Journal of Medical Internet Research最新文献

Background: Osteoporosis (OP) is projected to be a major issue significantly impacting the well-being of middle-aged and old populations. Machine learning (ML) and deep learning (DL) models developed based on medical imaging have enhanced clinicians' diagnostic accuracy and work efficiency. However, the diagnostic performance of different types of medical imaging for OP has not been systematically assessed.

Objective: By summarizing related literature, this study aims to elucidate the role of DL models based on different medical imaging modalities in OP detection.

Methods: PubMed, Embase, the Cochrane Library, and Web of Science were systematically searched for studies using ML for the diagnosis of OP based on medical imaging. The final search was conducted on May 16, 2024. The risk of bias in the included studies was assessed using the Quality Assessment of Diagnostic Accuracy Studies-2 tool. A bivariate mixed-effects model was applied to perform meta-analyses of sensitivity (SEN) and specificity (SPC), stratified by imaging modality (x-ray, computed tomography [CT], magnetic resonance imaging [MRI]). In addition, subgroup analyses were carried out based on the type of ML algorithm, the method of validation dataset generation, and the anatomical site of assessment.

Results: A total of 60 studies comprising 66,195 participants were encompassed in this systematic review and meta-analysis. Among these, 22 studies used x-ray imaging, 37 applied CT imaging, and 3 used MRI for ML-based OP diagnosis. For x-ray-based models, the pooled SEN and SPC for studies focusing on the appendicular skeleton were 0.97 (95% CI 0.83-0.99) and 0.90 (95% CI 0.75-0.96), respectively. For studies using the mandible as the target site, SEN and SPC were 0.94 (95% CI 0.89-0.97) and 0.80 (95% CI 0.56-0.93), respectively. For those focusing on the lumbar spine, the pooled SEN and SPC were 0.87 (95% CI 0.77-0.93) and 0.82 (95% CI 0.75-0.87), respectively. For CT-based models, studies targeting the hip joint reported a pooled SEN and SPC of 0.87 (95% CI 0.83-0.90) and 0.92 (95% CI 0.81-0.96), respectively. For the thoracic spine, SEN and SPC were 0.91 (95% CI 0.86-0.94) and 0.94 (95% CI 0.92-0.95), respectively, while for the lumbar spine, they were 0.91 (95% CI 0.87-0.94) and 0.92 (95% CI 0.86-0.95), respectively.

Conclusions: ML based on medical imaging demonstrates high diagnosis accuracy for OP, particularly DL models using x-ray and CT modalities. However, this study included only a limited number of original studies using MRI-based ML, and there remains a lack of adequate external validation across studies, which poses interpretative limitations. Future research should aim to develop artificial intelligence tools with broader applicability and enhanced diagnostic precision.

Background: The advent of tirzepatide has transformed obesity care; yet, real-world weight loss outcomes necessarily depend on patient engagement with behavioral support. Digital platforms offering coaching, self-monitoring, and automated feedback have the potential to further augment pharmacological efficacy.

Objective: The aim of the study is to examine associations between digital engagement and weight loss outcomes among adults prescribed tirzepatide in routine care over 12 months and to identify baseline correlates of engagement.

Methods: In this retrospective cohort study, we included adults (18-75 years; BMI ≥30 or ≥27.5 kg/m² with comorbidities) who initiated tirzepatide between February 2024 and August 2025 via a UK digital weight loss service. Engagement was defined by all 3: attendance at ≥1 coaching session AND ≥1 weekly weight log AND ≥1 app login over 12 months. Percent weight loss was analyzed at months 2, 4, 6, 8, 10, and 12 using a mixed model repeated measures adjusted for age, sex, baseline BMI, and comorbidities. Time-to-event analyses (Kaplan-Meier) assessed attainment of ≥5%, ≥10%, ≥15%, and ≥20% weight loss thresholds. Multivariable logistic regression identified correlates of engagement, reporting odds ratios (ORs) per decade of age and per 5 kg/m² BMI.

Results: Among 126,553 participants, 6746 (5.3%) were maximally engaged. Cohort demographics were a mean age of 42.3 (SD 12.4) years, 78.9% (99,905/126,553) female, and a mean BMI of 35.3 (SD 6.2) kg/m². Engaged users achieved greater adjusted weight loss at month 12 (-22.9%, 95% CI -23.2 to -22.6) versus nonengaged users (-17.5%, 95% CI -17.7 to -17.4), an absolute difference of 5.3 percentage points (P<.001; Cohen d=0.54). Differences emerged by month 2 (-7.4% vs -6.4%; P<.001) and widened steadily. Engaged participants reached all clinically significant weight loss thresholds faster (5%-20%; log-rank P<.001), and engaged participants were nearly 3 times more likely to achieve ≥20% weight loss compared to nonengaged participants (1079/6746, 16% vs 6710/119,807, 5.6%; risk ratio 2.88; P<.001). Older age (OR 1.18 per decade, 95% CI 1.15-1.20; P<.001), higher BMI (OR 1.14 per 5 kg/m², 95% CI 1.12-1.16; P<.001), and the presence of polycystic ovary syndrome (OR 1.59, 95% CI 1.45-1.74; P<.001) or fatty liver disease (OR 1.52, 95% CI 1.32-1.76; P<.001) correlated with engagement. Male sex (OR 0.86, 95% CI 0.81-0.92; P<.001) and diabetes (OR 0.83, 95% CI 0.73-0.95; P=.009) were associated with lower engagement.

Conclusions: Digital engagement was associated with substantially greater tirzepatide-associated weight loss in real-world practice. Integrating structured digital support with pharmacotherapy represents a promising strategy for optimizing obesity management.

Large language models are rapidly transitioning from pilot schemes to routine clinical practice. This creates an urgent need for clinicians to develop the necessary skills to strike the right balance between seizing opportunities and taking accountability. We propose a 3-tier competency framework to support clinicians' evolution from cautious users to responsible stewards of artificial intelligence (AI). Tier 1 (foundational skills) defines the minimum competencies for safe use, including prompt engineering, human-AI agent interaction, security and privacy awareness, and the clinician-patient interface (transparency and consent). Tier 2 (intermediate skills) emphasizes evaluative expertise, including bias detection and mitigation, interpretation of explainability outputs, and the effective clinical integration of AI-generated workflows. Tier 3 (advanced skills) establishes leadership capabilities, mandating competencies in ethical governance (delineating accountability and liability boundaries), regulatory strategy, and model life cycle management-specifically, the ability to govern algorithmic adaptation and change protocols. Integrating this framework into continuing medical education programs and role-specific job descriptions could enhance clinicians' ability to use AI safely and responsibly. This could standardize deployment and support safer clinical practice, with the potential to improve patient outcomes.