E‐mail address: santoshpbk@gmail.com Study programme: Masters in electronics design, 120p Examiner: Börje Norlin Scope: 14,819 words inclusive of appendices Date: 2011‐12‐20
M.Sc. Thesis report within Electronics Engineering,
International Masters programme in Electronics
Design, 30hp.
Tremor Quantification and
Parameter Extraction
Abstract
Tremor is a neuro degenerative disease causing involuntary muscle movements in human limbs. There are many types of tremor that are caused due to the damage of nerve cells that surrounds thalamus of the front brain chamber. It is hard to distinguish or classify the tremors as there are many reasons behind the formation of specific category, so every tremor type is named behind its frequency type. Proper medication for the cure by physician is possible only when the disease is identified.
Because of the argument given in the above paragraph, there is a need of a device or a technique to analyze the tremor and for extracting the parameters associated with the signal. These extracted parameters can be used to classify the tremor for onward identification of the disease.
There are various diagnostic and treatment monitoring equipment are available for many neuro‐muscular diseases. This thesis is concerned with the tremor analysis for the purpose of recognizing certain other neurological disorders. A recording and analysis system for human’s tremor is developed.
The analysis was performed based on frequency and amplitude parameters of the tremor. The Fast Fourier Transform (FFT) and higher‐ order spectra were used to extract frequency parameters (e.g., peak amplitude, fundamental frequency of tremor, etc). In order to diagnose subjects’ condition, classification was implemented by statistical significant tests (t‐test).
Keywords: Tremor analysis; FFT; Higher‐order statistics; Data acquisition; Statistical test
Acknowledgements
It is with great pleasure that I wish to acknowledge several people who have helped me tremendously during the difficult, challenging, yet rewarding and exciting path towards my thesis. Without their help and support, none of this work could have been possible.
I would like to thank Dr. Borje Norlin, Associate professor, Mid‐Sweden University, Sweden for permitting me to take up the project and his encouragement and support for finishing the masters program in electronics design.
I am greatly indebted to my guide Mr. M.Venkateswara Rao, Associate professor, Osmania University, India for his guidance, encouragement, motivation and continued support throughout my thesis work. He has allowed me to pursue my research interests with sufficient freedom, while always being there to guide me. Working with him has been one of the most rewarding experiences of my professional life.
I am also deeply indebted to Dr.Roopam Borgohain, HOD, Neurology department, NIMS, for supporting my work. Special thanks to Dr.Aruna, Neurology department, NIMS Hospital for providing the records of the Patients and helping me in recording the Tremor Movements Bejugam Santosh Kumar
Table of Contents
