ArticlePDF Available

Development of Urine Hydration System Based on Urine Color and Support Vector Machine

Authors:

Abstract and Figures

Nowadays, to maintain hydration in human body has becoming an important issue in health research. The water needs depend on many factors like body size, dietary intake, gender and physical activity, consequently the indicator of hydration status is highly individual. Based on recent research that said the urine color is reliable indicator for hydration status, we would like to develop a prototype for detecting hydration status automatically. To use the color indicator precisely, we tested several color sensors to performance in detecting color and chose TCS34725 as the color sensor. The prototype of urine hydration system (UHS) is designed to record hydration status data in daily basis in cloud computing and the urine information can be accessed by Android smartphone. The prototype employs a set of microcontrollers and sensors as IoT devices and support vector machine (SVM) as a classifier. To evaluate the accuracy of hydration status, we compared the prediction with urine specific gravity (USG) as golden standard. The accuracy of our UHS prototype can reach up to 84%, which is nearly similar with the accuracy of manual prediction using only human eyes.
Content may be subject to copyright.
ScienceDirect
Available online at www.sciencedirect.com
Procedia Computer Science 135 (2018) 481–489
1877-0509 © 2018 The Authors. Published by Elsevier Ltd.
This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0/)
Selection and peer-review under responsibility of the 3rd International Conference on Computer Science and Computational
Intelligence 2018.
10.1016/j.procs.2018.08.200
10.1016/j.procs.2018.08.200
© 2018 The Authors. Published by Elsevier Ltd.
This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0/)
Selection and peer-review under responsibility of the 3rd International Conference on Computer Science and Computational
Intelligence 2018.
1877-0509
Available online at www.sciencedirect.com
ScienceDirect
Procedia Computer Science 00 (2018) 000000
www.elsevier.com/locate/procedia
1877-0509 © 2018 The Authors. Published by Elsevier Ltd.
This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0/)
Selection and peer-review under responsibility of the 3rd International Conference on Computer Science and Computational Intelligence 2018.
3rd International Conference on Computer Science and Computational Intelligence 2018
Development of Urine Hydration System Based on Urine Color and
Support Vector Machine
Alexander A S Gunawan a*, David Brandon b, Velinda Dwi Puspa b, Budi Wiweko c
a Mathematics Department, School of Computer Science, Bina Nusantara University, Jakarta, Indonesia 11480
b Computer Science Department, School of Computer Science, Bina Nusantara University, Jakarta, Indonesia 11480
c Department of Obstetrics and Gynecology, Faculty of Medicine, Universitas Indonesia, Jakarta, Indonesia
Abstract
Nowadays, to maintain hydration in human body has becoming an important issue in health research. The water needs depend on
many factors like body size, dietary intake, gender and physical activity, consequently the indicator of hydration status is highly
individual. Based on recent research that said the urine color is reliable indicator for hydration status, we would like to develop a
prototype for detecting hydration status automatically. To use the color indicator precisely, we tested several color sensors to
performance in detecting color and chose TCS34725 as the color sensor. The prototype of urine hydration system (UHS) is
designed to record hydration status data in daily basis in cloud computing and the urine information can be accessed by Android
smartphone. The prototype employs a set of microcontrollers and sensors as IoT devices and support vector machine (SVM) as a
classifier. To evaluate the accuracy of hydration status, we compared the prediction with urine specific gravity (USG) as golden
standard. The accuracy of our UHS prototype can reach up to 84%, which is nearly similar with the accuracy of manual
prediction using only human eyes.
© 2018 The Authors. Published by Elsevier Ltd.
This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0/)
Selection and peer-review under responsibility of the 3rd International Conference on Computer Science and Computational
Intelligence 2018.
Keywords: urine hydration system; color sensor; cloud computing; support vector machine; urine specific gravity
* Corresponding author. Tel.: +0-000-000-0000 ; fax: +0-000-000-0000 .
E-mail address: aagung@binus.edu
Available online at www.sciencedirect.com
ScienceDirect
Procedia Computer Science 00 (2018) 000000
www.elsevier.com/locate/procedia
1877-0509 © 2018 The Authors. Published by Elsevier Ltd.
This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0/)
Selection and peer-review under responsibility of the 3rd International Conference on Computer Science and Computational Intelligence 2018.
3rd International Conference on Computer Science and Computational Intelligence 2018
Development of Urine Hydration System Based on Urine Color and
Support Vector Machine
Alexander A S Gunawan a*, David Brandon b, Velinda Dwi Puspa b, Budi Wiweko c
a Mathematics Department, School of Computer Science, Bina Nusantara University, Jakarta, Indonesia 11480
b Computer Science Department, School of Computer Science, Bina Nusantara University, Jakarta, Indonesia 11480
c Department of Obstetrics and Gynecology, Faculty of Medicine, Universitas Indonesia, Jakarta, Indonesia
Abstract
Nowadays, to maintain hydration in human body has becoming an important issue in health research. The water needs depend on
many factors like body size, dietary intake, gender and physical activity, consequently the indicator of hydration status is highly
individual. Based on recent research that said the urine color is reliable indicator for hydration status, we would like to develop a
prototype for detecting hydration status automatically. To use the color indicator precisely, we tested several color sensors to
performance in detecting color and chose TCS34725 as the color sensor. The prototype of urine hydration system (UHS) is
designed to record hydration status data in daily basis in cloud computing and the urine information can be accessed by Android
smartphone. The prototype employs a set of microcontrollers and sensors as IoT devices and support vector machine (SVM) as a
classifier. To evaluate the accuracy of hydration status, we compared the prediction with urine specific gravity (USG) as golden
standard. The accuracy of our UHS prototype can reach up to 84%, which is nearly similar with the accuracy of manual
prediction using only human eyes.
