Trends in Food Science & Technology 91 (2019) 634–639

Contents lists available at ScienceDirect

Trends in Food Science & Technology

journal homepage: www.elsevier.com/locate/tifs

Commentary

eNutrition - The next dimension for eHealth? T

a,⁎ a,b
Mike Boland , John Bronlund
a
Riddet Institute, Massey University, Palmerston North, 4442, New Zealand
b
School of Food and Advanced Technology, Massey University, Palmerston North, 4442, New Zealand

A B S T R A C T

Rapid communication technology, inexpensive sensing devices and big data enable “eNutrition” – the use of ICT to develop and support personalized healthy
nutrition. eNutrition combines personal health records, monitoring of body functions and activity by smart wearables, monitoring food consumption, and datasets of
food composition. An AI system, a personal virtual dietitian, can use this information to provide nutritional advice and alerts at the time of purchase and consumption
of foods. eNutrition has the potential to significantly lessen the public heath burden of non-communicable diseases such as hypertension, type 2 diabetes and allergic
reactions to foods.

1. Introduction
The presence of digital technologies in almost all aspects of life
together with rapid worldwide electronic communications, often re-
ferred to as the 4th Industrial Revolution, or I4, is in the process of
changing the world and how we live. eNutrition is a term for the
emergence of digital technologies that can guide our nutrition and lead
to better health at both personal and (by implication) national and
global health and lifestyle.
2. Emergence of eHealth and mHealth
Today's Information and Communications Technology (ICT) gives
the possibility of widespread management of health through the in-
ternet and smart devices. “eHealth” is defined by the World Health
Organization as “the use of information and communication technologies
(ICT) for health”. The scope of eHealth was acknowledged by the World
Health Assembly in resolution WHA58.28 (2005): “eHealth is the cost-
effective and secure use of ICT in support of health and health-related fields,
including health-care services, health surveillance, health literature, and
health education, knowledge and research.”
mHealth is a component of eHealth. The WHO Global Observatory
for eHealth (GOe) has defined mHealth or mobile health as “medical and
public health practice supported by mobile devices, such as mobile phones,
patient monitoring devices, personal digital assistants (PDAs), and other
wireless devices”. mHealth usually involves the use of a mobile phone's
communication functionality by voice, short messaging service (SMS)
and internet connectivity as well as more diverse applications including
the use of global positioning system (GPS), and Bluetooth technology.
This can be further enhanced by interaction with large cloud-based data
systems, including eHealth repositories, and by the use of sensing de-
vices that use either the native capability of a mobile phone (such as an
accelerometer or built-in camera) or by a range of connected sensing
devices that currently include smart watches and fitness bands, but also
blood glucose monitors, blood pressure monitors, electrocardiographs
and even pocket-sized cardiac ultrasound units.
Online health records are an essential part of an eHealth system,
primarily aimed at patient management and making personal health
data immediately available to a medical professional as the need arises.
Development and alignment of these systems has been the main focus of
eHealth. Another new focus has been TeleHealth: the remote mon-
itoring of patients, in some cases leading to “hospitals without beds”,
whereby patients are remotely monitored while living at home instead
of having to live in a hospital. eHealth systems can provide a platform
that will support a smart personalized nutrition system, providing di-
agnosis information to guide diet, information on allergies to avoid
foods that may contain undesired ingredients and information about
drugs to avoid adverse drug reactions (for example patients taking
statins must avoid grapefruit in the diet). This sets the stage for
eNutrition. We have deliberately used the term eNutrition in this con-
text because it adds the dimension of nutrition to the eHealth system.
We believe this is the proper use of this term.
3. Importance of nutrition and NCDs
Non-communicable diseases (NCDs) are estimated to be responsible
for 71% of all deaths (41 million people per annum) globally (World
Health Organization). These diseases include cardiovascular and re-
spiratory disease, cancers and diabetes which are collectively re-
sponsible for 80% of all premature NCD deaths. There is a strong

⁎
Corresponding author.
E-mail address: m.boland@massey.ac.nz (M. Boland).