Abstract ... i Acknowledgements ... ii Table of Contents ... iii Terminology ... vi Abbreviations ... vi 1 Introduction ... 1 1.1. Characteristic of tremor ... 2 1.2. Measurement of Tremor ... 3 1.3. Goal of the project ... 3 2 Theory ... 5 2.1. Tremor ... 5 2.2. Physiological Tremor ... 6 2.3. Diseases Inducing Tremor ... 6 2.3.1. Ataxia ... 7 2.3.2. Athetosis ... 7 2.3.3. Agraphia ... 7 2.3.4. Alcoholism ... 8 2.3.5. Basal ganglia ... 8 2.3.6. Dyskinesia ... 8 2.3.7. Dystonia ... 8 2.3.8. Huntington’s disease (HD) ... 9 2.4. Types of Tremor ... 9 2.4.1. Parkinson Tremor ... 9 2.4.2. Essential Tremor ... 9 2.4.3. Resting or Static tremor ... 11 2.4.4. Postural Tremor ... 11 2.4.5. Action Tremor ... 11 2.4.6. Kinetic Tremor ... 11 2.4.7. Intention Tremor ... 11 2.4.8. Cerebellar Tremor ... 12 2.4.9. Dystonic Tremor ... 12 2.4.10. Psychogenic Tremor ... 13 2.4.11. Orthostatic Tremor ... 14 2.4.12. Physiologic Tremor ... 142.4.13. Holmesʹ Tremor ... 14 2.4.14. Drug‐induced Tremor ... 14 2.5. Accelerometer Sensor ... 15 2.5.1. Principle of Operation ... 16 2.5.2. Pin diagram of Accelerometer ... 17 2.5.3. Pin description of Accelerometer ... 18 2.6. Microcontroller Circuit ... 20 2.6.1. Pin Diagram of R8C/1B MCU ... 20 2.6.2. Block Diagram of R8C/1B MCU ... 20 2.7. MAX 232 ... 22 2.7.1. Features ... 23 2.7.2. Pin Diagram of MAX232 ... 23 2.7.3. Max 232 Logic diagram ... 24 2.7.4. MAX 232 pin description ... 25 2.8. RS 232 Serial port ... 25 2.8.1. Pin Description of DB9 Pin ... 27 3 Methodology ... 28 3.1. Relevant work using accelerometer in the past ... 30 3.1.1. Measurement of Involuntary Hand Motion in the past ... 31 3.1.2. Tracking of human limb movement using accelerometer by ... 32 Morris 32 3.1.3. Measuring Chemical Effects on Involuntary Hand Tremor ... 33 3.1.4. Measurement of Hand Acceleration using Accelerometer ... 34 3.2. Types of accelerometer sensors ... 35 3.2.1. Strain Gauge Accelerometer ... 35 3.2.2. Piezoelectric Accelerometer ... 36 3.2.3. Piezo‐resistive Accelerometer ... 37 3.2.4. Capacitive Accelerometer ... 39 4. Design ... 41 4.1. Hardware design ... 41 4.1.1. Power Supply ... 41 4.1.2. Accelerometer Interface ... 44 4.1.3. Interfacing between MAX232 and DB9 connector ... 45 4.2. Tremor Hardware Kit Pin Configuration ... 45 4.3. Software design ... 47 4.3.1. Embedded software ... 48 4.3.2. MATLAB Software ... 48 5. Results ... 51 5.1. Recordings Using Accelerometer ... 51 5.2. Measurement Using Electromyography (EMG) ... 55
5.3. Comparison between EMG and Accelerometer ... 56 6. Conclusions ... 58 7. References ... 60 8. Appendix A: Documentation of own developed program code ... 62 8.1. Microcontroller Program for data acquisition from accelerometer ... 62 8.2. Mat lab program for data acquisition from microcontroller ... 63 8.2.1. Fast Fourier analysis using Mat lab ... 65 9. Appendix B: Paper Presented in International Conference on Electrical and Electronics Engineering ... 67
Terminology
Abbreviations
AC Alternating Current ADC Analog to Digital Converter CMOS Complementary Metal Oxide Semiconductor CPU Central Processing Unit DAC Digital to Analog Converter DC Direct Current DT Dystonic Tremor EIA Electronics Industry Alliance EMG Electromyography ET Essential Tremor FFT Fast Fourier Transform IIC Inter Integrated Circuit I/O Input/output KT Kinetic Tremor MATLAB Matrices Laboratory MCU Micro Controller Unit MEMS Micro Electro Mechanical System MPU Micro Processors Unit NIMS Nizam Institute of Medical Science PC Personal Computer PCB Printed Circuit Board PD Parkinson disease PT Parkinson tremor RAM Random Access Memory RT Resting Tremor ROM Read Only Memory RS232 Recommended Standard 232 RXD Receive Data SCI Serial Communication Interface SPI Serial Peripheral Interface SPP Standard Parallel Port TTL Transistor‐Transistor Logic TIA Telecommunication Industry Association TXD Transmit Data
1 Introduction
Tremors are rhythmic, involuntary and oscillatory movements of muscles at joints (4). Loss of neurotransmitters like dopamine, gamma‐ amino butyric acid, and serotonin activity in the brain leads to the tremor (5).
