OBJECTIVE: This review aims to identify the contribution of user experience features and underlying technical features to sustained engagement in unguided chatbots for improving health-related behaviors. MATERIALS AND METHODS: Following PRISMA-2020 guidelines, we conducted a systematic review, searching PubMed, ACM, APA PsycINFO, Cochrane, Web of Science, and IEEE Xplore from June to September 2022 and updated in April 2025. Data was analyzed via Synthesis without Meta-Analysis (SWiM), to understand the relationship between user engagement overall and individual experience metrics. RESULTS: Customizable avatars and flexible input interactions may enhance overall user engagement. Conversely, pre-scripted content that lacks personalization and emotional support negatively impacts user satisfaction and adherence to health interventions. Other features contributing to sustained engagement are in-app technical assistance, user learning features, and crisis support systems. A strong positive correlation (r = 0.808, n = 16) was observed between user satisfaction and engagement, specifically for satisfaction dimensions including need fulfillment (r = 0.872, n = 6), willingness to recommend chatbot (r = 0.817, n = 4) and user enjoyment (r = 0.971, n = 3) in SWiM analysis. The limited application of large language models and retrieval augmented generation techniques may constrain the quality of support available to users and overall sustained engagement. CONCLUSION: Effective unguided chatbot design requires an emphasis on interactive educational elements, in-app technical assistance and crisis support, and personalized content. This can be achieved with high context awareness, input understanding, and quality content generation. Our findings suggest that user satisfaction is a primary driver of sustained engagement, though further research is needed to validate individual user satisfaction features for sustained engagement.

Alzheimer's disease causes neural damage, including brain atrophy in the patient. Consequently, ventricles that contain cerebral fluid a re e xpanded to filling th oseregions, which increases the proportional volume of ventricles in the brain. Therefore, abnormal growth of ventricle volume is an important indicator for estimating neural damage and, in turn, for the progression of Alzheimer's diseases. The rate of this volumegrowth, i.e., neural damage, can be predicted by predictive and machine learning models using the patient's current status. These predictions help assess the effectiveness of a particular treatment for a patient, in addition to providing some expectation of the disease timeline. In this work, we propose a convolutional neural network (CNN) model using region-level features for predicting ventricle volume biomarkers. The region-level representation with domaindriven features benefits from the CNN spatial pattern recognition capability. It also prevents learning irrelevant features and overfitting tot he t raining d ata a s a r esult 0 fd ata scarcity. Our model is applied to the ADNI dataset in the TADPOLE competition and outperforms the best leaderboard results.

Over 1.1 million tweets were collected using the four hashtags (# health, # fitness, # diet, # weightloss) from August 2018 to January 2019 for the purpose of understanding obesity-related image sharing practice in Twitter. Of these tweets, more than 200,000 (17.76%) had images, which were shared by more than 60,000 unique twitter users (22.13%). Over 130,000 unique images were computationally processed by applying a very deep VGG16 model (convolutional neural network) and 967 unique labels were used to describe objects within the images extracted from obesity related tweets. Informational content was most frequently re-shared, followed by images of individuals posing, comical content, commercial content, physical activities, and motivational quotes. Although a general deep learning model provided high level insights into the image sharing activity on Twitter, manual analysis revealed that a tailored model is needed to learn detailed information.

This study aims to demonstrate the feasibility of using a novel distance-based representation of 3D CT-scan images to train a deep learning model for dose predictions in radiation treatment planning. The distance representation is inspired by previous knowledge of the domain to increase the generalizability of the deep learning models for radiation treatment planning. Conventional knowledge-based planning methods extract engineered features from 3D CT-scan images, as well as other patients' features, to predict the best achievable dose in a cancerous area and other organs at risk. Recent studies have shown higher accuracy in voxel-level dose prediction using deep learning models compared to the conventional machine learning approaches. Since the data resources for training these models are limited, most of the studies use 2D contour information to represent the patient anatomy. This representation loses volumetric information, and it is sensitive to small changes in patient orientation and translation. The distance-based representation introduced in this paper is inspired by the domain knowledge and is able to maintain the volumetric distance information despite the 2D slicing of 3D CT-image. According to prior studies in the radiation treatment planning domain, there is a strong association between the organs-at-risk distance from the cancerous volume and the patient's vulnerability to receive excessive dose. Therefore, the contour value in prior representation was replaced by voxel distance from cancerous volume. This modification in representation makes it transition and orientation invariant and adds potential robustness to patient positioning differences during the imaging/planning process. We evaluated the distance-based deep learning models through experiments for prediction of prostate cancer patients' vulnerability and voxel-level dose distribution using convolutional neural network and U-net models, respectively. The results were compared with contour-based U-net model as well as conventional machine learning with engineered representations. We found that the performance was comparable or higher than the prior state-of-the-art results for prostate-cancer dose distribution prediction.

It is important for data scientists to have a good understanding of the availability of relevant datasets as well as the content, structure, and existing analyses of these datasets. While a number of efforts are underway to integrate the large amount and variety of datasets, there is a lack of information resources that focus on specific learning needs of some targeted audiences. To address this gap, we have been developing a semantic Dataset Information Resource (DIR) framework to specifically address the challenges of entry-level data scientists in learning to identify, understand, and analyze major datasets with an initial focus on healthcare. The DIR does not contain actual data from the datasets but aims to provide comprehensive knowledge about the datasets and their analyses. The framework leverages Semantic Web technologies and the W3C Dataset Description Standard for knowledge integration and representation and includes natural language processing (NLP)-based methods to enable knowledge extraction and question answering. The prototype DIR implementation includes four major components-dataset metadata and related knowledge, search modules, question answering for frequently-asked questions, and blogs. And the DIR currently includes information on three commonly-used large and complex healthcare datasets: HCUP, MarketScan, and MIMIC. Initial usage evaluation based on health informatics students is encouraging. Further development is underway.

