Citation: END BibTex RIS

Download PDF Download XML

Limitations of Binary Classification for Long-Horizon Diagnosis Prediction and Advantages of a Discrete-Time Time-to-Event Approach: Empirical Analysis

Limitations of Binary Classification for Long-Horizon Diagnosis Prediction and Advantages of a Discrete-Time Time-to-Event Approach: Empirical Analysis

Each observation was represented by the triplet {X,T,S}, where X⊆Rd is a d-dimensional feature vector, T∈(0,Emax] is an observed event or censoring time over a finite time horizon, and S∈{0,1} indicates whether T is a right-censoring time (S=0) or an event time (S=1). The observed time T is the minimum of the event time E and the right-censoring time C, that is, T=min(E, C).

De Rong Loh, Elliot D Hill, Nan Liu, Geraldine Dawson, Matthew M Engelhard

JMIR AI 2025;4:e62985

Download Citation: END BibTex RIS

Download PDF Download XML