https://journals-sol.sbc.org.br/index.php/jdhbi/issue/feed Journal of Digital Health and Biomedical Informatics 2026-05-29T21:32:08+00:00 Marcia Ito jdhbieditorchefe@gmail.com Open Journal Systems <p data-start="124" data-end="540">The <em data-start="128" data-end="182">Journal of Digital Health and Biomedical Informatics</em> (JDHBI) is an open-access journal of the Brazilian Computing Society, dedicated to <strong data-start="266" data-end="284">digital health</strong> and the use of technologies to support <strong data-start="324" data-end="340">human health</strong>. It publishes original research and reviews on digital innovations, health information systems, artificial intelligence, telemedicine, data analysis, and other digital technologies applied to human health.</p> <p data-start="547" data-end="754">The journal also accepts studies in the <strong data-start="587" data-end="601">One Health</strong> context, provided they explore the intersection between technology, human health, animal health, and ecosystems, with a direct impact on human health.</p> <p data-start="761" data-end="894"><strong data-start="761" data-end="892">Articles focusing exclusively on animal or veterinary health, without relation to human health or ecosystems, are not accepted.</strong></p> https://journals-sol.sbc.org.br/index.php/jdhbi/article/view/7916 Tooth Detection and Segmentation on Occlusal Surfaces in Early Mixed Dentition Using Deep Learning 2026-03-24T12:17:11+00:00 Bruna Cristine Dias brunacristined@gmail.com Mateus Felipe de Cássio Ferreira mateus.fecassio@gmail.com Luan Matheus Trindade Dalmazo luantrindade@ufpr.br Luciana Reichert da Silva Assunção lurassuncao@yahoo.com.br Lucas Ferrari de Oliveira lferrari@inf.ufpr.br <p><span dir="ltr" role="presentation">Artificial intelligence, particularly Convolutional Neural Networks (CNNs), has shown great </span><span dir="ltr" role="presentation">potential for the detection of oral health conditions, including dental caries. Objectives: To evaluate the performance </span><span dir="ltr" role="presentation">of CNNs in the detection and segmentation of posterior teeth during the early phase of mixed dentition.</span> <span dir="ltr" role="presentation">Materials and </span><span dir="ltr" role="presentation">Methods:</span> <span dir="ltr" role="presentation">A total of 945 images of posterior teeth were collected in a school-based setting. The dataset was divided into </span><span dir="ltr" role="presentation">training and testing subsets. Three CNN architectures — YOLOv11, U-Net, and DeepLabv3 — were compared based on </span><span dir="ltr" role="presentation">precision, recall, and their ability to detect partially erupted teeth.</span> <span dir="ltr" role="presentation">Results:</span> <span dir="ltr" role="presentation">YOLOv11 achieved a precision of 0.967 and a </span><span dir="ltr" role="presentation">recall of 0.938, successfully identifying 96.1% of partially erupted permanent teeth. U-Net demonstrated a precision of </span><span dir="ltr" role="presentation">0.953 and a recall of 0.951, detecting 78.4% of partially erupted teeth, while DeepLabv3 achieved a precision of 0.963 and </span><span dir="ltr" role="presentation">a recall of 0.944, detecting 76.5% of these teeth. Conclusions: All three CNNs demonstrated high accuracy in detecting </span><span dir="ltr" role="presentation">and segmenting posterior teeth in children. However, YOLOv11 outperformed both U-Net and DeepLabv3 in detecting </span><span dir="ltr" role="presentation">partially erupted teeth, highlighting its potential for use in studies involving early mixed dentition.</span> <span dir="ltr" role="presentation">Clinical Relevance: </span><span dir="ltr" role="presentation">Identifying the most suitable CNN architecture for detecting caries in mixed dentition may help standardize diagnosis and </span><span dir="ltr" role="presentation">reduce clinical time.