Apart from the above mentioned, there are other reasons for Tremor cause i.e. due to the loss of body supportive forces like arteries or heart dysfunction, and the death of certain nerve cells by the usage or stoppage of certain drugs. Static, Dynamic form of tremors and physiological or pathological cause of tremors are the major classifications, besides which dynamic or action tremor are of few types called postural, kinetic and isometric or rest (12).
Sensors provide information on physical or chemical parameters that are not directly received by human senses, thus the sensors receives an input signal and gives a readable output signal. They respond to physical stimulus such as pressure, vibration, acceleration, temperature, humidity, distance, sound, light etc and transforms in to signal that is used for measurement or further analysis purpose (20). Semiconductor sensors are widely used in many applications as they not only contribute with low cost and size but also a possibility to integrate with microcontroller circuits (20).
Deposition, Lithography and Etching are the main steps involve in the fabrication of semiconductor sensors. Conductive and non‐conductive layers of silicon are etched and deposited together using planar techniques to form a Micro‐Electro‐Mechanical Systems, or MEMS based sensor. Planar techniques use different two dimensional patterns to stack the layers together to form a three dimensional devices (20).
In order to improve the sensor performance (i.e. higher sensitivity, selectivity and fast response time), to reduce its size and also to lower the cost of sensor, micromachining process is extensively used since 1980. Thus the MEMS based sensors broaden their applications by integrating with micro electronics circuits (20).
The current project design uses MEMS based accelerometer sensor for measuring either linear or angular acceleration. A tri‐axial MEMS linear accelerometer sensor fabricated by Free‐scale semiconductor is used for measuring the acceleration of human hand motion. Remaining signal conditioning circuit is built by commercial available equipments.
1.1. Characteristic of tremor
Tremor is subtle in every human being but visible to naked eye depending on the physiological status like stress, fatigue, anxiety, fright, excitement and during the alcoholic intoxication (1). Major aspect of tremor is its frequency, which is the main parameter to distinguish the tremor analysis. Oscillatory activities are the possible initiators of tremors caused in the front brain. Neurons that surround the hypothalamus called basal ganglions, nerve cells associates thalamus called bulbar olive, red cerebellar nucleus, ventromedial nucleus of thalamus are responsible for tremor disorders (1).
Levy bodies are abnormal proteins form in neurons destroy their functionality by not allowing them to produce neurotransmitter called dopamine essential for regulating body movements, this is one of the main reasons of tremor cause(1).
1.2. Measurement of Tremor
Micro Electro Mechanical System (MEMS) based sensor with small size and cost is used to convert the physical parameter i.e. acceleration or vibration of limbs into electrical signal. As the body can have movement in all the three axes, a tri‐axial accelerometer is needed. The three analogue voltages are fed to the microcontroller, which in turn forwards this information to a PC for subsequent analysis in order to get desired parameters (13).
Methods reported for tremor analysis using single or dual access instruments were not quantitatively apparent for satisfied results, as the movements at specific anatomical site are three dimensional. So, tri‐axial measurements of tremor analysis using accelerometer evaluates clinical problems pertained (6).
Accelerometers are of many types, but the main types that can be incorporated with MEMS are 1.Piezoelectric, and 2.Capacitive. Acceleration generates voltage in relation with the vibration forces. As the voltage generated is analogue, it is fed to ADC (analogue to digital converter) embedded in the microcontroller and stored in digital format for analysis using computer techniques. Pathological tremors are evaluated by spectral methods namely (FFT) Fast Fourier Transforms (1).
1.3. Goal of the project
The main aim of this project is to develop a three axis micro electro mechanical system based accelerometer sensor for measuring involuntary hand motion.
Initially hand movement or tremor of the patient is to be measured using accelerometer sensor, and thereafter microcontroller must be used to acquire the output data from sensor for onward transmission to personal computer (PC) via recommended standard 232 (RS232) serial communication port. The collected data must be used for signal processing and parameter determination using matrices laboratory (Mat lab) software.
Finally, the acquired data need to be processed in such a manner to identify the parameters that can be used to diagnose / assess the subjects for certain neurological problems.
In the present report project related theory, methodology, designing steps, results of tremor quantification and features like frequency and voltage extraction are given in a sequence.