We describe the architecture of the Patient Centered Outcomes Research Institute (PCORI) funded Scalable Collaborative Infrastructure for a Learning Healthcare System (SCILHS, http://www.SCILHS.org) clinical data research network, which leverages the $48 billion dollar federal investment in health information technology (IT) to enable a queryable semantic data model across 10 health systems covering more than 8 million patients, plugging universally into the point of care, generating evidence and discovery, and thereby enabling clinician and patient participation in research during the patient encounter. Central to the success of SCILHS is development of innovative 'apps' to improve PCOR research methods and capacitate point of care functions such as consent, enrollment, randomization, and outreach for patient-reported outcomes. SCILHS adapts and extends an existing national research network formed on an advanced IT infrastructure built with open source, free, modular components.

BACKGROUND: Current image sharing is carried out by manual transportation of CDs by patients or organization-coordinated sharing networks. The former places a significant burden on patients and providers. The latter faces challenges to patient privacy. OBJECTIVE: To allow healthcare providers efficient access to medical imaging data acquired at other unaffiliated healthcare facilities while ensuring strong protection of patient privacy and minimizing burden on patients, providers, and the information technology infrastructure. METHODS: An image sharing framework is described that involves patients as an integral part of, and with full control of, the image sharing process. Central to this framework is the Patient Controlled Access-key REgistry (PCARE) which manages the access keys issued by image source facilities. When digitally signed by patients, the access keys are used by any requesting facility to retrieve the associated imaging data from the source facility. A centralized patient portal, called a PCARE patient control portal, allows patients to manage all the access keys in PCARE. RESULTS: A prototype of the PCARE framework has been developed by extending open-source technology. The results for feasibility, performance, and user assessments are encouraging and demonstrate the benefits of patient-controlled image sharing. DISCUSSION: The PCARE framework is effective in many important clinical cases of image sharing and can be used to integrate organization-coordinated sharing networks. The same framework can also be used to realize a longitudinal virtual electronic health record. CONCLUSION: The PCARE framework allows prior imaging data to be shared among unaffiliated healthcare facilities while protecting patient privacy with minimal burden on patients, providers, and infrastructure. A prototype has been implemented to demonstrate the feasibility and benefits of this approach.

Multiscale analysis is often required in image processing applications because image features are optimally detected at different levels of resolution. With the advance of high-resolution 3D imaging, the extension of multiscale analysis to higher dimensions is necessary. This paper extends an existing 2D scale selection method, known as the minimum reliable scale, to 3D volumetric images. The method is applied to 3D boundary detection and is illustrated in examples from biomedical imaging. The experimental results show that the 3D scale selection improves the detection of edges over single scale operators using as few as three different scales.

Detection of the set of centers of maximal disks is a key step toward generation of accurate skeletons on the basis of distance maps. Algorithms using approximate distance metrics are abundant and their theory has been well established. However, the resulting skeletons may be inaccurate and sensitive to rotation. In this paper, we study methods for detecting maximal disks from distance maps based on the exact Euclidean metric. We first show that no previous algorithm identifies the exact set of discrete maximal disks under Euclidean distance metric. We then propose new algorithms and show that they produce the exact set of maximal disks. The effectiveness of our algorithms is demonstrated with numerous examples.

The skeleton is an important representation for shape analysis. A common approach for generating discrete skeletons takes three steps: 1) computing the distance map, 2) detecting maximal disks from the distance map, and 3) linking the centers of maximal disks (CMDs) into a connected skeleton. Algorithms using approximate distance metrics are abundant and their theory has been well established. However, the resulting skeletons may be inaccurate and sensitive to rotation. In this paper, we study methods for generating skeletons based on the exact Euclidean metric. We first show that no previous algorithms identify the exact set of discrete maximal disks under the Euclidean metric. We then propose new algorithms and show that they are correct. To link CMDs into connected skeletons, we examine two prevalent approaches: connected thinning and steepest ascent. We point out that the connected thinning approach does not work properly for Euclidean distance maps. Only the steepest ascent algorithm produces skeletons that are truly medially placed. The resulting skeletons have all the desirable properties: they have the same simple connectivity as the figure, they are well-centered, they are insensitive to rotation, and they allow exact reconstruction. The effectiveness of our algorithms is demonstrated with numerous examples.

Analysis of brain images often requires accurate localization of cortical convolutions. Although magnetic resonance (MR) brain images offer sufficient resolution for identifying convolutions in theory, the nature of tomographic imaging prevents clear definition of convolutions in individual slices. Existing methods for solving this problem rely on heuristic adaptation of brain atlases created from a small number of individuals. These methods do not usually provide high accuracy because of large biological variations among individuals. The authors propose to localize convolutions by linking realistic visualizations of the cortical surface with the original image volume. They have developed a system so that a user can quickly localize key convolutions in several visualizations of an entire brain surface. Because of the links between the visualizations and the original volume, these convolutions are simultaneously localized in the original image slices. In the process of the authors' development, they have implemented a fast and easy method for visualizing cortical surfaces in MR images, thereby making their scheme usable in practical applications.