https://doi.org/10.1016/j.tifs.2019.08.001
Received 8 April 2019; Received in revised form 24 June 2019; Accepted 2 August 2019
Available online 09 August 2019
0924-2244/ © 2019 Elsevier Ltd. All rights reserved.
M. Boland and J. Bronlund
linkage between many NCDs and diet, and the World Health
Organization is leading a campaign to decrease the risk of NCDs. The
importance of diet and the burden of disease has recently been high-
lighted in a major international study that found that in 2017, 11
million deaths and 255 million disability-adjusted life-years were at-
tributable to dietary risk factors (GBD 2017 Diet Collaborators, 2019).
In terms of deaths attributable to NCDs, the leading metabolic risk
factor is elevated blood pressure (to which 19% of global deaths are
attributed), followed by overweight and obesity and raised blood glu-
cose. All of these factors can, be managed or mitigated by diet, espe-
cially when coupled with appropriate exercise.
Properly designed nutrition has the potential to significantly reduce
the effects and occurrence of NCDs, and eNutrition could play a sig-
nificant role towards achieving this reduction.
4. Issues with current methods for nutrition management
Management of nutrition has generally been left up to the in-
dividual. Most people don't keep any record of what they have eaten
over any time, and even those with a diagnosed condition and are under
nutritional management are generally not good at keeping accurate
records (Cade, 2017). Moreover, there is no current system that can
monitor what you are about to eat and advise on portion size, what else
should be consumed, or what should be avoided.
In addition to this, advice on nutrition is not generally linked to a
measurement of recent physical activity. Such activity will affect the
caloric intake needed and can be measured by many devices available
today (even if only on a crude basis). Nor is nutritional advice usually
personalized: we are all different and have different nutritional re-
quirements, based on body parameters, lifestyle, genetic and epigenetic
makeup (nutritional phenotype), dentition and oral health, and any
history of disease or allergies. All of these factors, in principle, are
available to feed into a nutritional model and provide timely nutritional
advice.
5. Enablement through mobile communications and smartphone
apps
A critical enabler of e-Nutrition is the Smartphone. These devices
can communicate widely through a variety of protocols (typically voice,
SMS, Bluetooth and internet), connecting personal devices to the wider
world. Smartphones also have native computing ability that allows the
use of programs (apps) that can carry out a wide range of useful
functions.
There is a wide range of apps for Smartphones that can advise on
diet, some even linking to the level of physical activity (as measured by
accelerometers native to smartphones, smartwatches and fitness
bands). However, a recent review identified more than 165,000 mobile
health apps and stated: “These apps have the potential to provide low-cost,
around-the-clock access to high-quality, evidence-based health information
to end users on a global scale. However, they have not yet lived up to their
potential due to multiple barriers, including lack of regulatory oversight,
limited evidence-based literature, and concerns of privacy and security. The
future directions may consist of improving data integration into the health
care system, an interoperable app platform allowing access to electronic
health record data, cloud-based personal health record across health care
networks, and increasing app prescription by health care providers. For
consumer mobile health apps to fully contribute value to health care delivery
and chronic disease management, all stakeholders within the ecosystem must
collaborate to overcome the significant barriers.” (Kao & Liebovitz, 2017).
There are concerns about the accuracy of nutrition and diet apps. A
recent survey in Germany of three apps selected from a random sample
of 100 nutrition apps from Google Play, identified significant variances
in the caloric values ascribed to different foods, with errors of as much
as 50% (Holzmann, Proell, Hauner, & Holzapfel, 2017). It can be ex-
pected that, as the use of diet apps becomes more mainstream, they will
635
Trends in Food Science & Technology 91 (2019) 634–639
become more reliable, and that nutritionist-supported apps will emerge
that can be reliably used for diet advice and management (Chen,
Gemming, Hanning, & Allman-Farinelli, 2018). However, few of these
yet relate specifically to what food is actually eaten or should be eaten,
and fewer still are personalized to any degree.
One of the strengths of smartphones is internet connectivity. This
potentially enables immediate access to two important datasets. The
first is the person's health data: Most countries now have patients'
health information available online (eHealth systems) and this can be
used to inform diet, based on general health data, diagnosed conditions
that require dietary management (such as blood pressure, blood cho-
lesterol), any medications that require avoidance of certain foods (e.g.
statins) and known allergies. The second dataset is the composition of
foodstuffs, and again many large databases, usually government-run,
contain independently-obtained nutritional and compositional data on
most available foodstuffs, both natural and packaged (for example the
USDA Food Composition Databases: https://ndb.nal.usda.gov/ndb/,
soon to be changed to the USDA Food and Nutrient Data System). These
datasets are coordinated internationally by INFOODS, an instrument of
the United Nations Food and Agriculture Organization. These datasets
are an important resource for providing advice on what food a con-
sumer should purchase and consume.
The other important role of the smartphone is through its Bluetooth
connectivity, which allows communication with wearable and con-
nected devices such as fitness bands and smart scales.
6. Enablement through wearable and connected devices
One of the big trends in the past few years has been the emergence
of smart watches and fitness bands. Through the inbuilt accelerometer,
these devices “measure” the number of steps that have been taken
during a day and estimate the expenditure of calories. While the idea is
right, caution must be observed. It is well recognized that these devices
are prone to errors such as under-measurement of steps if worn on the
wrist, and over-measurement if the wearer uses tools or other devices
that cause vibration (such as an orbital sander, for example).
In addition to fitness bands and smart watches, there is a wide range
of wearable devices that can report on a range of activities and body
functions. Some of these are listed in Table 1.
Bathroom scales are important non-wearable but Internet of Things
(IoT) connected devices. In addition to reporting and tracking weight,
many of these devices measure body fat, using body conductance from
one foot to the other. While this form of measurement is subject to
systematic error, trends and changes over time are useful metrics.
7. Monitoring of diet and food intake
An important aspect of eNutrition is monitoring what food is actu-
ally consumed by an individual, to provide a direct measure of the
intake of energy, nutrients and bioactives. In the past, individuals on
strict diets, have kept food diaries. This has often resulted in erratic and
inaccurate records, as these are usually filled in from memory.
Moreover, there is tendency for people to underestimate portion sizes
(Cade, 2017; Hassannejad et al., 2017) and to forget between-meal
snacks. New app-supported record-keeping may improve this (Maria
Bejar, Adrian Reyes, & Dolores Garcia-Perea, 2018). Automated re-
cording, that requires no effort from the consumer would be a better
option but does not yet fully exist. New technologies can enable the
automatic monitoring of food intake. This can be done two ways: one is
by inference from use of recipes, changes in food inventory in the home
and food purchases and disposals; the other is from a range of direct
measurements including food image capture and analysis and the use of
various on-body sensing systems to monitor eating behaviors.
M. Boland and J. Bronlund Trends in Food Science & Technology 91 (2019) 634–639