© 2018 The Authors. Published by Elsevier Ltd.
This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0/)
Selection and peer-review under responsibility of the 3rd International Conference on Computer Science and Computational
Intelligence 2018.
Keywords: urine hydration system; color sensor; cloud computing; support vector machine; urine specific gravity
* Corresponding author. Tel.: +0-000-000-0000 ; fax: +0-000-000-0000 .
E-mail address: aagung@binus.edu
482 Alexander A S Gunawan et al. / Procedia Computer Science 135 (2018) 481–489
2 Author name / Procedia Computer Science 00 (2018) 000000
1. Introduction
Water is essential for human health. Without water, human body can survive only for a few days. Depending on
the stage of human life, total body water (TBW) of infant is about 75% body weight, but in elderly is only about
55% 1. Nevertheless, there are many unanswered questions about water in of our body metabolism. To prevent
dehydration humans have evolved a finely sensitive system of physiological controls to maintain TBW by thirst.
Humans have many reasons for drinking, but most of drinking is due to the thirst which triggered by water
deficiency. Nowadays, the mechanism of thirst is understood well, and the reason why human may drink only for
pleasure is associated with large capacity of the kidney to quickly remove excess water with urine secretion 2. To
maintain hydration in human body requires a complex coordination of many sensors in the body, which is done by
integrative centers in the brain to process the data. Most fluid balance is controlled by mechanism of homeostasis
which correspond to the TBW. This mechanism is sensitive to TBW change both water deficits or excesses, and
precise to only a few hundred milliliters 2. In the brain, there are two controlling sensors: one for drinking by
sending a signal to the brain and the other for urine secretion by sending a signal to the kidneys. The kidneys play an
important role in regulating TBW, which work better in the presence of a plentiful water supply. If there is only a
few of water in the body, the kidneys will produce a more concentrated urine and make their tissues work harder.
Therefore, drinking enough water will protect this vital organ.
It is important to know about the effects of variation in TBW on health and human performance. In the past,
researches on water requirements were just based on observing the water intake from foods and drinks of the healthy
individuals. Beyond this hydration state, we do not really understand how dehydration affects human, for example
the influence of water intake to chronic diseases. We want to know about the influence of water intake, which is
important to prevent illness and maintain health. Recently, Isabelle Guelinckx et al 3 addressed the dehydration
question based on the human physiology. Previous research has associated total fluid intake or water intake with the
risk for chronic kidney disease (CKD). Urine biomarkers have been linked to the risk of CKD, and these
biomarkers, mainly arginine vasopressin (AVP) respond quickly to variations in fluid intake. AVP is molecule that
is recognized as a very important part in water homeostasis. However, the AVP also have impacts to human health.
When humans have lower intake of water, they have higher circulating AVP and thus higher urine concentration,
they tend to have increased risk of developing pathologies, increased risk for diabetes mellitus or cardiovascular
disease and chronic kidney disease (CKD). The interconnections between AVP, water intake and metabolic health
can be seen in Fig 1.
Fig. 1. Interconnection between total fluid intake to urine variables and AVP 3
The water needs depend on internal human factors, such as body weight, dietary intake, gender, exercises, and
external factors such as climate, temperature, and disease. Therefore, the indicator of hydration status is highly
individual. Three simple indicators of hydration status are: body weight, thirst, and urine osmolality (the total
number of solutes in a urine solution). Based on Thornton 4, increased water intake is related with loss of body
weight. But its mechanisms through decreased feeding and increased lipolysis take a relatively long time, thus body
weight is not suitable for tracking hydration. The second option, thirst naturally lefts behind an acute change in
hydration mainly in exercise session. Moreover, the thirst responses of older persons are less sensitive comparing to
Alexander A S Gunawan et al. / Procedia Computer Science 135 (2018) 481–489 483
Author name / Procedia Computer Science 00 (2018) 000000 3
younger persons 2. These reasons make the thirst is viewed as not ideal method of tracking hydration. The last
choice to monitor individual hydration status is by tracking urine osmolality or urine concentration. Urine
concentration is the end-result of human metabolism required to maintain TBW in response to variation of hydration
status. Thus, measuring urine concentrations has great benefits for individual assessment of daily water intake.
Perrier et al 5 stated that the urine color correlates with urine concentration in situations and can be used for
monitoring hydration status accurately. Substantial changes in urine color happen daily because of the amount of
fluid intake. This fact suggests that individuals can use urine color to monitor their hydration status in simple way.
The results of Perrier et al 5 research states that a change in urine color by 2 shades on an 8-color scale (see Fig
2), can be attained with an assessable change in water intake about 1200 mL/day. Moreover, the change in urine
color shades correspond to to change in urine volume and its specific gravity. This result has inspired us to use urine
color as indicator of hydration status.
Fig. 2. 8-color scale urine hydration chart
Nowadays, many people live in sedentary lifestyle especially in urban areas. In sedentary lifestyle, people are
often sitting while engaged in an activity like working, reading, socializing, watching television or using computer
for much of the day. This lifestyle tends to make people ignore their daily water intake because their activity does
not make a sense of thirst. Therefore, the sedentary group inclines to consume lower amounts of water and other
beverages than in the other groups 6. This habit can lead to chronic dehydration, where water that out from our body
was bigger than the water continually. The risk of dehydration in the elderly is due to increased fluid loss coupled
with a decrease in fluid intake associated with decreased thirst 7.
