Why we need a small data paradigm

Abstract

Background There is great interest in and excitement about the concept of personalized or precision medicine and, in particular, advancing this vision via various ‘big data’ efforts. While these methods are necessary, they are insufficient to achieve the full personalized medicine promise. A rigorous, complementary ‘small data’ paradigm that can function both autonomously from and in collaboration with big data is also needed. By ‘small data’ we build on Estrin’s formulation and refer to the rigorous use of data by and for a specific N-of-1 unit (i.e., a single person, clinic, hospital, healthcare system, community, city, etc.) to facilitate improved individual-level description, prediction and, ultimately, control for that specific unit. Main body The purpose of this piece is to articulate why a small data paradigm is needed and is valuable in itself, and to provide initial directions for future work that can advance study designs and data analytic techniques for a small data approach to precision health. Scientifically, the central value of a small data approach is that it can uniquely manage complex, dynamic, multi-causal, idiosyncratically manifesting phenomena, such as chronic diseases, in comparison to big data. Beyond this, a small data approach better aligns the goals of science and practice, which can result in more rapid agile learning with less data. There is also, feasibly, a unique pathway towards transportable knowledge from a small data approach, which is complementary to a big data approach. Future work should (1) further refine appropriate methods for a small data approach; (2) advance strategies for better integrating a small data approach into real-world practices; and (3) advance ways of actively integrating the strengths and limitations from both small and big data approaches into a unified scientific knowledge base that is linked via a robust science of causality. Conclusion Small data is valuable in its own right. That said, small and big data paradigms can and should be combined via a foundational science of causality. With these approaches combined, the vision of precision health can be achieved.

Applying the Deep Learning Method for Simulating Outcomes of Educational Interventions

“…In general, researchers in the field have considered that the deep learning methods are suitable for analyzing large datasets and the majority of studies have been conducted with bigdata [48]. Even if this might be the case in general, several researchers, for example, those who are interested in applying the deep learning methods in medical and clinical science, have been trying to test whether the methods can contribute to improving automatized diagnosis procedures with relatively small datasets [49]. In their studies, they have found that even with small datasets, the deep learning methods can more accurately predict clinical outcomes compared with traditional analysis methods and they can be well used in the aforementioned context dealing with small datasets [50][51][52].…”

Bayesian Structural Time Series for Biomedical Sensor Data: A Flexible Modeling Framework for Evaluating Interventions

“…These data demonstrate properties that are commonly found across other wearable sensor data, which are increasingly gaining in popularity (3). In order to effectively integrate such data with personalized medicine and health in the future, we must understand how specific timings and types of interventions impact individual patients (2)(3)(4).…”

“…The wearable sensor market was valued at $10.8 billion USD in 2019 and is expected to triple in value over the next five years (1). Investment is bolstered by the great potential for advancing personalized medicine in the near future (2)(3)(4). Whereas medicine has previously focused on determining the right interventions, it is now more focused on for whom and when (5).…”

“…A variety of such data can be seen in Figure 1, all of which come from various contexts. It is important to discern these properties and to establish a flexible model for emerging sensor and wearable data, as it will have broad implications in fields such as personalized medicine (2). Therefore, in this study we aim to contextualize a model and establish a flexible statistical framework to model various sensor and mobile health data.…”

Environmental Exposures during Puberty: Window of Breast Cancer Risk and Epigenetic Damage

“…Currently, there is a significant movement underway to develop interactive research tools that can support public health advocacy and environmental justice. Examples include (1) the United States Department of Agriculture (USDA) food desert locator [98], (2) Environmental Protection Agency environmental justice tools [51], and (3) some of the dynamic neighborhood/environmental mapping work performed at the University of California, San Diego [99][100][101].…”