Journal of Medical Systems最新文献

Good mental health is crucial for well-being. Unfortunately, despite the advancements of automatic detection solutions in the mental health field, along with the existence of effective treatments, a large percentage of affected people receive no care for their disorder. Thus, this research proposes an innovative framework integrating counterfactual explanations into a multi-label detection system for anxiety and depression, combining large language models for feature extraction and multi-label machine learning for final prediction. The solution is designed to operate in a streaming mode, addressing the need to process information in real-time. Moreover, sliding window techniques manage the data's evolution, preserving temporal relevance while ensuring robust, user-centered interpretation capabilities. The latter is reinforced by the generation of counterfactual explanations, which contribute to the interpretability, adaptability, and accountability of the results in a critical context, such as mental health. The results surpass the 90% accuracy, indicating very few misclassifications per label. Ultimately, this solution contributes to the literature with timely and transparent decision-making in mental healthcare.

Acute Myeloid Leukemia (AML) is a complex and heterogeneous disease identified by severe clinical progression, fast cellular proliferation, and often high mortality rates. Incorporating diverse longitudinal information on patients' medical histories is essential for developing effective disease predictive models applicable to both research and clinical settings. Here, we present a robust methodology for discovering the regulation of disease progression dynamics from a novel longitudinal, multimodal clinical dataset of patients diagnosed with AML. The medical data were analyzed to reveal the main clinical, genetic, and treatment features modulating disease progression. To discover dynamic mathematical models at the systems level-including the necessary regulatory interactions, parameters, and disease drivers-predictive of AML progression, we developed a de novo inference algorithm based on high-performance evolutionary computation. The results demonstrate that the predictive methodology could accurately estimate the drivers and clinical dynamics of AML progression in terms of blast percentages for both training and novel patients. This approach effectively predicted AML drivers, their mechanistic interactions, and disease progression by leveraging the heterogeneous and longitudinal dynamics of patients' clinical data. Importantly, this methodology shows significant potential for modeling progression dynamics in other acute diseases, providing a flexible and adaptable framework for advancing clinical and translational research.

Meaningful community engagement is an essential component of impactful public health and implementation research. Multiple community engagement methods have been defined, including co-creation. Co-creation involves an iterative process that advances from identifying opportunities for value creation and solutions, to defining partner priorities, to evaluating co-created outcomes. This study reports our methods to co-create culturally and linguistically meaningful mHealth messages to promote preventive healthcare engagement for Arabic, Spanish, and Vietnamese - speaking communities. This multi-method study is part of a larger program of research, "Working towards Empowered community-driven Approaches to increase Vaccination and preventive care Engagement" (WEAVE), that aims to co-create and test a preventive healthcare program that includes mHealth and care coordination with medically underserved patients at multiple federally qualified health center (FQHC) locations near the US/Mexico border and surrounding region. A multi-level partner process was used to engage in co-creation across six partner groups (n = 27): (1) Community Advisory Boards (CAB), (2) Community Weavers (individuals with lived experience as members of an underserved community who act as cultural brokers between communities, public health systems, and researchers), (3) FQHC Care Coordinators, (4) FQHC Administrators, (5) a FQHC Clinical Expert, and (6) Research Experts in health communication, vaccine behavior research, and/or mHealth. Each of these partner groups was distinctly engaged through structured CAB meetings, weekly research and operations team meetings, topic-specific meetings, and e-review of content. The Research Engagement Survey Tool (REST) was used as a global assessment of partner engagement in the co-creation process. Results are organized by a co-creation framework anchored to identify, analyze, define, and design steps. Across four CAB meetings and engagement activities with the other co-creation partners, 262 mHealth messages (89 Arabic, 85 Spanish, 88 Vietnamese) were refined and approved. A message cadence and delivery mode were finalized. On the REST, the average ratings were over 4.50 (out of 5), indicating strong perceptions of engagement with the co-creation process and members. We successfully engaged six co-creation partner groups to develop and approve the content, cadence, and delivery mode of mHealth preventive care messages. These messages will be embedded in the multicomponent health program that will be tested in a randomized adaptive trial. NCT05841810, registration date: 03/28/2023.

Objective: To develop a tool for the clinical hybrid imaging workflow which combines morphologic and functional measurements. And to quantify the number of clicks saved per positron emission tomography/computed tomography (PET/CT) interpretation.

Methods: A tool was developed where a volume of interest (VOI) is automatically created around line distance measurements. VOI statistics for both PET and CT component, and line distances are generated and displayed. Usage data for the first two months after introduction of the tool was analyzed.