Table 1
Some wearable e-health devices.
Name of device What it does Where to find out more
(web sites accessed March 2019)

Fitness band Activity trackers and smart watches that provide health and fitness https://www.fitbit.com/
monitoring:
• Heart rate
• Steps taken (km walked)

Other fitness bands and

• Sleep time
A variety of functions as above that may also include: http://www.cronovo.com/
smartwatches • Blood pressure (by either wrist cuff or photoplethysmography) https://omronhealthcare.com/products/heartguide-wearable-
• Body composition (conductance) blood-pressure-monitor-bp8000m/
• Blood glucose (prototype) https://www.tomtom.com/en_nz/sports/fitness-trackers/fitness-
• Pulse oximetry tracker-touch/black-large/
• ECG https://www.pkvitality.com/ktrack-glucose/
https://www.garmin.com.sg/garmin-technology/pulse-ox/
https://www.engadget.com/2019/01/06/withings-move-ecg/
Blood glucose monitor Continuously measures blood glucose levels and reports via local Rodbard (2016)
station or smartphone https://www.dexcom.com/continuous-glucose-monitoring
https://freestylediabetes.co.uk/our-products/freestyle-libre
http://www.medtronicdiabetes.com/products/guardian-connect-
continuous-glucose-monitoring
VitalPatch Monitors vital signs: https://vitalconnect.com/solutions/vitalpatch/
•ECG
•Heart rate
•Respiration
•Temperature
•Posture
•Activity

Polar H10
•Falls
Wearable (chest strap) heart rate sensor https://www.polar.com/au-en/products/accessories/h10_heart_
rate_sensor
QardioCore Wearable (chest strap) continuous ECG monitor. https://store.getqardio.com/products/qardiocore