Research of The Indonesian Regional Hydration Study (THIRST) 8 found that as many as 46.1% of the 1200
Indonesians in DKI Jakarta, West Java, East Java and South Sulawesi were mildly dehydrated. It is also found that
the number of adolescents who experienced mild dehydration higher than adults due to low knowledge about the
function of water for the body. Another study 9 found that the dehydration risk factors in adolescents are ecological
areas, gender, body temperature, hydration knowledge, and fluid intake, when to explain why adolescents got
dehydration higher than adult. Another study 10 found that the dehydration factors are depends on age, gender,
working time and also work load. To increase hydration knowledge, there is book 11 that shows how water can cure
diseases by only adjusting the water and salt intake. Nevertheless, this approach is not effective because we do not
know exactly our hydration status. Even though we can check daily the hydration status using urine color chart, we
do not remember or care our history of hydration status. Due to this motivation, we would like to develop urine
hydration system based on IoT technology which can monitor and store our hydration status and give
recommendation about our health condition. The urine hydration system will be based on urine color aligned with
the solid study 5 to identify the hydration status. There is a research 12 which use color and ammonia level in human
484 Alexander A S Gunawan et al. / Procedia Computer Science 135 (2018) 481–489
4 Author name / Procedia Computer Science 00 (2018) 000000
urine to identify the hydration status. This research uses low quality color sensor and naïve Bayes classifier, so their
experiments give low accuracy in the identification step. To build a better accuracy in identification of hydration
status, we develop the system based on high quality color sensor and support vector machine (SVM) as classifier.
The remainder of this paper is composed as follows: first we discuss the design of urine hydration system (UHS)
in section 2, and then is followed by hardware design and its implementation., in section 3. In section 4, we discuss
software design including SVM algorithm for classifying the urine color. Furthermore, we report the experiment
result in hydration status identification based on urine color in section 5. Finally, we summarize our work with
suggestions to our future research in section 6.
2. Urine Hydration System
The urine hydration system (UHS) is designed to detect hydration status automatically based on the urine color.
We relied on Perrier et al 5 research that said the urine color is reliable indicator for hydration status. To use the
color indicator precisely, we tested several color sensors to performance in detecting color and compared their
datasheets. The used color sensor is TCS34725 (see Fig. 3) where this color sensor has better ability in determining
red-green-blue (RGB) comparing the other color sensors. The TCS34725 13 has RGB and clear light sensing
elements and also an IR blocking filter for allowing color measurements to be made accurately. The TCS34725
sensor also has an incredible 3,800,000:1 dynamic range with adjustable integration time and gain.
Fig. 3. TCS34725 color sensor
UHS is also designed to record hydration status data in daily basis in cloud computing Firebase. Firebase can be a
platform to connect IOT with Android smartphone. It is mobile development platform that provides a database
hosted in the cloud. Firebase stores data as JSON and can update the data in real-time by synchronizing to connected
client. After the sensor read the urine color, UHS will send the color data to the local server for color classification.
Local server which is built using python, will classify the color data based on urine hydration chart by using support
vector machine (SVM) algorithm. Finally, the user profile from NFC tag, time and date, urine color data, and its
classification will be sent to cloud computing Firebase. Fig 4 depicts the design of the UHS components and their
interactions .
Fig. 4. The design of urine hydration system (UHS)
Furthermore, UHS is designed such that the users can monitor their hydration status conditions in the form of
texts or graphs any time dan can also update their profiles. The next two sections will discuss hardware and software
development of UHS in detail.
Alexander A S Gunawan et al. / Procedia Computer Science 135 (2018) 481–489 485
Author name / Procedia Computer Science 00 (2018) 000000 5
3. Hardware Design and Implementation
In this section, we would like to describe about the color detection device, which illustrates in Fig. 4. The color
detection device, in principle, is an embedded device to get urine color data based on the color sensor. The
components of this device consist of microcontroller Arduino Uno and Wemos D1, color sensor TCS34725, and
RFID MRC522. To build the UHS prototype, the microcontroller Arduino needs to be combined with the sensor
color sensor and RFID reader. Wemos D1 is Wi-Fi based board that uses layout similar to Arduino Uno. It is used to
connect the UHS device with local server and cloud computing Firebase. Fig. 5 is scheme of the UHS prototype,
including its wiring.
Fig. 5. The scheme of UHS prototype
The procedure to use the color detection device as follow: (1) users attach NFC or RFID tags, such as electronic
ID card to the reader RFID MRC522 which connected on Arduino Uno, (2) user ID profiles which is read will be
sent to Wemos D1 and integrate with the urine color data which read by the color sensor, (3) urine color data is sent
to the local server for color classification, (4) Finally, all data including time and date will be sent to cloud
computing. The appearance of or color detection device can be seen in Fig.6 together with illustration how to take
the urine color data.
(a)
(b)
Fig. 6. (a) a whole prototype (b) taking urine color data on the UHS
4. Software Design and Support Vector Machine
There are three programming languages which used in the software design of our UHS. To control and connect
the microcontroller (Arduino Uno and Wemos D10) to other devices, it is used Arduino -based C language. While
486 Alexander A S Gunawan et al. / Procedia Computer Science 135 (2018) 481–489
6 Author name / Procedia Computer Science 00 (2018) 000000
the local server, which implemented SVM for urine color classification, used Python as programming language.