</span></p> 2026-05-27T00:00:00+00:00 Copyright (c) 2026 Bruna Cristine Dias, Mateus Felipe de Cássio Ferreira, Luan Matheus Trindade Dalmazo, Luciana Reichert da Silva Assunção, Lucas Ferrari de Oliveira https://journals-sol.sbc.org.br/index.php/jdhbi/article/view/6590 Development of predictive risk algorithms for mental health in indoor settings: a cohort study 2025-09-20T15:59:59+00:00 Ana Laura de Sene Amâncio Zara analaurazara@ufg.br Mariana Maria Batista marianamariabatista@egresso.ufg.br Bianca Ribeiro Soares soaresbianca@egresso.ufg.br Myllena D’ávila Mariana Gomes Silva myllena.mariana@discente.ufg.br Joédan Silva Santos joedansantos@egresso.ufg.br Myrella Alves Bordado myrellaalves@discente.ufg.br Júlia Gomes Moreira julia.moreira@egresso.ufg.br Melyssa Mariana Gomes Silva melyssamariana@discente.ufg.br Antônio Emílio Pereira antoniopereira@discente.ufg.br Daniel Fazzioni fazzioni@discente.ufg.br Murilo de Oliveira Guimarães muriloguimaraes@discente.ufg.br Alex Victor Echeverria Ugarte Silva ugarte_alex@egresso.ufg.br Aldo André Díaz Salazar aldo.diaz@ufg.br Arthur Ricardo de Sousa Vitória arthursvrr@gmail.com Paula Geralda Barbosa Coelho paulacoelho@ufg.br Paulo Roberto D’Aurea Machado paulo@dmhealth.com.br Rodrigo Silva Dias rodrigo@dmhealth.com.br Iwens Gervásio Sene Junior iwens@ufg.br Renato de Freitas Bulcão-Neto rbulcao@ufg.br <p>This article describes the methodology and characteristics of a cohort designed to support the development of intelligent predictive algorithms for anxiety and depression risk in indoor settings. This prospective cohort study included students and staff from a university. Symptoms of anxiety and depression were assessed using self-reported questionnaires (PROMIS®). Indoor environments were monitored using devices that measured air quality parameters, including humidity, temperature, and carbon dioxide levels. A total of 601 individuals were recruited, with a retention rate of 85.2%. The majority were male (55.3%), with a median age of 22 years. Dropouts and retained participants differed significantly in terms of sex, age, occupation, chronic illnesses, and general health status. Among participants, the prevalence of anxiety and depression was 54.7% and 41.8%, respectively. The observed prevalence rates of anxiety and depression exceed those typically reported in Brazilian academic settings, which, although concerning, provide a methodological advantage. Including participants with diverse clinical outcomes, particularly those with extreme scores, is crucial for enhancing the precision of predictive algorithms. Therefore, the ideal sample composition should prioritize clinical heterogeneity, reflecting the diversity of real-world healthcare scenarios and directly contributing to improved computational modeling accuracy. Differential sample attrition did not compromise the high prevalence of anxiety and depression, a key feature for developing predictive algorithms with the potential to accurately identify mental health risks in indoor settings.</p> 2026-11-13T00:00:00+00:00 Copyright (c) 2026 Ana Laura de Sene Amâncio Zara, Mariana Maria Batista, Bianca Ribeiro Soares, Myllena D’ávila Mariana Gomes Silva, Joédan Silva Santos, Myrella Alves Bordado, Júlia Gomes Moreira, Melyssa Mariana Gomes Silva, Antônio Emílio Pereira, Daniel Fazzioni, Murilo de Oliveira Guimarães, Alex Victor Echeverria Ugarte Silva, Aldo André Díaz Salazar, Arthur Ricardo de Sousa Vitória, Paula Geralda Barbosa Coelho, Paulo Roberto D’Aurea Machado, Rodrigo Silva Dias, Iwens Gervásio Sene Junior, Renato de Freitas Bulcão-Neto https://journals-sol.sbc.org.br/index.php/jdhbi/article/view/6746 A Platform for Remote Monitoring of Older Adults: The Value of Heart Rate Variability 2025-10-24T14:52:59+00:00 Eujessika Rodrigues eujessikars@gmail.com Paulo Barbosa paulo.barbosa@nutes.uepb.edu.br Daniella Carvalho daniella.lima@nutes.uepb.edu.br Elisa Yumi Nakagawa elisa@icmc.usp.br Sabrina Fernandes sabrinaggf@hotmail.com Ana Tereza Fernandes aninha.sales@gmail.com Álvaro Maciel alvaro.maciel@ufrn.br Lina Garcés linagarces@usp.br <p>Background: <span style="font-weight: 400;">Despite the inherent disabilities of human aging, the number of older adults is growing. Previous works have proposed and explored technologies to improve the quality of life of older adults and mitigate this challenge in the health field. In particular, technologies based on Heart Rate Variability (HRV), measure the decrease of vagal tone and sedentary behaviors in age. Although this measure can be used remotely in older adults’ care, it has been only made available for clinical testing environments, laboratories, and hospitals. To address this gap, in this paper, we present Senior Mobile Health (SMH), a platform that supports the development of a complete system composed of mobile apps deployed in wearables and web systems for the remote monitoring of the HRV data of older adults.