7.1. Inferential approaches 7.2. Direct approaches

A smart kitchen has the potential to track whole food and ingredient
use in the home through changes in inventory, for example in a smart
refrigerator; and an assistive kitchen (such as the Miele “Invisible
Kitchen”) will be able to track food use through recipes and cooking
activities. This will give an overview of food consumption in the home,
but not how it is divided between cohabiting individuals or how much
is consumed or disposed of. Plating and serving sizes can be managed
by tracking weights on scales or weight-sensitive surfaces.
Snacking is an important component of food intake that is often
missed. This could be tracked through purchasing events for out of
home snacks and from inventory changes in the home. The use of so-
called “loyalty cards” with food retailers enables the purchaser's de-
tailed purchasing records to be linked to their personal details. At
present this is used to estimate demographics of purchasing for mar-
keting purposes, but there is no technical reason why these data could
not be used as part of the food record for that person.
It has been estimated that nearly half the world's population eats out
at least once a week (Nielsen, 2016). Understanding of the nutritional
values for restaurant and other food service dishes is therefore im-
portant. The nutritional composition of food from the major chains of
restaurants in the United States is available through the website
Menustat.org, and these values are likely to be valid for those chains
where they extend outside the United States, although this would need
to be verified. A difficulty in using the nutritional values from many of
these restaurants is that while the composition of their dishes is known,
serving sizes can vary, thus altering intake values (Niederman, Leonard,
& Clapp, 2017). Small independent restaurants cannot provide reliable
nutritional data: being able to report and manage meal composition is
beyond the means of most, both technically and in terms of compliance
costs (Brown et al., 2017).
Direct approaches to monitoring diet include image analysis and
wearable devices (Hassannejad et al., 2017). There are smartphone
apps that can identify, sometimes with surprising accuracy, what is on a
plate, and wearable cameras (e-buttons) can take a photo of a plate of
food and identify the components of a serving. These methods are not
always accurate, and improvements may be possible when coupled with
other contextual information that may be available, such as location
(from smartphone GPS, for example for specific types of restaurant) or
known preferences (Herranz, Jiang, & Xu, 2017).
Smartphones are the most convenient device for diet monitoring.
The image processing-based methods currently in use are too resource-
demanding for smartphone capability. Weaknesses of image analysis
systems are that they give only a two-dimensional image of the food,
and so they are not good at estimating serving size, and that they
analyze only the surface of the food, which, in the case of a non-
homogeneous food, means they are of little use. The recent develop-
ment of cameras that take 3-dimensional images, such as the Intel de-
vice called RealSense (https://www.intel.com/content/www/us/en/
architecture-and-technology/realsense-overview.html), a little camera
that gives 3D surface scans with mm scale accuracy, may go some way
to better estimation of portion size. Another weakness is that condi-
ments that may be sprinkled on food, most importantly salt (blood
pressure) and sugar (blood glucose) are not able to be seen and mea-
sured. A recent review concluded: “We believe that food image analysis is
not likely to provide a general fully automated mobile stand-alone solution in
the near future. Nevertheless, it can be integrated with other technologies to
form comprehensive solutions and improve the monitoring system's effec-
tiveness and accuracy.” (Hassannejad et al., 2017).
Smart Plate Top View appears to be a partial solution – it is a system
that combines a segmented, weighed plate with smartphone photo-
graphy, image recognition software and artificial intelligence to
monitor what is on your plate, interpret food intake and make re-
commendations. This system has the same limitation concerning foods

