608
P.-Y. S. Hsueh et al.
13.1 Introduction
In this chapter, we provide a snapshot of the state-of-the-art research in mobile and
IOT e-health studies that leverage AI technologies for making sense of personal
health measurement and assessment, as well as for delivering situational, actionable
insights in care flows. In recent years, the proliferation of consumer and pervasive
health technologies has enabled a whole new generation of sensor-based precision
measurement technologies and mobile ecological momentary assessments that are
able to capture patient-specific characteristics in context [1–3]. The captured physiomes (i.e., a collection of quantitative and integrated descriptions of the functional
behavior of the physiological state of an individual [4]) can help detect physiological
macro-phenotypes such as inflammatory response and fatigue [5], as well as critical
conditions such as seizure and atrial fibrillation [6, 7]. The accumulated longitudinal
records of such phonemes are also expected to capture patterns that can help
distinguish individual physiological differences, e.g., being insulin-sensitive or
insulin-resistant, which will make a difference in disease diagnosis and prognosis
[5].
Moreover, the patient-specific characteristics include not only physiome patterns
exhibited in physiological measures but also behavioral patterns and its associated
contexts in daily-life settings, which explain more than 50% of the premature deaths
in the United States for the past 50 years and become one of the major categories
of health determinants [8]. Take cardiovascular health for an example: Public
health experts have advocated on the importance of healthy behaviors for reducing
cardiovascular risk, which estimatedly sums up to 41% of global cardiovascular
disease burden [9, 10]. Past research has also demonstrated the benefit of healthy
behavior (e.g., regular physical activity) on emotional well-being [11, 12].
Being able to capture behavioral patterns in context would have tremendous
implications to healthcare applications. However, despite the success of capturing
the varying patient-specific characteristics, early evidence of using them to improve
outcomes yields only mixed results [3, 13–15].
On the one hand, simply providing technologies for self-monitoring is not
sufficient to activate and sustain behavioral changes. Individuals in different stages
of change would need different ways to communicate and different types of
intervention. This, in turn, resulted in the failure of many implementation plans
for integrating mobile and IOT e-health data to guide behavioral change that often
require more intuitive reasoning and explanation capability [16].
On the other hand, a number of theoretical model-driven interventions have
been shown to have positive impact on activating and sustaining health behavioral
changes. For example, implementation intentions (i.e., representations of simple
plans that can translate goal intentions into behavior under specific conditions)
have a medium to large effect on goal attainment of physical activity behavior [17],
especially when coupled with e-health capabilities [18–22].
As a middle ground, behavioral scientists and healthcare service researchers are
now working together to employ e-health technologies as an interactive platform
P.-Y. S. Hsueh et al.
13.1 Introduction
In this chapter, we provide a snapshot of the state-of-the-art research in mobile and
IOT e-health studies that leverage AI technologies for making sense of personal
health measurement and assessment, as well as for delivering situational, actionable
insights in care flows. In recent years, the proliferation of consumer and pervasive
health technologies has enabled a whole new generation of sensor-based precision
measurement technologies and mobile ecological momentary assessments that are
able to capture patient-specific characteristics in context [1–3]. The captured physiomes (i.e., a collection of quantitative and integrated descriptions of the functional
behavior of the physiological state of an individual [4]) can help detect physiological
macro-phenotypes such as inflammatory response and fatigue [5], as well as critical
conditions such as seizure and atrial fibrillation [6, 7]. The accumulated longitudinal
records of such phonemes are also expected to capture patterns that can help
distinguish individual physiological differences, e.g., being insulin-sensitive or
insulin-resistant, which will make a difference in disease diagnosis and prognosis
[5].
Moreover, the patient-specific characteristics include not only physiome patterns
exhibited in physiological measures but also behavioral patterns and its associated
contexts in daily-life settings, which explain more than 50% of the premature deaths
in the United States for the past 50 years and become one of the major categories
of health determinants [8]. Take cardiovascular health for an example: Public
health experts have advocated on the importance of healthy behaviors for reducing
cardiovascular risk, which estimatedly sums up to 41% of global cardiovascular
disease burden [9, 10]. Past research has also demonstrated the benefit of healthy
behavior (e.g., regular physical activity) on emotional well-being [11, 12].
Being able to capture behavioral patterns in context would have tremendous
implications to healthcare applications. However, despite the success of capturing
the varying patient-specific characteristics, early evidence of using them to improve
outcomes yields only mixed results [3, 13–15].
On the one hand, simply providing technologies for self-monitoring is not
sufficient to activate and sustain behavioral changes. Individuals in different stages
of change would need different ways to communicate and different types of
intervention. This, in turn, resulted in the failure of many implementation plans
for integrating mobile and IOT e-health data to guide behavioral change that often
require more intuitive reasoning and explanation capability [16].
On the other hand, a number of theoretical model-driven interventions have
been shown to have positive impact on activating and sustaining health behavioral
changes. For example, implementation intentions (i.e., representations of simple
plans that can translate goal intentions into behavior under specific conditions)
have a medium to large effect on goal attainment of physical activity behavior [17],
especially when coupled with e-health capabilities [18–22].
As a middle ground, behavioral scientists and healthcare service researchers are
now working together to employ e-health technologies as an interactive platform