Malaysian Healthy Ageing Society 1st WORL RLD CONGRE GRESS SS ON - - PowerPoint PPT Presentation

Organised by:

Malaysian Healthy Ageing Society

Co-Sponsored:

UNIVERSITI VERSITI KEBANGS GSAAN AAN MALAY AYSIA SIA Biomedical Engineering Research Cluster (Medical Devices Group)

1st WORL RLD CONGRE GRESS SS ON HEALTHY HY AGEIN ING

Vascular Age Sustainability in Ageing Health Management

Author/Presenter: Kalaivani Chellappan (PhD) Co-Author:

• Prof. Dr. Mohd Alauddin Mohd Ali
• Assoc. Prof. Dr. Nor Anita Megat Mohd Nordin

Introduction……Ageing

Ageing

• The progressive deterioration of the

near total of the functions of the

• rganization in the course of time1.
• Extrinsic Ageing - These are from

external factors. Contributing factors can be smoking, lifestyle habits such as sun exposure (sun baking), pollution, nutrition/diet, alcohol, drugs, lack of sleep, exhaustion, and

• f course, poor maintenance of
• verall health.
• Intrinsic Ageing - These are

genetically determined.

1Austad, Steven NR.; Why we age; New York; John Wiley & Sounds, Inc. ; 1997

Introduction……Healthy Ageing

Healthy Ageing

Healthy aging is the development and maintenance

• f optimal mental, social and

physical well-being and function in older adults. This is most likely to be achieved when communities are safe, promote health and well-being, and use health services and community programs to prevent or minimize disease.

1Austad, Steven NR.; Why we age; New York; John Wiley & Sounds, Inc. ; 1997

Introduction….Sustainability

• Health Sustainability
• Capacity to implement

(ABILITY) by patient

• Appropriateness in

implementation (OBERVATION) by medical practioners

• Impact of implementation

(MEASUREMENT) by ?

Unsustainable? Sustainable?

Introduction…. Health Management

Health Management

• What is needs?

Plan, design and manage

• Who’s is going to

manage? Vendor, …

• Type? Individual or

Group

• Cost? $$$$$$$$

Introduction…. Vascular Age

Vascular Age: You're only as old as your arteries

The number of candles on your birthday cake may add up to your chronological age, but it doesn't necessarily equal your biological age: environmental factors, such as stress and diet, and genetics can speed up

• r slow down how the body ages.

Proposed Solution Noninvasive & Low Cost Vascular Health Screener

NIVAR

Introduction Cardiovascular Risk Factors

Risk Factors:

1. Ageing 2. Diabetes 3. Hypertension 4. Hyperlipidemia 5. Smoking 6. Obesity

50 percent of death and disability from CVD can be reduced by a combination of simple effective national efforts and individual actions to reduce major CVD risk factors.1

• 1. Integrated Management of Cardiovascular Risk. Report of a WHO Meeting, Geneva, July 2002.

Introduction Current Clinical Approach in Vascular Risk Assessment

Noninvasive Method

• Ankle Brachial Index
• Echocardiographic

assessment

• Ultrasonography
• Others:

– Measurement of:

• body fat
• body mass index
• waist circumference
• blood pressure
• Lipid profile
• glucose levels
• Arterial stiffness

Invasive Method

• Coronarography

The present techniques are found to be:

• Limited information
• Disease base diagnostic
• Too Complex
• Costly

Need a more affordable and precise vascular assessment technique for all cardiovascular related risk.

Introduction Photoplethysmography

Photoplethysmography (PPG)

(Bhattacharya et al. 2001; Webster 1997):

• Optoelectronic method (using LED

and PD).

• Measures blood volume changed

associated with cardiac contraction.

• Obtained from finger, ear lobes and

toes.

• Low optical absorption
• High degree of vasculature.
• Widely used in:
• Blood oxygenation saturation.
• Heart rate.

Objective

Establish a vascular risk prediction index through a noninvasive assessment technique using finger PPG waveform.

Methodology

Signal processing techniques

Signal Pre-processing

• Single channel recording.

– Left or right finger

(based on whether subject is left or right handed)

• Sampling rate at 275 Hz.
• Signal detrending to remove
• utliers, drifts, offset and

movement artifacts.

• Bandpass filtering at range of 0.6 –

15Hz.

• Scaling to one: signal was

normalized to unity (range at 0 to 1).

Methodology

Clinical data acquisition

Statistical Analysis

Type of Analysis Gender Health Status Total Population Male Female Without Risk With Risk

Data Set 1 (All subjects) 142 161 135 168 303 Data Set 2F (Female only) None 161 56 105 161 Data Set 2M (Male only) 142 None 79 63 142 Data Set 3A Age(24 – 66) w/o ref 127 147 123 151 274 Data Set 3B Age(24 – 66) with ref 128 148 125 151 276 Data Set 4L Age(20 – 44) with ref 71 78 53 96 149 Data Set 4U Age(45 – 66) with ref 58 70 72 56 128 Data Set 5L