636
M. Boland and J. Bronlund
Table 2
Smart devices that could be used to track food consumption.
Name of device What it does
Diet eyeglasses Recognizes chewing
Electromyography Recognizes swallowing
Autodietary Acoustic recognition of eating
Chefling Smartphone or tablet app - records inventory and suggests recipes
Neo Smart Jar Jar that is programmed for what it will contain, nutritional values, monitors contents and
reports/reorders via Bluetooth and an app.
SmartPlate Intelligent nutrition platform that uses advanced photo recognition, inbuilt scales and AI
technology to identify, analyze, and track what is eaten.
GeniCan Barcode scanning attachment for garbage can – reorder when you use something.
HapiFork Monitors eating rate and eating habits
SCio Scanner Near Infrared scanner smartphone attachment – gives surface composition of food
KogniChef Prototype assistive kitchen – provides instruction and controls kitchen devices for meal
preparation
that are non-homogeneous as other image recognition apps, but does
allow identification between suggested options, and does manage por-
tion size better. It does not link to health and other biological data (such
as nutritional phenotype) and it appears to be designed only for non-
animal foods. An earlier version of this device has been withdrawn, and
a new improved version is projected to become available in July 2019.
Another device that has potential but is not currently being widely
used for nutrition purposes is the smartphone linked pocket-size near
infrared scanner (SCiO - www.consumerphysics.com). This device has
been shown to be useful for measuring the ripeness (sugar content) of
fruit (Kaur, Kunnemeyer, & Mcglone, 2017) and may have wider ap-
plication, as near infrared spectroscopy can be used for measuring the
composition of food items.
The use of on-body sensors is another possibility. A wide range of
different wearable sensors have been developed, with many available
commercially; some are listed in Table 2. These devices are still in their
infancy, especially when applied to food and eating, but more accurate
and more sophisticated ways to determine what and when we eat are to
be expected.
A recent comprehensive review of technology-based tools for
dietary intake assessment concluded: “Dietary assessment methods that
utilize technology offer many advantages for research and are often pre-
ferable to consumers over more traditional methods. In order to meet general
quality standards, new technology tools require detailed publications de-
scribing tool development, food identification and quantification, customi-
zation, outputs, food composition tables used, and usability/validity testing.”
(Eldridge et al. (2019).
8. Putting it all together
The full eNutrition solution will be a virtual personal dietitian, and
assistive system that monitors what food is consumed, provides alerts
about foods that should be consumed or avoided, or inappropriate
portion sizes, at the same time monitoring health and body functions on
a semi-continuous basis. It should also be able to advise on recipes and
menu choices for food. Most of the necessary components exist in some
form at present, although many are quite rudimentary and not espe-
cially accurate.
A schematic of how a fully functional assistive Personalized
eNutrition management system might work is shown in Fig. 1. We note
that similar systems have been proposed by Bashiardes, Godneva,
Elinav, and Segal (2018) with a particular focus on metabolic syndrome
and by Boland, Alam, and Bronlund (2019) with a focus on personali-
zation.
The heart of the system is what we have here designated an “App”.
This App is, in fact, an artificial intelligence system that uses the various
data streams – about aspects of health, activity, preferences, body data,
food purchasing and food consumption – and provides timely advice,
637
Trends in Food Science & Technology 91 (2019) 634–639
Where to find out more (web sites accessed March 2019)
Zhang, Bernhart, Amft, and Ieee (2016)
Nederkoorn, Smulders, and Jansen (1999)
Bi et al. (2016)
http://www.chefling.net/
https://innovationessence.com/understand-pantries-accordily/
https://www.indiegogo.com/projects/smart-plate-topview-
your-personal-nutritionist#/
https://www.genican.com/
https://www.hapi.com/product/hapifork
https://www.consumerphysics.com/
https://www.dietsensor.com/product/scio-scanner/
Neumann et al. (2017)
warnings or reminders about aspects of food purchasing and con-
sumption to support the health of the client. Whether the App can be
based solely on a smartwatch, on a smartphone or needs to be based in
the cloud is not clear and it is possible that a mixture of the three will be
used. The functionality and data flows of the system are shown sche-
matically in Fig. 2.
The Artificial Intelligence Engine (AI) must offer dietary advice
from inputs such as food intake, physical exercise and medical history.
The model itself will be based around balances, carried out to predict
energy, macro and micro nutrient intake vs consumption based on food
composition databases and recommended daily allowances. Machine
learning allows improvement of the AI engine advice based on the in-
dividual's health response resulting from adoption of past advice, en-
abling personalization of the system. These ‘personalization's’ will ac-
count for differences between people such as those due to metabolic
phenotype diversity (Holmes et al., 2008) and microbiome. Training of
the AI to an individual could be done by applying the AI to their current
diet for some initial period prior to offering advice to achieve nutrition
and health goals.
Metabolomic researchers are developing sophisticated models of
human metabolism, where key metabolites, networks of metabolism
pathways and reaction and transport fluxes are mathematically mod-
elled. One such model is the community-developed Recon 2.2
(Swainston et al., 2016), which includes 5324 metabolites, 7785 reac-
tions and 1675 genes and can predict ATP production on different
carbon sources and in different conditions. While these models are
developed for disease study, identification of new biomarkers, under-
standing host-microbe interactions (Thiele et al., 2013) and drug dis-
covery (Kell & Goodacre, 2014), they could be used to develop sim-
plified metabolic models to extend the accuracy of AI predictions. A
potential outcome of this approach could be characterization of an in-
dividual's “Biochemical Profile” that links dietary intake and exercise to
physiological status. This information could assist medical profes-
sionals, nutritionists and the AI itself to advise on changes towards
health aspirations (e.g. prescription of metformin to change carbohy-
drate metabolism and storage).
A personalized nutrition AI must not only be personalized with re-
spect to the response of the individual to dietary uptake, it must also
consider preference when proposing dietary suggestions. The food
market place is driven by personal choice and to be accepted by con-
sumers, a personalized nutrition app must facilitate choice as well as
tailoring suggestions that will be readily accepted by its user. As well as
avoiding foods that have drug or allergic responses, foods that a con-
sumer dislikes should also be avoided. Foods or recipes aligned with
cultural or favorite cuisines should be prioritized while still maintaining
choice or variety. Behavioral cues may be derived from wearable device
data to aid in providing appropriate dietary advice. For example, signs
of ongoing stress from elevated blood pressure or heart rate could
M. Boland and J. Bronlund Trends in Food Science & Technology 91 (2019) 634–639