Finally, to develop android applications is used Java programming language in Android studio.
We implemented three communication types in the microcontroller based on Arduino-based C language for
sensor data collection and device communication, as following:
1. To communicate between Arduino Uno and Wemos D1, we used serial port, which also known as a universal
asynchronous receiver-transmitter (UART). In Arduino Uno, it communicates via digital pins 0 (RX) and 1
(TX) to the pair pins in Wemos D1 (see Fig. 5). Serial communication on pins TX/RX in Arduino Uno uses 5V
TTL logic levels. This type communication is asynchronous serial communication using transmitter (TX) and
receiver (RX) in which the data format and transmission speed can be configured 14.
2. To communicate between Arduino Uno and RFID reader MRC522, it is used SPI (Serial Peripheral Interface)
in which communication is synchronous 14. SPI is a synchronous serial data protocol used by microcontrollers
for communicating with peripheral devices rapidly over short distances. To implement SPI communication, it
needs three pins in Arduino Uno (see Fig.5), that is: (1) digital pin 11 for MOSI (Master Out Slave In), (2)
digital pin 12 for MISO (Master In Slave Out), and (3) digital pin 13 for SCK (Serial Clock).
3. To communicate between Wemos D1 and the color sensor TCS34725, we used I2C (Inter-Integrated Circuit)
communication protocol. I2C is a serial protocol for two-wire interface to connect microcontrollers in
embedded systems, which has the following features: synchronous, multi-master and multi-slave. In Wemos
D1, we can implement I2C communication by using two pins (see Fig.5), that is: (1) digital pin 16 for SDA
(data line) and (2) digital pin 17 for SCL (clock line).
In local server, we developed simple HTTP server and classification function based on support vector machine
(SVM). SVM is a discriminative classifier formally defined by a separating hyperplane 15. Hyperplane in an n-
dimensional Euclidean space is a flat, n-1 dimensional subset of that space that divides the space into two
disconnected parts. SVM find the optimum separation hyperplane for a given finite set of learning patterns. We
implemented SVM classifier by using scikit-learn library in Python programming language. For training SVM
algorithm, we used 100 dataset, consisting of three RGB color as input data and one label of hydration status based
on urine specific gravity (USG). The dataset is generated by normal random generator based on urine color chart
(Fig. 2) and its related urine specific gravity as label (Table 1) and time data. This table is based on Minton et al 16.
Example of the training dataset can be seen in Table 2.
Table 1. Hydration status based on urine specific gravity (USG)
USG Value
Label
Meaning
1000 - 1010
0
hydrated and healthy (HH)
1011 - 1020
1
properly dehydrated (PD)
1021 - 1030
2
seriously dehydrated (SD)
>1030
3
highly dehydrated (HD)
Table 2. Example of training dataset
R
G
B
Label
USG
Time
85
90
75
0
1005
Noon
65
88
88
0
1000
Afternoon
77
85
85
0
1000
Noon
108
78
49
2
1023
Morning
76
88
84
0
1000
Afternoon
In our UHS prototype, Android application is mainly used to retrieved information for cloud computing Firebase.
When the users register or update the profile in the application, their profile will be stored at Firebase. The user data
is synchronized to their RFID or NFC tag by attaching the tag to the smartphone (see Fig. 7a) Furthermore, the
application will display the urine information based on periodically measurements, in the form of sequences of text
Alexander A S Gunawan et al. / Procedia Computer Science 135 (2018) 481–489 487
Author name / Procedia Computer Science 00 (2018) 000000 7
data. The information which are showed are user profile, date and time of measurements, urine RGB color, and the
hydration status (see Fig. 7b). Finally, we can see the result of hydration status in graphical form. Consequently, we
can see changes in hydration status more easily because the graph shows comparisons on every hour of each day
(see Fig. 7c).
5. Experiment Result
For testing data, we collected urine data from 52 participants. Each participant collected one urine sample in
morning, afternoon and noon randomly. As the training data, urine specific gravity of each sample as golden
standard is measured by using digital refractometer and labeled based on Table 1 after that the urine color is
measure using our prototype. The confusion matrix of experiments can be seen in Table 3. In general, the average
accuracy of the UHS prototype is about 84%.
(a)
(b)
(c)
Fig. 7. (a) Form of user profile (b) urine history in text form (c) urine history in graphical form
488 Alexander A S Gunawan et al. / Procedia Computer Science 135 (2018) 481–489
8 Author name / Procedia Computer Science 00 (2018) 000000
Table 3. Confusion matrix of the experiments
Actual
Prediction
n =52
HH
PD
SD
HD
HH
24
5
0
0
PD
3
2
0
0
SD
0
0
18
0
HD
0
0
0
0*
Note: hydrated and healthy (HH), properly dehydrated (PD), seriously dehydrated
(SD), highly dehydrated (HD). There no participants who is highly dehydrated.
