e.g. mhealth
Search Results (1 to 10 of 16 Results)
Download search results: CSV END BibTex RIS
Skip search results from other journals and go to results- 9 JMIR Medical Informatics
- 2 Journal of Medical Internet Research
- 1 JMIR AI
- 1 JMIR Formative Research
- 1 JMIR Infodemiology
- 1 JMIR Mental Health
- 1 JMIR mHealth and uHealth
- 0 Medicine 2.0
- 0 Interactive Journal of Medical Research
- 0 iProceedings
- 0 JMIR Research Protocols
- 0 JMIR Human Factors
- 0 JMIR Public Health and Surveillance
- 0 JMIR Serious Games
- 0 JMIR Rehabilitation and Assistive Technologies
- 0 JMIR Preprints
- 0 JMIR Bioinformatics and Biotechnology
- 0 JMIR Medical Education
- 0 JMIR Cancer
- 0 JMIR Challenges
- 0 JMIR Diabetes
- 0 JMIR Biomedical Engineering
- 0 JMIR Data
- 0 JMIR Cardio
- 0 Journal of Participatory Medicine
- 0 JMIR Dermatology
- 0 JMIR Pediatrics and Parenting
- 0 JMIR Aging
- 0 JMIR Perioperative Medicine
- 0 JMIR Nursing
- 0 JMIRx Med
- 0 JMIRx Bio
- 0 Transfer Hub (manuscript eXchange)
- 0 JMIR Neurotechnology
- 0 Asian/Pacific Island Nursing Journal
- 0 Online Journal of Public Health Informatics
- 0 JMIR XR and Spatial Computing (JMXR)
In this study, we used a multihead attention transformer encoder, drawing inspiration from BERT [8,9]. The model architecture, illustrated in Figure 3, includes an embedding module, a multihead attention transformer encoder, a feed-forward layer, and 2 classifier heads. The embedding module integrates medical codes with their associated temporal information.
JMIR Med Inform 2025;13:e68138
Download Citation: END BibTex RIS
Enhancing Antidiabetic Drug Selection Using Transformers: Machine-Learning Model Development
The field of ML has undergone significant advancement since the introduction of the transformer approach in 2017 [28,29]. Transformers enable contextual interactions for natural language processing tasks and have become a core technology across diverse domains [30-32]. The transformer model incorporates an attention mechanism and has shown remarkable performance in tasks involving the extraction of temporal and semantic relationships, leading to success in tasks such as generation and classification [33].
JMIR Med Inform 2025;13:e67748
Download Citation: END BibTex RIS
The tokenized input was then processed through the Ko BERT transformer encoder, followed by a fully connected classification layer to predict the sentiment as positive, neutral, or negative. For training, we used the Adam optimizer with a learning rate of 5×10-5, a warmup ratio of 0.1, and gradient clipping with a max norm of 1 to stabilize learning. The model was trained for 3 epochs with a batch size of 64. To evaluate computational performance, the training and inference times were measured.
JMIR Med Inform 2025;13:e65127
Download Citation: END BibTex RIS
Pretrained transformer language models, like bidirectional encoder representations from transformers (BERT) [16-18], have outperformed the existing deep neural network models, including convolutional neural networks and recurrent neural networks. Examples of transformer-based models trained on biomedical data include Bio BERT [19], Bio Link BERT [20], Blue BERT [21], and Bio Med BERT [22], which are pretrained on biomedical literature and clinical text.
JMIR Med Inform 2025;13:e63267
Download Citation: END BibTex RIS
The technique of modeling words in a large context has been referred to as transformer-based large language modeling [8]. This may not only facilitate the automatic analysis of large amounts of text data [9,10] but, by modeling words in a large context, also allow the generation of meaningful text and the interactive use of this technology [5,10]. Thus, the application of LLMs may improve efficiency and effectiveness of data processing in various fields—including health care [5].
JMIR Ment Health 2025;12:e57986
Download Citation: END BibTex RIS
Transformers for Neuroimage Segmentation: Scoping Review
In this review, we focused on the deep transformer–based techniques that have gained more attention recently. From the proposed models, we can find transformer-based, CNN with transformer-based, and generative adversarial network with transformer-based techniques. At the same time, the methods based on Trans BTS, Trans UNet, Sein UNet, and U-Net with transformer are the most used models for neuroimage segmentation. Figure 3 illustrates these models in terms of architecture.
J Med Internet Res 2025;27:e57723
Download Citation: END BibTex RIS
Zhang and Jankowski [25] proposed average pooling transformer layers handling token-, sentence-, and document-level embeddings for classifying International Classification of Diseases codes. Their model outperformed the BERT-base model by 11 points.
For the reading comprehension task, BERT can be used to determine the answer span within a given text. Pampari et al [26] proposed the electronic medical record question answering (emr QA) dataset to determine the answer span to a question in a clinical context.
JMIR Med Inform 2024;12:e52897
Download Citation: END BibTex RIS
Since its introduction in 2017, the transformer model has excelled in various time-series predictive tasks, solidifying its position as a core technology across multiple fields [27-32]. The transformer model incorporates an attention mechanism simplifying the extraction of temporal relationships and setting it apart from other models [33-35].
JMIR AI 2024;3:e56700
Download Citation: END BibTex RIS
We proposed a novel E4-tailored transformer (E4mer) architecture (Figure 2) [43] and showed that SSL is a viable paradigm, outperforming both fully supervised E4mer and classical machine learning (CML) models using handcrafted features in distinguishing acute MD episodes from clinical stability (euthymia in psychiatric parlance), that is, a time-series (binary) classification task.
We investigated what makes SSL successful.
JMIR Mhealth Uhealth 2024;12:e55094
Download Citation: END BibTex RIS
Although similar, this study differs in the following ways: (1) we survey a wide range of supervised ML algorithms as opposed to only using lasso regularized logistic regression, (2) we use transformer embeddings of sentences to represent information in each patient’s records and aggregate these embeddings with multilayer perceptron (MLP) networks to create a patient vector representation for patient phenotyping, and (3) we performed an ablation study of processing methods to compare the impact on performance
JMIR Form Res 2024;8:e52200
Download Citation: END BibTex RIS