RESUMO
Machine learning (ML) has seen impressive growth in health science research due to its capacity for handling complex data to perform a range of tasks, including unsupervised learning, supervised learning, and reinforcement learning. To aid health science researchers in understanding the strengths and limitations of ML and to facilitate its integration into their studies, we present here a guideline for integrating ML into an analysis through a structured framework, covering steps from framing a research question to study design and analysis techniques for specialized data types.
Assuntos
Aprendizado de Máquina , Reforço Psicológico , Humanos , Projetos de Pesquisa , PesquisadoresRESUMO
Introduction/Purpose: Wearables that include a color light sensor are a promising measure of electronic screen use in adults. However, to extend this approach to children, we need to understand feasibility of wear placement. The purpose of this study was to examine parent perceptions of children's acceptability of different sensor placements and feasibility of free-living 3- to 7-day wear protocols. Methods: This study was conducted in three phases. In phase 1, caregivers (n=161) of 3- to 8-year-old children completed an online survey to rate aspects of fitting and likelihood of wear for seven methods (headband, eyeglasses, skin adhesive patch, shirt clip/badge, mask, necklace, and vest). In phase 2, children (n=31) were recruited to wear one of the top five prototypes for three days (n=6 per method). In phase 3, children (n=23) were recruited to wear prototypes of the top three prototypes from phase 2 (n=8 per method) for 7 days. In phases 2 and 3, parents completed wear logs and surveys about their experiences. Parents scored each wearable on three domains (ease of use, likelihood of wear, and child enjoyment). Scores were averaged to compute an everyday "usability" score (0, worst, to 200, best). Results: Phase 1 results suggested that the headband, eyeglasses, patch, clip/badge, and vest had the best potential for long-term wear. In phase 2, time spent wearing prototypes and usability scores were highest for the eyeglasses (10.4 hours/day, score=155.4), clip/badge (9.8 hours/day, score=145.8), and vest (7.1 hours/day, score=141.7). In phase 3, wearing time and usability scores were higher for the clip/badge (9.4 hours/day, score=169.6) and eyeglasses (6.5 hours/day, score=145.3) compared to the vest (4.8 hours/day, score=112.5). Conclusion: Results indicate that wearable sensors clipped to a child's shirt or embedded into eyeglasses are feasible and acceptable wear methods in free-living settings. The next step is to asses the quality, validity, and reliability of data captured using these wear methods.
RESUMO
ERATS decreased length of stay, postoperative complications, and readmission.
RESUMO
In the twenty years since Dr. Leo Breiman's incendiary paper Statistical Modeling: The Two Cultures was first published, algorithmic modeling techniques have gone from controversial to commonplace in the statistical community. While the widespread adoption of these methods as part of the contemporary statistician's toolkit is a testament to Dr. Breiman's vision, the number of high-profile failures of algorithmic models suggests that Dr. Breiman's final remark that "the emphasis needs to be on the problem and the data" has been less widely heeded. In the spirit of Dr. Breiman, we detail an emerging research community in statistics - data-driven decision support. We assert that to realize the full potential of decision support, broadly and in the context of precision health, will require a culture of social awareness and accountability, in addition to ongoing attention towards complex technical challenges.