Fig. 1. Overview of the hardware of theeNutrition system and its interactions.

promote the suggestion of comfort food or something that can be pre-
pared rapidly. The flexibility and adaptive nature of AI and machine
learning can facilitate development of these outcomes.
The smartphone is an essential component for implementation of an
AI engine, enabling access to the internet and datasets in the cloud, and
the camera function allows photography of food items and menu list-
ings. The code-reading function of the camera will allow conversion of
barcodes and QR codes on packaging to nutritional data about food
products by using the online datasets.
The smartwatch component has two main functions, one providing
timely physical and physiological data about the client to the App,
while also being the primary carrier of alerts. Most smartwatches today
produce a reasonable estimate of energy expenditure based on the
number of steps taken (for a review of this function see (Xie et al.,
2018)) and also pulse rate, based on an interpretation of blood vessel
expansion and contraction, from reflectance of the skin (photo-
plethysmography) (Kao, Chao, & Wey, 2018). While this is a reasonably
good measure for pulse rate, further extension to measure blood

Fig. 2. Schematic of the data linkages and flows in the eNutrition system.

638
M. Boland and J. Bronlund
pressure and interpret heart function is still in its infancy and although
it is available on some smartwatches, it cannot yet be relied on for
accurate measurement (although consistency appears to be good).
9. A possible scenario of the eNutrition system in action
The eNutrition system for an individual should initially be set up in
consultation with a nutritionist or other appropriate professional. This
should set daily targets, to inform the AI system, probably via an online
questionnaire. In the longer term, the AI system may refine or modify
those targets.
The kinds of advice and warnings that can be expected from the
system, via the text function of the smartwatch (and alarm if necessary)
can include:
• Meal suggestions
• Recipe suggestions (may give links to existing online recipes)
• Advice on menu choices in restaurants
• More exercise needed (now standard on many smartwatches)
• Calories needed (e.g. following exercise, when previous intake has
been insufficient)
• Calories not needed
• Protein needed (important for sports nutrition and elderly nutrition)
• Sugar warnings – for diabetics and prediabetics – linked to blood
glucose monitoring
• Salt warnings and caffeine warnings – linked to blood pressure
monitoring
• Allergy or incompatible food ingredient warnings – after scanning or
photography of food intended for consumption
• More healthy alternative foods – when food is being purchased, a
function currently available in the FoodSwitch App (Dunford et al.,
2014).
In addition to advice, the system should provide reports on a regular
basis, usually daily, via the smartphone on nutrients consumed, energy
expended and energy balance with comparison to norms and targets.
One of the problems with all diets is lack of compliance, although
there is some evidence that smartphone apps promote improved com-
pliance (Ipjian & Johnston, 2017; Patel, Hopkins, Brooks, & Bennett,
2019). This can also be the case for an eNutrition system, but the im-
mediacy and proactive nature of the system will provide prompts: the
system should suggest better alternatives and try to adapt the client's
nutrition to a healthier pattern, while not being over-assertive. To get
good compliance, it will probably be necessary to incorporate some
kind of reward system (beyond the obvious benefits of better health,
which may not be immediately appreciated). How this might work is