</span> Results:<span style="font-weight: 400;"> We conducted a real-world case study with 89 older adults using the SMH and validated the results by comparing them with those provided by a reference clinical tool, the Kubios HRV. </span>Conclusions:<span style="font-weight: 400;"> Our results show evidence that SMH aggregates an important value to older adults’ care assisting in rehabilitation processes by health professionals. </span></p> 2026-11-13T00:00:00+00:00 Copyright (c) 2026 Eujessika Rodrigues, Paulo Barbosa, Daniella Carvalho, Elisa Yumi Nakagawa, Sabrina Fernandes, Ana Tereza Fernandes, Álvaro Maciel, Lina Garcés https://journals-sol.sbc.org.br/index.php/jdhbi/article/view/8115 Improving the visualization of strategic information for public health 2026-04-29T02:26:49+00:00 Lúcia Vilela Leite Filgueiras lfilguei@usp.br Leandro Manuel Reis Velloso leandrovelloso@usp.br Mailson Fontes de Carvalho mailsoncarvalho@ufpi.edu.br Luiz Sérgio de Souza luizsds@usp.br Luciano Anísio Garcia luciano.garcia@usp.br Johne Marcus Jarske johne.jarske@usp.br Tânia Letícia dos Santos tania.leticia2011@gmail.com Gabriela Momberg Araújo gabriela.momberg.araujo@alumni.usp.br Henrique Mathias Fernandes fernandeshm1997@usp.br Solange Nice Alves de Souza ssouza@usp.br Márcia Ito marciaito2000@gmail.com <p>This article presents the results of a project carried out to improve the visualization of strategic information in the Strategic Management Support Room (SAGE) of the Ministry of Health based on the user experience. Using participatory methodology and aspects of the agile cycle, the project started by recognizing the needs of public health managers in the three spheres of government, using Infovis design techniques and user experience evaluation, to develop prototypes of information visualization appropriate to the needs and demands of the user managers. The prototypes were prioritized and elaborated: installed capacity, monitoring results, arbovirosis risk map, mobile emergency care service, crossing indicators, interfederative pacting indicators, and covid-19 projection. The project enhances improvements in access and understanding of strategic information needed for decision making in health, essential for the qualification of the management of the Unified Health System.</p> 2026-11-13T00:00:00+00:00 Copyright (c) 2026 Lúcia Vilela Leite Filgueiras, Leandro Manuel Reis Velloso, Mailson Fontes de Carvalho, Luiz Sérgio de Souza, Luciano Anísio Garcia, Johne Marcus Jarske, Tânia Letícia dos Santos, Gabriela Momberg Araújo, Henrique Mathias Fernandes, Solange Nice Alves de Souza, Márcia Ito https://journals-sol.sbc.org.br/index.php/jdhbi/article/view/8362 Retinal Vessel Segmentation in OCTA Images with U-Net: A Comparative Analysis of Training Strategies 2026-05-29T21:32:08+00:00 Henrique Christoph Bohn hbohn@edu.unisinos.br Cristiano André da Costa cac@unisinos.br Fabio Lavinsky ylavinsky@htomail.com Ismael Garcia Santos ismael@edu.unisinos.br Gregory Stein steingregory@edu.unisinos.br Felipe André Zeiser felipe_zei@hotmail.com <p>Optical coherence tomography angiography (OCTA) is a non-invasive, depth-resolved imaging modality that enables visualization of retinal vascular slabs, including superficial and deeper plexuses. Accurate vessel segmentation is essential for deriving biomarkers such as vessel density and foveal avascular zone area, which support the assessment of diabetic retinopathy and age-related macular degeneration. This study investigates whether task-specific data augmentation or ImageNet-based transfer learning provides greater benefit for large-vessel segmentation under limited labeled data. We evaluated four U-Net-based configurations on the OCTA\_6mm subset of OCTA-500 using a 70/15/15 subject-level split: a U-Net without pretraining or augmentation, the same model with augmentation, and two variants with ImageNet-pretrained ResNet34 and EfficientNet-B0 encoders. The augmented U-Net achieved Dice = 0.887 and IoU = 0.799, within the range of previously reported OCTA-500 baselines from 2D projection maps and close to the reported IPN volumetric baseline (DSC = 0.882). The pretrained encoders did not outperform the U-Net without pretraining. To isolate the effect of augmentation from concurrent training-recipe changes, we conducted a controlled multi-seed ablation varying only the augmentation policy. Augmentation yielded a modest but consistent Dice gain of $+0.007$ over a matched no-augmentation baseline, driven almost entirely by geometric transformations; intensity and elastic components added no measurable benefit. These findings suggest that, at this data scale, task-preserving geometric invariances are more effective than natural-image pretraining for 2D large-vessel OCTA segmentation.</p> 2026-11-13T00:00:00+00:00 Copyright (c) 2026 Henrique Christoph Bohn, Cristiano André da Costa, Felipe André Zeiser, Fabio Lavinsky, Ismael Garcia Santos, Gregory Stein