2 Theory
2.1. Tremor
Central nervous system produces most of the tremor related issues in Parkinson tremor(PT), Essential tremor(ET) and the tremor intensity is asymmetry in PT and ET where as frequency and frequency dispersion (frequency width around the centre frequency) in PT is asymmetry but symmetry in ET(2). Symmetry in frequency is the similarity in either sides or both hands when combined, which is used to differentiate the tremor type whether it is ET or PT (2).
Tremor is one of the major features of Parkinson disease (PD) called Parkinson’s tremor but the absence of tremor frequency is hard for diagnosis. It is asymmetrical or unilateral depending on its effects to more or one side of the body, usually hand shaking during rest position called rest tremor is seen in PD patients (1).
Essential tremor is a movement disturbance with its own pattern; it is inversely related between severity and age of the patient. It affects head and speech apart from hands (1). Upper Limbs and head are the main affected parts of the essential tremor which is the kinetic and postural kind of tremor and its frequency ranges from 4.5‐ 8 Hz (3).
Postural tremor occurs when the subject attempts to maintain a posture, such as maintaining the upper limbs outstretched. The following conditions are associated with postural tremor: physiological tremor, essential tremor, cerebellar tremor, post‐traumatic tremor, peripheral neuropathy (3).
Kinetic tremor occurs during purposeful movement; for example, during finger‐to‐nose test (the patient is asked to put the index finger on the nose). Kinetic tremor is highly suggestive of a cerebellar disorder (cerebellar ataxia) or a disease involving cerebellar pathways. Midbrain tremor combines rest, postural, and kinetic tremor (3). Task‐specific tremor appears when performing goal‐oriented tasks such as handwriting, speaking, or standing. This group consists of primary writing tremor, vocal tremor, and orthostatic tremor. Task‐specific tremor can be viewed as a form of kinetic tremor that appears during specific tasks (3).
2.2. Physiological Tremor
A slow vibration of approximately 10 cycles per second contributes the normal burst or contraction of voluntary muscle. It appears to be a mechanism called hunting in the reflex arc that controls the muscle.Physiological tremor can occur in a state of normality or in an exagge‐ rated form due to phenomenon such as anxiety, fever, hyperthyroidism, hypo glycaemia, excess caffeine and medication etc. Usually it is sym‐ metrical, bilateral and non‐progressive over time. There may be a chance of family history but this is less often than the other types of tremors such as essential tremor (16).
2.3. Diseases Inducing Tremor
There are many neurological diseases that are responsible for tremor in human muscles or limbs; they are Ataxia, Athetosis, Agraphia,
Alcoholism, Basal ganglia, Dyskinesia, Dystonia, and Huntington’s disease. In this section, brief words are given on all of these disorders.
2.3.1. Ataxia
Ataxia is a disease with shaky movements due to the failure in the regulation of the body’s posture, strength and direction of limb movements by the brain. Usually ataxia is the result of brain damage in the cerebellum or spinal cord occurring from head injury, infection, brain tumour, toxins, multiple sclerosis etc.
The other major types of ataxias include cerebellar ataxia, sensory ataxia and Fried Reich’s ataxia. In cerebellar ataxia, clumsiness in intentional movements occurs, i.e. walking, speaking, and eye movements.
Sensory ataxia occurs due to lack of feedback in sensory organs such as unstable movements when the patients close their eyes. Friedreich’s ataxias, a fatal genetic disease caused by the degeneration of the motor nerves in the spinal cord, causing a loss of coordination and a disturbance in gait (15).
2.3.2. Athetosis
Athetosis is the behaviour of involuntary slow writing movements due to the lack of proper functioning deep with in the brain. It is mostly found in the patients suffering from Huntington’s disease, Cerebral Palsy, Encephalitis, or other brain disorders. It also occurs as a side effect to certain medications (15).