Average accuracy: 84% based on 52 data
(24+2+18/52)
Here some examples of prediction of the testing data comparing to USG measurement using digital refractometer:
Table 4. Example of testing data
Urine Detection System
Comparison
Matching
Red
Green
Blue
Time
Prediction Status
USG
Hydration Status
83
89
71
Afternoon
HH
1009
HH
85
91
73
Noon
HH
1004
HH
90
85
71
Morning
SD
1018
PD
x
92
92
71
Afternoon
HH
1014
HH
6. Conclusion
In this paper, we have successfully developed prototype of urine hydration system (UHS) to detect and monitor
urine hydration status based the urine color. Based on the experiment results, the average accuracy of this prototype
is about 84%. We can still improve the prototype accuracy by adding more training dataset and tuning the SVM
parameters. This result is nearly similar with the study of THIRST 8 that compares manual prediction using only
human eyes with USG measurement using digital refractometer. The current implementation is still restricted for
manual urine measurement. Thus, we would like to design further so that our UHS prototype can be installed in
toilette as Benavides et al 14 and then detect the hydration status automatically. These are remaining for our future
works.
References
1.
Horswill CA, Janas LM. Hydration and Health. American Journal of Lifestyle Medicine. 2011;: p. 304-315.
2.
Popkin BM, D'Anci KE, Rosenberg IH. Water, hydration, and health. Nutrition Reviews. 2010 August; 68(8).
3.
Isabelle Guelinckx , Mariacristina Vecchio , Erica T. Perrier , Guillaume Lemetais. Fluid Intake and
Vasopressin: Connecting the Dots. Annals of Nutrition and Metobolism. 2016 June; 68.
4.
Thornton SN. Increased Hydration Can Be Associated with Weight Loss. Frontiers in Nutrition. 2016; 3(18).
5.
Perrier ET, Johnson EC, McKenzie AL, Ellis LA, Armstrong LE. Urine colour change as an indicator of change
in daily water intake: a quantitative analysis. European Journal of Nutrition. 2015 August; 55.
6.
Aparicio-Ugarriza R, Luzardo-Socorro R, Palacios G, Bibiloni MdM, Julibert A, Tur JA, et al. Impact of
physical activity and sedentarism on hydration status and liquid intake in Spanish older adults. Nutrición
Hospitalaria. 2016; 33(Supl. 3): p. 4-8.
7.
Thomas DR, Cote TR, Lawhorne L, Levenson SA, Rubenstein LZ, Smith DA, et al. Understanding Clinical
Dehydration and Its Treatment. Journal of the American Medical Directors Association. 2008; 9(5): p. 292-301.
Alexander A S Gunawan et al. / Procedia Computer Science 135 (2018) 481–489 489
Author name / Procedia Computer Science 00 (2018) 000000 9
8.
Santoso BI, Hardinsyah , Siregar P, Pardede SO. Air bagi kesehatan. 3rd ed.: Centra Communications; 2017.
9.
Adiningsih S, Hardinsyah , Lestari KS, Nindya TS, Mahmudiono T. Habitual Water Intake Among Adolescent
in Low Land and High Land Areas at East Java. The Indonesian Journal of Public Health. 2012 July; 9(1).
10.
Yi W, Chan APC. Effects of Heat Stress on Construction Labor Productivity in Hong Kong: A Case Study of
Rebar Workers. International Journal of Environmental Research and Public Health. 2017; 14(1055): p. 1-14.
11.
Batmanghelidj F. Air - Untuk Menjaga Kesehatan dan Menyembuhkan Penyakit Jakarta: Gramedia Pustaka
Utama; 2007.
12.
Amani RZ, Maulana R, Syauqy. Sistem Pendeteksi Dehidrasi Berdasarkan Warna dan Kadar Amonia pada Urin
Berbasis Sensor TCS3200 Dan MQ135 dengan Metode Naive Bayes. Jurnal Pengembangan Teknologi
Informasi dan Ilmu Komputer. 2017;: p. 436-444.
13.
Adafruit. RGB Color Sensor with IR filter and White LED - TCS34725. [Online]. [cited 2018 February.
Available from: https://www.adafruit.com/product/1334.
14.
Benavides M, Mailier J, Hantson AL, Muñoz G, Vargas A, Impe JV, et al. Design and Test of a Low-Cost RGB
Sensor for Online Measurement of Microalgae Concentration within a Photo-Bioreactor. Sensors. 2015; 15: p.
4766-4780.
15.
Kung SY. Kernel Methods and Machine Learning. 1st ed.: Cambridge University Press; 2014.
16.
Minton DM, O'neal EK, Torres McGehee TM. Agreement of urine Specific Gravity Measurements Between
Manual and Digital Refractometers. Journal Of Athletic Training. 2015;: p. 59-64.
... Table 1. Criteria for classification of a dehydrated state [12,18,19,36,37]. ...
... Urine color [19,37] 5-8 ...
Article
Full-text available
The aim of this study was to determine the effects of different seasons of the year and the time of day (before work vs. after work) on hydration status in men. The study involved sixty foresters who spent most of the work outdoors. During three seasons of the year (summer, autumn, and winter), indices of hydration status (body mass (BM) and percentage change of BM, total body water (TBW) and percentage change of TBW, serum osmolality (Sosm) and percentage change of Sosm, urine osmolality, urine-specific gravity (USG), urine color, and thirst) were determined before work on the first day (time point 1 used as baseline), immediately after work on the first day (time point 2), and before work on the following day (time point 3). USG decreased at time point 2 compared to time point 1 (p < 0.001) and time point 3 (p = 0.03). At time point 2 (p = 0.002) in winter and time point 3 in autumn (p = 0.049), serum osmolality was higher than in summer. In conclusion, the differences in hydration status depended on the time of day and season. A large percentage of foresters come to work inadequately hydrated, especially in colder seasons compared to summer.