beyond the scope of this discussion, but needs to be explored.
The consequence of a well-developed personalized eNutrition
system should be better food choices made by the client, which will
lead to better health outcomes and lower demand on public health
systems: benefits to the individual and to society.
Acknowledgement
This work was supported by The Riddet Institute Centre of Research
Excellence funding.
References
Bashiardes, S., Godneva, A., Elinav, E., & Segal, E. (2018). Towards utilization of the
human genome and microbiome for personalized nutrition. Current Opinion in
Biotechnology, 51, 57–63.
Bi, Y., Lv, M., Song, C., Xu, W., Guan, N., & Yi, W. (2016). Auto dietary: A wearable
acoustic sensor system for food intake recognition in daily life. IEEE Sensors Journal,
16, 806–816. https://doi.org/10.1109/jsen.2015.2469095 Available:.
Boland, M., Alam, F., & Bronlund, J. (2019). Modern technologies for personalized nu-
trition. In C. Galankis (Ed.). Trends in personalized nutrition (pp. 137–164). (1st ed.).
Academic Press.
639
Trends in Food Science & Technology 91 (2019) 634–639
Brown, T., Vanderlinden, L., Birks, A., Mamatis, D., Levy, J., & Sahay, T. (2017). Bringing
menu labelling to independent restaurants: Findings from a voluntary pilot project in
toronto. Canadian Journal of Dietetic Practice and Research, 78, 177–181. https://doi.
org/10.3148/cjdpr-2017-014 Available:.
Cade, J. E. (2017). Measuring diet in the 21st century: Use of new technologies.
Proceedings of the Nutrition Society, 76, 276–282. https://doi.org/10.1017/
s0029665116002883 Available:.
Chen, J., Gemming, L., Hanning, R., & Allman-Farinelli, M. (2018). Smartphone apps and
the nutrition care process: Current perspectives and future considerations. Patient
Education and Counseling, 101, 750–757. https://doi.org/10.1016/j.pec.2017.11.011
Available:.
Dunford, E., Trevena, H., Goodsell, C., Ka Hung, N., Webster, J., Millis, A., et al. (2014).
FoodSwitch: A mobile phone app to enable consumers to make healthier food choices
and crowdsourcing of national food composition data. Jmir Mhealth and Uhealth, 2.
https://doi.org/10.2196/mhealth.3230 Available:.
Eldridge, A. L., Piernas, C., Illner, A.-K., Gibney, M. J., Gurinovic, M. A., De Vries, J. H.
M., et al. (2019). Evaluation of new technology-based tools for dietary intake as-
sessment-an ILSI europe dietary intake and exposure task force evaluation. Nutrients,
11. https://doi.org/10.3390/nu11010055 Available:.
GBD 2017 Diet Collaborators. (2019). Health effects of dietary risks in 195 countries,
1990–2017: A systematic analysis for the global burden of disease study 2017. Lancet
Published online April 3, 2019 https://doi.org/10.1016/S0140-6736(19)30041-8.
Hassannejad, H., Matrella, G., Ciampolini, P., De Munari, I., Mordonini, M., & Cagnoni, S.
(2017). Automatic diet monitoring: A review of computer vision and wearable
sensor-based methods. International Journal of Food Sciences & Nutrition, 68, 656–670.
https://doi.org/10.1080/09637486.2017.1283683 Available:.
Herranz, L., Jiang, S., & Xu, R. (2017). Modeling restaurant context for food recognition.
IEEE Transactions on Multimedia, 19, 430–440. https://doi.org/10.1109/tmm.2016.
2614861 Available:.
Holmes, E., Loo, R. L., Stamler, J., Bictash, M., Yap, I. K., Chan, Q., & Daviglus, M. L.
(2008). Human metabolic phenotype diversity and its association with diet and blood
pressure. Nature, 453(7193), 396.
Holzmann, S. L., Proell, K., Hauner, H., & Holzapfel, C. (2017). Nutrition-Apps: Quality
and limitations an explorative investigation on the basis of selected example-apps.
Ernahrungs Umschau, 64, M260–M269.
Ipjian, M. L., & Johnston, C. S. (2017). Smartphone technology facilitates dietary change
in healthy adults. Nutrition, 33, 343–347. https://doi.org/10.1016/j.nut.2016.08.003