2.3.3. Agraphia
Loss of ability to write is a form of Agraphia, though the patient has normal hand and arm‐muscle function Agraphia may be caused by brain damage such as from a brain tumour or head injury. Writing needs a complex sequence of processes in brain, such as word selection, spelling recall, functioning of hand movements, and visual agreement to match their mental representation. These varied processes apparently take place in a number of connected brain areas. Damage to any of them leads to different types of Agraphia. It often appears together with a disturbance in speaking and writing skills. But, some of the lost writing skills may eventually return. Agraphia rarely occurs alone but there is no specific treatment (15).
2.3.4. Alcoholism
In moderate amounts, alcohol imparts feelings of relaxation and confidence as it withdraws control from higher brain centres. Moreover, tests show that alcohol interacts with the brain’s activities by slowing reactions. Even few drinks a week affect the brain, but most physicians agree that limited quantities of alcohol do not usually cause nerve cell damage.
Mental deterioration and muscle damage are the different types of neurological disorders appear due to excess drinking or alcoholism. Alcohol withdrawal cause minor tremors. Long‐term alcohol abuse damages the right frontal lobe of the Cerebral Cortex, which is responsible for spatial skills and perception. This is the reason of unharmed verbal skills (15).
2.3.5. Basal ganglia
Basal ganglion is a nerve cell found in the mid brain, these neurons normally release a neurotransmitter called dopamine. So, its absence leads to Parkinson’s disease with involuntary movements, trembling and weakness. Parkinsonism is a disorder with masklike face, rigidity and slowed movements. Dopamine levels decreases as the basal ganglia nerve cells die in the brain levels (16).
2.3.6. Dyskinesia
Dyskinesia is a brain disorder of abnormal muscular movements with irresistible jerking. It occurs in either entire body or a part of body muscle groups. Different types of Dyskinesia include chorea (jerking movements), athetosis (writing), choreoathetosis (a combination of jerking and writing), tics (repetitive movements), tremors, or myoclonus (muscle spasms) (15).
2.3.7. Dystonia
Dystonia is an abnormal muscle rigidity with painful muscle spasms and strange movements. It is caused due to the side effect of antipsychotic drugs and PD (16).
2.3.8. Huntington’s disease (HD)
Huntington’s disease is hereditary disorder with involuntary movements and memory loss. This disease slowly finishes the affected individual’s ability to walk, think, talk, and reason. Eventually, people with HD become totally dependent upon others for their care (15).
2.4. Types of Tremor
Tremor occurs in any age but is most common for middle‐aged and older persons. It is occasional, temporary or occurs intermittently and affects men and women equally. Tremors are of many types and are described in this section (17, 18).
2.4.1. Parkinson Tremor
Damage in the brain structures controlling movements responsible for Parkinson tremor. This resting tremor occurs alone or associates with other neurological disorders explained in the above section. It is classically a “pill‐rolling” action of hands that also affect chin, lips, legs, and trunk, and can be slowly increased by stress or emotions.
Parkinson’s tremor usually occurs after the age of sixty; initially it starts in one side of the body and spreads progressively to other parts.
Parkinsonʹs Tremor takes place in association with other symptoms, such as micrographia, bradykinesia (slowness) and rigidity. This type of tremor is not hereditary as the family history is not seen (4).
2.4.2. Essential Tremor
Essential tremor (ET) is the predominant type of tremor among more than twenty types of tremor. ET is mild and non‐progressive in some people but it is slowly progressive in others by starting on one side of the body and spreading both sides in three years.
In general, hands are most often affected but head, voice, tongue, legs, and trunk may also be involved. Head tremor is seen as a “yes‐yes” or “no‐no” motion. Tremor frequency decreases in the life time but the
severity increases, affecting the person’s ability to perform certain tasks or activities of daily living.
Emotion, stress, fever, physical exhaustion or low blood sugar may trigger tremors and increase their severity. ET is common after age of forty, although symptoms can appear in any age. It is a hereditary disease having family history (15).