... The use of artificial intelligence (AI) enabling software is used to employ AI abilities while detecting colour [35]. The colour thresh-holding technique found in MATLAB is in conjunction with the RGB model to detect selected colour and can improve the impact of AI in colour classification as reported in [24]. The development of a urine colour system for a urine hydration system is discussed in [25]. ...
Method
Full-text available
The conventional way of analyzing urine colour is done by visual examination, which is often paired with urine clarity. A visual examination of urine for its colour and clarity is done to identify abnormal urine colouration, the darkness of the urine, blood in the urine, foamy urine, etc. The naked eye is used to know how clear or cloudy and the colouration of urine. It employs physical, microscopic, and chemical examination of urine. This approach is primitive and poses a great level of incompetence, which is below the present state of technology and advancement in our society. This paper gives an overview of the traditional colour sensing techniques and the use of sensors to improve colour identification accuracy. The review studies the use of various technologies such as colour sensors, smartphones, and LDR to get a better outcome in detecting urine colour using sensors and integrating artificial intelligence in the area of urine analysis.
... Example of a urine color chart for hydration assessment. 141 drugs. 72 MWCNTs are also being developed for the simultaneous detection targets with overlapping electrochemical oxidation peaks. ...
Article
Full-text available
Urine testing is a powerful clinical diagnostic tool. The noninvasive collection of samples and wide range of diagnostic targets found in urine makes urinalysis well suited for point-of-care (PoC) monitoring applications. Complete urinalysis testing faces many limitations due to the large quantity of samples processed, the time required for testing, and the labor involved in sample preparation and processing. Development of PoC urinalysis devices with microfluidic technology can enable the detection of infections and monitoring of chronic disease while reducing the demand on testing facilities. In this article, current approaches in clinical urinalysis are reviewed. Emerging sensing and imaging technologies specifically suitable for point-of-care examination of urine samples are discussed with an outlook on the future of point of care urinalysis devices as well as emerging applications enabled by these technologies such as in situ monitoring of Activities of Daily Living (ADL).
Chapter
A 20-year increase in life expectancy among the general population since the 1960s (World Bank) brings with it several health-related challenges, especially if people are to continue to live independently in their own homes. Long term conditions such as stroke, diabetes and dementia and their impact on independent living are well studied, however ways to prevent more intermittent conditions such as dehydration - the loss of body fluid, and hypothermia - the decrease of body temperature, both of which can result in hospitalization, have been studied less. Our research is focused on developing a novel Ambient Assisted Living (AAL) system, HyDeAdCo (Hypothermia and Dehydration Advising Companion), to help detect when an older person might be at risk of becoming dehydrated or developing hypothermia and advise them on how to reduce this risk. HyDeAdCo is an Internet of Things (IoT) system comprised of a social robot interface, commercial-of-the-shelf (COTS) devices and cloud computing to analyse, store, and send responses to the user based on sensor data. This paper describes the premise of the HyDeADCo prototype system and presents the role of the social robot, Temi, that acts as an interface between the user and the system in order that the risk of a person becoming dehydrated or developing hypothermia is reduced.
Article
The purpose of the current study was to assess effects of a hydration management program on the fluid consumption and dehydration parameters (i.e., dehydration risk scores, urine specific gravity, and urine color) of older adults. The study was conducted as a pre-/post-test quasi-experimental study with a control group (intervention group: n = 42, control group: n = 37). The hydration management program applied to the intervention group was completed in 1 week. After the intervention, the knowledge and awareness of the intervention group regarding fluid consumption increased (p = 0.001). The dehydration risk score of the intervention group was significantly lower than that of the control group (p = 0.0001). There was a significant change in average 3-day water intake (p = 0.0001), urine specific gravity value (p = 0.01), and urine color scale score (p = 0.0001) in the intervention group between pre- and post-test. The hydration management program provided a significant improvement in daily water intake and dehydration risk score and increased knowledge and awareness among older adults concerning the importance of fluid intake. [Journal of Gerontological Nursing, 47(9), 40-48.].
Article
United States’ animal shelters take in 3.2 million cats per year. Caged cats likely experience stressful situations during their stay at an animal shelter. However, a feasible and efficient way to determine which cats will be at the highest risk for severe stress, and subsequent health and behavior deterioration, is needed. We aimed to determine whether urinary output was suppressed immediately following a presumably stressful move from one location to another in the shelter and whether this suppression correlated with behavior in cats admitted to an animal shelter. Prior to addressing this aim, we first evaluated the use of litter clump weight to determine urine output. Then we aimed to determine whether behavior, using an established in-cage assessment, predicted urinary suppression in cats. Litter clump weight was significantly predicted by quantities of water added to the litter (50 mL, 100 mL, and 150 mL), suggesting that litter clump weight can be used as a marker of crude urine output. Next, newly-admitted shelter cats (n = 29) were subjected to an in-cage behavioral assessment. For the following week, urine clumps were weighed daily. Clump weight increased significantly (F = 17.926, p < 0.0001) after Day 1, suggesting an initial suppression perhaps due to acute stress (i.e. moving to the adoption floor). A principle component analysis (PCA) of in-kennel behavior followed by linear regression was conducted to predict Day 1 urine suppression and found that the sixth and eighth components (PC 6 and PC 8) showed modest prediction of urine suppression. A support vector machine (SVM) was also used to classify cats as having “less” vs. “more” urine suppression at three different cutoffs (10%, 25%, 50% suppression on Day 1). The SVM had an average training accuracy of 86.3%, 81.8%, 84.1% at 10%, 25%, and 50% suppression respectively. These data suggest that urinary suppression (measured through litter clump weights) may be a practical way to track wellbeing in shelter cats and that a SVM may have modest ability to classify cats at risk of suppressed urine output based on behavioral responses to a brief test.