Available:.
Kao, Y.-H., Chao, P. C. P., & Wey, C.-L. (2018). Towards maximizing the sensing accuracy
of an cuffless, optical blood pressure sensor using a high-order front-end filter.
Microsystem Technologies-Micro-and Nanosystems-Information Storage and Processing
Systems, 24, 4621–4630. https://doi.org/10.1007/s00542-018-3877-3 Available:.
Kao, C.-K., & Liebovitz, D. M. (2017). Consumer mobile health apps: Current state, bar-
riers, and future directions. Pm&R, 9, S106–S115. https://doi.org/10.1016/j.pmrj.
2017.02.018 Available:.
Kaur, H., Kunnemeyer, R., & Mcglone, A. (2017). Comparison of hand-held near infrared
spectrophotometers for fruit dry matter assessment. Journal of Near Infrared
Spectroscopy, 25, 267–277. https://doi.org/10.1177/0967033517725530 Available:.
Kell, D. B., & Goodacre, R. (2014). Metabolomics and systems pharmacology: Why and
how to model the human metabolic network for drug discovery. Drug Discovery
Today, 19(2), 171–182.
Maria Bejar, L., Adrian Reyes, O., & Dolores Garcia-Perea, M. (2018). Electronic 12-Hour
Dietary Recall (e-12HR): Comparison of a Mobile Phone App for Dietary Intake
Assessment With a Food Frequency Questionnaire and Four Dietary Records, Vol 6Jmir
Mhealth and Uhealthhttps://doi.org/10.2196/10409 Available.
Nederkoorn, C., Smulders, F. T. Y., & Jansen, A. (1999). Recording of swallowing events
using electromyography as a non-invasive measurement of salivation. Appetite, 33,
361–369. https://doi.org/10.1006/appe.1999.0268 Available:.
Neumann, A., Elbrechter, C., Pfeiffer-Lessmann, N., Koiva, R., Carlmeyer, B., Ruether, S.,
et al. (2017). KogniChef": A cognitive cooking assistant. Kunstliche Intelligenz, 31,
273–281. https://doi.org/10.1007/s13218-017-0488-6 Available:.
Niederman, S. A., Leonard, E., & Clapp, J. E. (2017). Restaurant nutrition reporting and
impact on surveillance. Journal of Food Composition and Analysis, 64, 73–77. https://
doi.org/10.1016/j.jfca.2017.04.011 Available:.
Nielsen (2016). WHAT’S IN OUR FOOD AND ON OUR MIND - INGREDIENT AND DINING-
OUT TRENDS AROUND THE WORLD. The Nielsen Company. Available: http://www.
nielsen.com/content/dam/nielsenglobal/eu/docs/pdf/Global%20Ingredient%20and
%20Out-of-Home%20Dining%20Trends%20Report.pdf.
Patel, M. L., Hopkins, C. M., Brooks, T. L., & Bennett, G. G. (2019). Comparing self-
monitoring strategies for weight loss in a smartphone app: Randomized controlled
trial. Jmir Mhealth and Uhealth, 7. https://doi.org/10.2196/12209 Available:.
Rodbard, D. (2016). Continuous glucose monitoring: A review of successes, challenges,
and opportunities. Diabetes Technology & Therapeutics, 18, 3–13. https://doi.org/10.
1089/dia.2015.0417 Available:.
Swainston, N., Smallbone, K., Hefzi, H., Dobson, P. D., Brewer, J., Hanscho, M., &
Kyriakopoulos, S. (2016). Recon 2.2: From reconstruction to model of human me-
tabolism. Metabolomics, 12(7), 109.
Thiele, I., Swainston, N., Fleming, R. M., Hoppe, A., Sahoo, S., Aurich, M. K., &
Thorleifsson, S. G. (2013). A community-driven global reconstruction of human
metabolism. Nature Biotechnology, 31(5), 419.
World Health Organization Noncommunicable diseases. Key facts. accessed March 2019
https://www.who.int/news-room/fact-sheets/detail/noncommunicable-diseases.
Xie, J., Wen, D., Liang, L., Jia, Y., Gao, L., & Lei, J. (2018). Evaluating the validity of
current mainstream wearable devices in fitness tracking under various physical ac-
tivities: Comparative study. JMIR mHealth and uHealth, 6. https://doi.org/10.2196/
mhealth.9754 e94-e94. Available:.
Zhang, R., Bernhart, S., Amft, O., & Ieee (2016). Diet eyeglasses: Recognising food
chewing using EMG and smart eyeglasses. 2016 ieee 13th international Conference on
Wearable and implantable body sensor networks.