Search Articles

View query in Help articles search

Search Results (1 to 8 of 8 Results)

Download search results: CSV END BibTex RIS


Automated Radiology Report Labeling in Chest X-Ray Pathologies: Development and Evaluation of a Large Language Model Framework

Automated Radiology Report Labeling in Chest X-Ray Pathologies: Development and Evaluation of a Large Language Model Framework

The extraction of labels from unstructured radiology reports is the task of radiology report labeling and it provides us structured information which can be used for many downstream tasks such as medical report generation and natural language explanation generation. It also enables training of large-scale medical imaging models [1].

Abdullah Abdullah, Seong Tae Kim

JMIR Med Inform 2025;13:e68618

Gamified Crowdsourcing as a Novel Approach to Lung Ultrasound Data Set Labeling: Prospective Analysis

Gamified Crowdsourcing as a Novel Approach to Lung Ultrasound Data Set Labeling: Prospective Analysis

Crowdsourced approaches to data set labeling are growing in popularity, and beneficial effects of crowdsourcing have been demonstrated in health care–related tasks including biomedical imaging analysis [10-14]. Using crowdsourcing for biomedical image labeling is challenged by the complexity of the tasks and the need to ensure label quality control. The user interface design for collecting crowd opinions and the metrics used for assessing opinion quality is key to successful results.

Nicole M Duggan, Mike Jin, Maria Alejandra Duran Mendicuti, Stephen Hallisey, Denie Bernier, Lauren A Selame, Ameneh Asgari-Targhi, Chanel E Fischetti, Ruben Lucassen, Anthony E Samir, Erik Duhaime, Tina Kapur, Andrew J Goldsmith

J Med Internet Res 2024;26:e51397

Scalable Approach to Consumer Wearable Postmarket Surveillance: Development and Validation Study

Scalable Approach to Consumer Wearable Postmarket Surveillance: Development and Validation Study

This set was manually labeled by a single data scientist, using a labeling guideline (Multimedia Appendix 1) that was developed as part of the test set generation. The development set contained 100 positive notes (prevalence=0.17). We then derived a labeler model that used weak supervision to probabilistically assign labels for the training set. Specifically, as shown in Figure 1, we used data programming [15], where labeling heuristics are expressed as code-based labeling functions.

Richard M Yoo, Ben T Viggiano, Krishna N Pundi, Jason A Fries, Aydin Zahedivash, Tanya Podchiyska, Natasha Din, Nigam H Shah

JMIR Med Inform 2024;12:e51171

The Effects of an Educational Intervention About Front-of-Package Labeling on Food and Beverage Selection Among Children and Their Caregivers: Protocol for a Randomized Controlled Trial

The Effects of an Educational Intervention About Front-of-Package Labeling on Food and Beverage Selection Among Children and Their Caregivers: Protocol for a Randomized Controlled Trial

Labeling of ultraprocessed products has been a common strategy that governments worldwide use to improve people’s diets [4]. In Mexico, it has been estimated that more than 30% of the total energy consumed comes from ultraprocessed foods, favoring the prevalence of overweight and obesity.

Diana Avila-Montiel, Jenny Vilchis-Gil, América Liliana Miranda-Lora, Lubia Velázquez-López, Miguel Klünder-Klünder

JMIR Res Protoc 2024;13:e54783

Agreement Between Experts and an Untrained Crowd for Identifying Dermoscopic Features Using a Gamified App: Reader Feasibility Study

Agreement Between Experts and an Untrained Crowd for Identifying Dermoscopic Features Using a Gamified App: Reader Feasibility Study

It is essential that some level of quality assurance take place for crowdsourced annotations in the absence of expert labeling for comparison, as would be the case in future studies. Although agreement is traditionally considered an indicator of data reliability, it has been suggested that participants’ competence and confidence should be taken into account [55].

Jonathan Kentley, Jochen Weber, Konstantinos Liopyris, Ralph P Braun, Ashfaq A Marghoob, Elizabeth A Quigley, Kelly Nelson, Kira Prentice, Erik Duhaime, Allan C Halpern, Veronica Rotemberg

JMIR Med Inform 2023;11:e38412

Interactive Medical Image Labeling Tool to Construct a Robust Convolutional Neural Network Training Data Set: Development and Validation Study

Interactive Medical Image Labeling Tool to Construct a Robust Convolutional Neural Network Training Data Set: Development and Validation Study

However, the process of labeling the images for the training of a CNN in a supervised algorithm is hard work and requires extensive time and effort by a health professional. In current CNN training models, the labeling of the data set samples is a critical and important phase. In pretrained classification networks, images have been labeled using polygonal contour tools that help detect objects, parts of a body, animals, and so on [12]. For tissue classification, more detailed labeling is required.

David Reifs, Ramon Reig-Bolaño, Marta Casals, Sergi Grau-Carrion

JMIR Med Inform 2022;10(8):e37284

Availability, Formulation, Labeling, and Price of Low-sodium Salt Worldwide: Environmental Scan

Availability, Formulation, Labeling, and Price of Low-sodium Salt Worldwide: Environmental Scan

However, little is known about the availability and accessibility of this emerging product in the global market or about factors that may affect equitable uptake including formulation, price, or labeling. We, therefore, performed a systematic search of low-sodium salts to understand better their availability, formulation, labeling, and the price in different countries. This study was a systematic search of low-sodium salts conducted from October 2019 to September 2020.

Xuejun Yin, Hueiming Liu, Jacqui Webster, Kathy Trieu, Mark D Huffman, J Jaime Miranda, Matti Marklund, Jason H Y Wu, Laura K Cobb, Ka Chun Li, Sallie-Anne Pearson, Bruce Neal, Maoyi Tian

JMIR Public Health Surveill 2021;7(7):e27423