Article
Full-text available
Global warming is bringing more frequent and severe heat waves, and the result will be serious for vulnerable populations such as construction workers. Excessive heat stress has profound effects on physiological responses, which cause occupational injuries, fatalities and low productivity. Construction workers are particularly affected by heat stress, because of the body heat production caused by physically demanding tasks, and hot and humid working conditions. Field studies were conducted between August and September 2016 at two construction training grounds in Hong Kong. Onsite wet-bulb globe temperature (WBGT), workers’ heart rate (HR), and labor productivity were measured and monitored. Based on the 378 data sets of synchronized environmental, physiological, construction labor productivity (CLP), and personal variables, a CLP-heat stress model was established. It was found that WBGT, percentage of maximum HR, age, work duration, and alcohol drinking habits were determining factors for predicting the CLP (adjusted R² = 0.68, p<0.05).The model revealed that heat stress reduces CLP,with the percentage of direct work time decreasing by 0.33% when the WBGT increased by 1 ◦C. The findings in this study extend the existing practice notes by providing scientific data that may be of benefit to the industry in producing solid guidelines for working in hot weather.
Article
Full-text available
In the last decade, cross-sectional and multiple cohort studies have associated total fluid intake or water intake with the risk for chronic kidney disease (CKD) and even the risk of developing hyperglycemia. Urine biomarkers have also been linked to the risk of CKD and lithiasis, and these biomarkers respond quickly to variations in fluid intake. High circulating copeptin levels, a surrogate marker of arginine vasopressin, have been associated with metabolic syndrome, renal dysfunction and increased risk for diabetes mellitus, cardiovascular disease and death. The aim of this paper was to explore how the various findings on water intake, hydration and health are interconnected, to highlight current gaps in our understanding and to propose a model that links water intake, homeostatic mechanisms to maintain water balance and health outcomes. Since plasma copeptin and vasopressin have been demonstrated to be sensitive to changes in water intake, inversely associated with 24-hour urine volume, and associated with urine biomarkers and fluid intake, vasopressin is proposed as the central player in this theoretical physiological model.
Article
Full-text available
This mini-review develops the hypothesis that increased hydration leads to body weight loss, mainly through a decrease in feeding and a loss of fat through increased lipolysis. The publications cited come from animal, mainly rodent, studies where manipulations of the central and/or the peripheral renin-angiotensin system lead to an increased drinking response and a decrease in body weight. This hypothesis derives from a broader association between chronic hypohydration (extracellular dehydration) and raised levels of the hormone angiotensin II (AngII) associated with many chronic diseases such as obesity, diabetes, cancer and cardiovascular disease. Proposed mechanisms to explain these effects involve an increase in metabolism due to hydration expanding cell volume. The results of these animal studies often can be applied to the humans. Human studies are consistent with this hypothesis for weight loss and for reducing the risk factors in the development of obesity and type 2 diabetes.
Article
Full-text available
Urine colour (U Col) is simple to measure, differs between low-volume and high-volume drinkers, and is responsive to changes in daily total fluid intake (TFI). However, to date, no study has quantified the relationship between a change in TFI and the resultant change in U Col. This analysis aimed to determine the change in TFI needed to adjust 24-h U Col by 2 shades on an 8-colour scale, and to evaluate whether starting U Col altered the relationship between the change in TFI and change in U Col. We performed a pooled analysis on data from 238 healthy American and European adults (50 % male; age, 28 (sd 6) years; BMI 22.9 (sd 2.6) kg/m(2)), and evaluated the change in TFI, urine volume (U Vol), and specific gravity (U SG) associated with a change in U Col of 2 shades. The mean [95 % CI] change in TFI and U Vol associated with a decrease in U Col by 2 shades (lighter) was 1110 [914;1306] and 1011 [851;1172] mL/day, respectively, while increasing U Col by 2 shades (darker) required a reduction in TFI and U Vol of -1114 [-885;-1343] and -977 [-787;-1166] mL/day. The change in U Col was accompanied by changes in U SG (lighter urine: -.008 [-.007;-.010]; darker urine: +.008 [.006;.009]). Starting U Col did not significantly impact the TFI change required to modify U Col by 2 shades. Our results suggest a quantifiable relationship between a change in daily TFI and the resultant change in U Col, providing individuals with a practical means for evaluating and adjusting hydration behaviours.
Article
Full-text available
Although water is actually a nutrient, there is a tendency that some people neglect its important roles for health; some said water is a forgotten nutrient. Public-private-academic partnerships are urgently required for better ecology, better drinking water supply, and better drinking water consumption for better healthy life and better physical performance of the Indonesian’s people. This study is part of The Indonesian Regional Study, called THIRST. The objectives of this study are to obtain scientific evidences of nutrition practice and knowledge of drinking rated of hydration status among teenagers. For this purpose, a cross sectional study design was applied in selected municipal or districts (Surabaya, Malang), which reflects different ecologies and regions of Indonesia. A total sample of 200 subjects was selected - 100 teenagers (15–18 yrs) from each selected municipal or districts with proportionally by sex. Data collected include socio-economic, food and beverages intake, body weight and height, physical activity. The results showed that most adolescents prefer to drink water. Tea and coffee is the second and the third most preferred beverages for them. Safety is the main reason for them in choosing type of beverages to be consumed. Only two-thirds of subjects knew about the amount of water needed every day; about mineral water is better than carbonated drink; and about mineral water is better than tea and coffee. About 60% of subjects did not know that water is required more during pregnancy and lactating, and during sweat and in the cold room. The study suggests the important of promoting adequate water intake as part of dietary guidelines; and developing a model for institutional-based dehydration prevention via schools, workplaces and community.