Results: Eleven radiologists and nuclear medicine physicians used the tool in 364 PET/CTs. In 19% of examinations, the novel tool was the only tool that needed to be used. The novel combined tool was used 1001 times, whereas the traditional spherical VOI had been placed 1131 times. The usage ratio of new to traditional tool differed significantly between examinations with ≤ 6 annotations (ratio 1.0) versus > 6 annotations (ratio 0.63, p = 0.030). The average number of saved clicks per PET/CT was estimated at 16.5.

Conclusion: A novel combined measurement tool for hybrid imaging was implemented and saved on average 16.5 clicks per examination. These improvements contribute to a smoother workflow and demonstrate the positive impact of thoughtful software design in medical practice.

While Fine needle aspiration cytology (FNAC) and mammography are both used to diagnose breast lesions, FNAC is generally more accurate than mammograms for predicting breast cancer. It is also gaining popularity as an early detection tool due to its rapid and straightforward procedure, cost-effectiveness, and minimal risk of complications. Deep learning enhances breast cancer detection by extracting crucial features, yielding highly accurate results compared to conventional techniques. Classical machine learning is less time-intensive and requires fewer parameter adjustments. This work is presented as a proof-of-concept study on FNAC images obtained from two centers. It explores eighteen hybrid architectures that are developed and evaluated, combining the strength of deep learning techniques- Inception-V3, MobileNet-V2, and DenseNet-121 for feature extraction, with three machine learning classifiers (Support Vector Machine, Decision Tree, and k-Nearest Neighbours) for binary classification of fine needle aspiration cytology images of the breast. Our study is based on an indigenously collected dataset of 427 images (152 benign and 275 malignant), which was later expanded through augmentation to 2,866 images (1216 benign and 1,650 malignant). The hybrid model, which combines feature extraction from MobileNet-V2 and DenseNet-121 in a concatenated architecture, achieves the highest internal test accuracy of 98.26% when paired with an SVM classifier. It also achieves the best-known sensitivity (97.95%) and specificity (98.48%). The explainability model, which utilizes Grad-CAM, achieved 95% positive clinical validation by expert pathologists, underscoring the model's trustworthiness and interpretability-critical for clinical adoption and decision-making support. The proposed hybrid model, with its impressive metrics and validation rate, underscores the model's ability to provide clear, interpretable insights that support clinical decision-making.

Large language models (LLMs) such as ChatGPT have gained attention for their potential to assist clinical decision-making in oncology. However, real-world validation of these models against multidisciplinary tumor board (MTB) recommendations-particularly in breast cancer treatment-remains limited. This retrospective study assessed the concordance between GPT-4o and the decisions of a breast cancer MTB over a six-month period. Thirty-three patients were included. Structured clinical data were entered into GPT-4o using standardized prompts, and treatment plans were generated in two independent sessions per case. Seven therapeutic domains were evaluated: surgery, radiotherapy, hormonal therapy, neoadjuvant therapy, adjuvant therapy, genetic counseling/testing, and dual HER2-targeted therapy. Two blinded reviewers scored concordance using a 5-point Likert scale. Inter-rater reliability and classification metrics were calculated. GPT-4o generated consistent recommendations across both sessions for all patients. Full concordance (5/5) with MTB decisions was observed in 31 of 33 cases (93.9%), while partial concordance (4/5) occurred in 2 cases (6.1%) due to differences regarding genetic counseling. Inter-rater agreement was perfect (Cohen's kappa = 1.00), and the mean concordance score was 4.94 out of 5. The model achieved an overall accuracy of 93.9%, precision of 93.9%, recall of 100%, and F1 score of 96.8%. GPT-4o demonstrated a high level of agreement with expert multidisciplinary decisions in breast cancer care when provided with structured clinical input. These findings support its potential as a reproducible, guideline-consistent decision-support tool in oncology workflows.

Nowadays, the presence of health-related content on social networks is rapidly increasing. With the effect of these networks, a large number of medical images, diagnosed and interpreted by various experts, are shared online. Therefore, concept detection and image classification from medical images remains a challenging task. In recent years, deep learning-based models have become increasingly popular for addressing these challenges. The primary objective of this study is to perform multi-label classification of radiological images shared on a social network by automatically assigning relevant medical concepts. These concepts are derived from the Unified Medical Language System (UMLS). In this study, Convolutional Neural Network (CNN) combined with feed forward neural networks and various image encoders, including VGG-19, DenseNet-121, ResNet-101, Xception, Efficient-B7, to predict the appropriate concepts. The proposed hybrid deep learning models were trained and evaluated using the ImageCLEF 2019 dataset. Further evaluation was performed using a custom dataset (Rdpd_Test_Ds) composed of radiological images and their associated comments collected from a social network. The performance of the models was assessed using precision, recall, and F1-score metrics. The evaluation results are promising, demonstrating high performance. To the best of our knowledge, this research is the first to apply deep learning-based models to radiological data collected from a social network, representing a novel and impactful contribution to the field.