Exact Age Match (20 –44) with ref

53 47 50 50 100 Data Set 5U

Exact Age Match(45 – 66) with ref

39 45 42 42 84

Statistical Analysis

Description Total Male Female Number of Subject 303 142 161 Minimum Age 17 17 19 Maximum Age 80 80 76 Mean 44.00 44.35 43.69 Std Deviation 12.66 13.62 11.77 Variance 160.25 185.68 138.64 Without\With Risk 168\135 79\63 56\105

Methodology

Empirical Data Analysis and Modelling

Data Analysis…1

Total number of subject: 184 Age range: 20 – 66 yrs Group 1: 100 age matched subjects (50 without risk & 50 with risk) Age range: 20 – 44 yrs Group 2: 84 age matched subjects (42 without risk & 42 with risk) Age range: 45 – 66 yrs

Remarks: Without Risk: Subjects without any clinically diagnosed cardiovascular disease or changeable risk factors and non- smokers. With Risk: Subjects with anyone of the risk only. Risk: Hypertension, Diabetes, Hyperlipidemia & Smoking.

Reference Signal Establishment

• 30 subjects aged 19 years old without risk for each

gender was recruited.

• PPG data recorded according to the experiment

protocol.

• Variability assessment was carried out.
• The variability between subjects for each gender was

less than 5% in all recording consistently.

Health Index Definition

• PPG Health Index measuring the percentage difference between the reference signal

and test signal. The reference signal is a 19 year old male and female. The health index assessment is gender base.

• Following are the steps:

– Removing mean from both reference and test signal. – Difference between reference and test sample standard deviation divide by test sample standard deviation.

• The standard deviation value will be small if the values are clustered tightly about

their mean and vice versa.

• As such the above fitness value can be defined as a Vascular Health Index

) ) ( )) ( ) (( 1 ( 100 _

2 2

 

       y y y y x x PCT fitness

Signal Processing

• PPG Peak Detection

algorithm will detect every single valley in the entire signal length.

• Best Pulse Selection

algorithm will calculate the fitness of every signal pulse in the entire signal length. The median of this pulses are calculated and the pulse which is closes to the average median will be selected and matched against the reference pulse.

) ) ( )) ( ) (( 1 ( 100 _

2 2

 

       y y y y x x PCT fitness

Application

Identification of a person at risk of cardiovascular risk factors. Major application: vascular risk-prediction

Model Output

25

Risk Index

Group 1 (20 – 44) yrs Group 2 (45 – 66) yrs

As such Residual (R): No Risk : R  6 Low Risk : 6  R  12 Moderate Risk : 12  R  20 High Risk : R > 20 As such Residual (R): No Risk : R  7 Low Risk : 7  R  16 Moderate Risk : 16  R  22 High Risk : R > 22

Area Under ROC Curve 1.00 0.963

Variability & Repeatability

Repeatability on 10 subjects: Plot of Coefficient of Variability vs recording If recoding done on the same day, interval time is 30 minutes

Variability & Repeatability

Repeatability on 10 subjects: Plot of Coefficient of Repeatability vs recording If recoding done on the same day, interval time is 30 minutes

Conclusion

• Variability

The variability between pulses are less than 10% in all recording consistently.

• Repeatability

The coefficient of repeatability is more than 90% for all recording consistently. The fitness for anyone subject recorded on a particular day do not vary much, but subjects recorded in far apart time frames shows large variation in fitness.

Prototype

Prototype GUI

References

1. Samer S. Najjar, Angelo Scuteri, Edward G. Lakatta. 2005. ‘Arterial Aging Is It an Immutable Cardiovascular Risk Factor?’. Hypertension;46;454-462. 2. Claessens P, Claessens C, Claessens M, Claessens M, Claessens J. 2002. ‘The 'CARFEM' vascular index as a predictor of coronary atherosclerosis.’ Medical Science Monitoring:8(1):MT1-9. 3. J Allen and A Murray. 2002. “Age-related changes in peripheral pulse timing characteristics at the ears, fingers and toes’. Journal of Human Hypertension (2002) 16, 711–717. 4. John R Cockcroft, Ian B Wilkinson, David J Webb. 1997.“Age, arterial stiffness and the endothelium”. Age and Aging: 26-S4: 53 – 60. 5. Aminbakhsh A, Frohlich J, Mancini GB. 1999.”Detection of early atherosclerosis with B mode carotid ultrasonography: assessment of a new quantitative approach.” Clinical Invest Med, 1999; 22(6): 265-74 6. Bernard I. Lévy. 2001. “Artery changes with aging: degeneration or adaptation?” Dialogues in Cardiovascular Medicine,Vol 6:2: 104-111 7. G Gerstenblith, J Frederiksen, FC Yin, NJ Fortuin, EG Lakatta and ML Weisfeldt. 1977.“Echocardiographic assessment of a normal adult aging population”. Circulation;56;273-278

Future Health Management

• Preventive
• Predictive
• Participative
• Personalize
• PRACTICAL >>>>> 5th P needs

– Affordability – Usability – Simplicity

Looking forward ……

• Collaboration:

– Malaysia

• Vascular Screening for improvement

– Studies on vascular improvement based on: » Exercise (Traditional & Non-traditional) » Therapy » Others

– Other countries

• Population establishment
• Other vascular risk (country specific)