Article
Full-text available
In this study, a low-cost RGB sensor is developed to measure online the microalgae concentration within a photo-bioreactor. Two commercially available devices, i.e., a spectrophotometer for offline measurements and an immersed probe for online measurements, are used for calibration and comparison purposes. Furthermore, the potential of such a sensor for estimating other variables is illustrated with the design of an extended Luenberger observer.
Article
Full-text available
Dehydration in clinical practice, as opposed to a physiological definition, refers to the loss of body water, with or without salt, at a rate greater than the body can replace it. We argue that the clinical definition for dehydration, ie, loss of total body water, addresses the medical needs of the patient most effectively. There are 2 types of dehydration, namely water loss dehydration (hyperosmolar, due either to increased sodium or glucose) and salt and water loss dehydration (hyponatremia). The diagnosis requires an appraisal of the patient and laboratory testing, clinical assessment, and knowledge of the patient's history. Long-term care facilities are reluctant to have practitioners make a diagnosis, in part because dehydration is a sentinel event thought to reflect poor care. Facilities should have an interdisciplinary educational focus on the prevention of dehydration in view of the poor outcomes associated with its development. We also argue that dehydration is rarely due to neglect from formal or informal caregivers, but rather results from a combination of physiological and disease processes. With the availability of recombinant hyaluronidase, subcutaneous infusion of fluids (hypodermoclysis) provides a better opportunity to treat mild to moderate dehydration in the nursing home and at home.
Article
Introduction: Data on hydration status in older adults are scarce and there are very few studies focusing on the impact of physical activity (PA) on drinking behavior. Objective: To assess the impact of physical activity and sedentarism on fluid intake in Spanish elderly. Methods: 433 non-institutionalized Spanish older adults (58% females), aged 55-88 years, volunteered for the PHYSMED study. PA data were obtained by means of the Minnesota and EXERNET questionnaires. Population was divided into four groups: ILS (inactive and low sedentary), IHS (inactive and high sedentary), ALS (active and low sedentary) and AHS (active and high sedentary). Serum from fasting blood samples was analysed for osmolarity. Results: The mean of total liquid intake was 1,751 ± 628 mL/d. Significant differences were observed for total liquid intake between ILS/ALS and IHS/ALS (p < 0.001). ALS subjects consumed a higher amount of beverages such as water, juice, milk, coffee, sport drink, beer, wine and distilled drinks than the other PA groups. There was a significant difference for water intake between PA groups (p < 0.01). Serum osmolarity values were within references ranges in all subjects, and there was a signifi cant difference between PA groups (p < 0.01). Conclusions: Spanish older adults meet the DACH recommendations set by the German, Austrian and Swiss nutrition societies' liquid intake recommendations in the mean independently of PA and sedentary level. All participants are within reference ranges of serum osmolarity. Subjects in the active and low sedentary group consumed higher amounts of water and other beverages than in the other PA groups.
Article
Offering a fundamental basis in kernel-based learning theory, this book covers both statistical and algebraic principles. It provides over 30 major theorems for kernel-based supervised and unsupervised learning models. The first of the theorems establishes a condition, arguably necessary and sufficient, for the kernelization of learning models. In addition, several other theorems are devoted to proving mathematical equivalence between seemingly unrelated models. With over 25 closed-form and iterative algorithms, the book provides a step-by-step guide to algorithmic procedures and analysing which factors to consider in tackling a given problem, enabling readers to improve specifically designed learning algorithms, build models for new applications and develop efficient techniques suitable for green machine learning technologies. Numerous real-world examples and over 200 problems, several of which are Matlab-based simulation exercises, make this an essential resource for graduate students and professionals in computer science, electrical and biomedical engineering. Solutions to problems are provided online for instructors.
Article
Daily adequate water intake to maintain euhydration is arguably the most important nutrient requirement for humans. Within a margin of error, the body regulates the maintenance of body fluid balance and especially that of the plasma volume, through mechanisms that stimulate thirst and/or modify the rate of urine production. However, there are circumstances such as with excessive sweating during exercise in the heat, osmotic diarrhea, or excessive fluid consumption, or water intoxication where normal mechanisms of regulation may be inadequate to compensate for acute changes in hydration status and result in life threatening consequences. Health and onset of disease may be affected by the chronic hydration state of individual. The risks of colorectal cancer, nephrolithiasis in those with a history of kidney stones, and bladder cancer may be reduced by more frequent water consumption. Recent research suggests that appropriate timing of water intake around meal occasions may help reduce energy intake and contribute to maintenance of body weight in overweight individuals. Definitive benefits of hydration on cardiovascular and oral health and general immune system function remain to be determined. It is also unclear whether the health benefits of water and fluid ingestion are a function of the process of frequent fluid intake or the maintenance of a potentially expanded state of hydration.