Comparison of Parkinson disease tremor and Essential Tremor are tabulated in Table 2.1 Tremor type Resting tremor (PT) Postural and action Tremor (ET) Age Older age (> 60 years) All age groups Family history Usually negative Positive in more than 60% of patients Alcohol Not beneficial Beneficial Tremor onset Unilateral Bilateral
Muscle tone Cogwheel rigidity Normal Facial expression Decreased Normal Gait Decreased arm swing Normal Tremor latency Longer 8‐9 seconds Shorter 1‐2 seconds Table 2.1: Comparison of Parkinsonʹs Tremor vs. Essential Tremor
2.4.3. Resting or Static tremor
Resting tremor (RT) is a shaking of the limb when the person is at rest. Here the muscles are not being voluntarily contracted and is completely supported against gravity. Normally when the limb is moved, the resting tremor disappears. Like all other tremors, RT is aggravated by stress or anxiety. Resting tremor is quite separate from other tremors. This type of tremor is often seen in patients with Parkinsonʹs disease (4). 2.4.4. Postural Tremor
Postural tremor appears when the person voluntarily maintains a posi‐ tion against gravity, e.g. holding the arms outstretched. Postural tremors require voluntary or purposive contraction of muscles. The common examples would be exaggerated physiological tremor and essential tremors. The other name has been given as ʺposition specific postural tremorʺ (4).
2.4.5. Action Tremor
Action tremor (AT) occurs during voluntary activation or contraction of muscles i.e. an arm outstretched requires muscle activity to hold it against gravity. AT includes postural tremors, kinetic tremors and intention tremors (4).
2.4.6. Kinetic Tremor
Kinetic Tremor (KT) is a purposeful voluntary movement of a body part. Physician determines KT by asking the subject to perform a simple rotary movements of the forearm or flexion and extension movements of the wrist. KT includes postural tremors (15).
2.4.7. Intention Tremor
Intention Tremor (IT) is a complex kinetic tremor and occurs during the muscle directing towards a particular target. It appears when the subject is asked by the clinician to touch their nose, and is called as ʺfinger to nose testingʺ.
Intentional tremor is due to the disturbance in cerebellum and its connections to other parts of the nervous system. Examples of IT include cerebellar tremor and multiple sclerosis tremors (4).
2.4.8. Cerebellar Tremor
Cerebellar Tremor (CT) is a slow, broad tremor of the extremities that appears during the purposeful movement, such as trying to press a button or touching a finger to the tip of one’s nose. It results from chronic alcoholism or overuse of particular medicines. In classic CT, a lesion on one side of the brain produces a tremor in the same side of the body and becomes worse with directed movement (4).
Cerebellar tremor produces a “wing‐beating” type of tremor called Holmes’ tremor which is a combination of rest, action, and postural tremors. The tremor is most prominent when the affected person is maintaining a particular posture. The diseases associated with CT includes dysarthria (speech problems), nystagmus (rapid, involuntary rolling of the eyes), gait problems, and postural tremor of the trunk and neck (16).
2.4.9. Dystonic Tremor
Dystonic tremor (DT) is derived from Dystonia; it is a movement disorder in which involuntary muscle contractions cause twisting and repetitive motions, painful and abnormal postures. DT can affect any muscle in the body and is seen often when the patient is in a certain position.
DT occurs irregularly and is relieved by complete rest. Touching the affected body part or muscle may reduce tremor severity. The tremor may be the initial sign of dystonia localized to a particular part of the body (4).
Different tremors that are discussed in this section are tabulated below in table 2.2 with their characteristics features and classification.
2.4.10. Psychogenic Tremor
Psychogenic tremor occurs at rest or during postural, kinetic movement. The characteristic features of this kind of tremor may vary but often include sudden onset and remission, increased incidence with stress, change in tremor direction when body part is affected, and greatly decreased tremor activity when the patient is disturbed.
Psychogenic tremor involves any part of the body, but it commonly affects the extremities. Usually, tremor onset is sudden and begins with an unusual combination of postural, action, and resting tremors (15).
