Search for: All records

Dynamic treatment regimes (DTRs) are critical to precision medicine, optimizing long-term outcomes through personalized, real-time decision making in evolving clinical contexts, but require careful supervision for unsafe treatment risks. Existing efforts rely primarily on clinician prescribed gold standards despite the absence of a known optimal strategy, and predominantly using structured EHR data without extracting valuable insights from clinical notes, limiting their reliability for treatment recommendations. In this work, we introduce SAFER, a calibrated risk-aware tabular-language recommendation framework for DTR that integrates both structured EHR and clinical notes, enabling them to learn from each other, and addresses inherent label uncertainty by assuming ambiguous optimal treatment solution for deceased patients. Moreover, SAFER employs conformal prediction to provide statistical guarantees, ensuring safe treatment recommendations while filtering out uncertain predictions. Experiments on two publicly available sepsis datasets demonstrate that SAFER outperforms state-of-the-art baselines across multiple recommendation metrics and counterfactual mortality rate, while offering robust formal assurances. These findings underscore SAFER’s potential as a trustworthy and theoretically grounded solution for high-stakes DTR applications. 

Recent advances in Large Language Models (LLMs) have demonstrated significant potential in the field of Recommendation Systems (RSs). Most existing studies have focused on converting user behavior logs into textual prompts and leveraging techniques such as prompt tuning to enable LLMs for recommendation tasks. Meanwhile, research interest has recently grown in multimodal recommendation systems that integrate data from images, text, and other sources using modality fusion techniques. This introduces new challenges to the existing LLM-based recommendation paradigm which relies solely on text modality information. Moreover, although Multimodal Large Language Models (MLLMs) capable of processing multi-modal inputs have emerged, how to equip MLLMs with multi-modal recommendation capabilities remains largely unexplored. To this end, in this paper, we propose the Multimodal Large Language Model-enhanced Sequential Multimodal Recommendation (MLLM-MSR) model. To capture the dynamic user preference, we design a two-stage user preference summarization method. Specifically, we first utilize an MLLM-based item-summarizer to extract image feature given an item and convert the image into text. Then, we employ a recurrent user preference summarization generation paradigm to capture the dynamic changes in user preferences based on an LLM-based user-summarizer. Finally, to enable the MLLM for multi-modal recommendation task, we propose to fine-tune a MLLM-based recommender using Supervised Fine-Tuning (SFT) techniques. Extensive evaluations across various datasets validate the effectiveness of MLLM-MSR, showcasing its superior ability to capture and adapt to the evolving dynamics of user preferences. 

This study focused on early adolescents’ stress of language brokering and examined the moderating role of family cumulative risk in the relation of language brokering to adjustment problems. Data came from self-reports of 604 low-income Mexican American adolescent language brokers (54% female; [Formula: see text]= 12.4; SD = 0.97; 75% born in the United States) and their parents (99% foreign-born) in central Texas. Path analyses revealed that brokering stress, but not frequency, was positively associated with adolescents’ adjustment problems, including depressive symptoms, anxiety, and delinquency. We also found that the relation between stress of brokering for mothers and adolescents’ depressive symptoms was stronger among families with a high cumulative risk. Further, with a high cumulative risk, adolescents exhibited delinquent behaviors regardless of the levels of stress from translating for fathers. Current findings underscore the importance of examining family contexts in assessing the consequences of language brokering for Mexican American early adolescents’ well-being. 

Language brokering is a prevalent phenomenon in ethnic minority immigrant populations. Although accruing evidence points to the beneficial impacts of healthy role identity development, research investigating the formation of a language broker role identity in language brokering adolescents is lacking in the literature. In a sample of 604 Latinx adolescents (54.3% female; Mage at Time 1 = 12.41, SD = .97), structured equation modeling was conducted with maternal warmth and hostility examined as antecedents and adolescents’ life meaning as a mediator for language broker role identities. Results revealed that life meaning mediated the positive association from maternal warmth to language broker role identity. However, the negative association from maternal hostility to language broker role identity was no longer significant when accounting for maternal warmth. Corroborating extant findings, reciprocal relations were observed between maternal parenting practices, life meaning and language broker role identity. The results attest to the importance of investigating culturally specific role identity development in immigrant populations and demonstrates the role of maternal parenting practices in affecting adolescents’ role identity formation, albeit with contrasting gender effects.