Table 2.2.Characteristics of Tremors and their classification Type of
Tremor Frequency Occurrence Tremor Classification Postural tremor 5 to 9 Hz When hand joints are positioned against gravity Physiologic tremor, essential tremor, alcohol or drug withdrawal, metabolic disturbances, drug‐induced tremor, psychogenic tremor. Rest tremor 3 to 6 Hz When limb is fully supported against gravity and the muscles are not voluntarily activated. Parkinsonʹs disease, multiple‐ systems atrophy, progressive supranuclear palsy, drug‐ induced tremor, rubral tremor, psychogenic tremor Action tremor† 3 to 10 Hz During any type of movement Cerebellar lesions, rubral tremor, psychogenic tremor
2.4.11. Orthostatic Tremor
Orthostatic tremor appears in legs and trunk immediately after standing. Patient suffers from uncontrollably shaking in legs when asked to stand in one spot, this shaking stops when the patient sits or is lifted off the ground. No other clinical signs or symptoms are present. Orthostatic tremor has high frequency of about 14 Hz to 18 Hz. It also occurs in patients with essential tremor. So, orthostatic tremor is considered to be a variant of essential tremor (4).
2.4.12. Physiologic Tremor
Physiologic tremor occurs in every normal individual when maintaining a posture or movement and has no clinical significance. It is rarely visible to the eye and is increased by physical exhaustion, hypoglycemia, strong emotion, hyperthyroidism, heavy metal poisoning, stimulants, alcohol withdrawal, and fever. It is detected by extending the arms and placing a piece of paper on top of the hands. Physiologic Tremor is in general not a neurological disease but appears by side effect to certain drugs, or reaction to alcohol withdrawal, and medical conditions including an overactive thyroid and hypoglycemia. It is usually reversible once the cause is corrected. Frequency of this Tremor is very low of around 6 Hz to 12 Hz, and is hardly visible to the naked eye (16).
2.4.13. Holmes' Tremor
Holmesʹ tremor is irregular a combination of rest, postural, and action tremors. The reason behind this tremor is midbrain lesions in the vicini‐ ty of the red nucleus. It has low frequency of 4.5 Hz. Signs of ataxia and weakness may be seen (4).
2.4.14. Drug-induced Tremor
There are various types of tremors induced by drugs and they are, enhanced physiologic tremor, rest tremor, and action tremor. Signs and symptoms of drug‐induced tremors depend on the drug used on a
patient its side effects. Some drugs cause extra pyramidal side effects resulting brady kinesia, rigidity, and tremor.
Table 2.3 below is a list of drugs that may induce tremor, along with the types of tremors and neurologic signs they produce (15).
Drug or Drug Class Tremor Type Neurologic Signs
Amiodarone Postural Rarely parkinsonism
Bronchodilators Postural, action None
Lithium Rest, postural, action Extrapyramidal
Metoclopramide Rest, postural Extrapyramidal
Neuroleptics Rest, postural Extrapyramidal
Theophylline Postural None
Valproate Postural Rarely parkinsonism
Table2.3Drug‐Induced Tremor and Corresponding Neurological signs
2.5. Accelerometer
Sensor
An accelerometer is an electromechanical device that measure acceleration forces that are directly attached to it. The Free Scale accelerometer sensor was selected for use in this project for its small mass and volume, the other feature is that it has serial interface techniques such as serial peripheral interface (SPI) and Inter Integrated Circuit (IIC) with easy communication between the sensor device and the external micro controller.
Amplification and analog‐to‐digital conversion of the signal take place on same silicon wafer which reduces the noise induction in to the system. Further the sensitivity and the filter characteristics are selected via software. The sensor module connected to the external world using only 5 wires, two for power supply and the remaining three for signal transmission (9).
Accelerometry is the measurement of acceleration with electronic equipment which is relatively new in biomechanics. However, the use of accelerometers was not widespread in biomedical applications until the end of 20th century or the beginning of twenty first century. Now‐a‐days, these are widely used in biomedical applications. The recent advances in wireless and embedded system technologies such as MEMS sensors hold a great promise with built‐in signal conditioning unit. Types of accelerometers include piezo‐resistive, strain gauge, piezoelectric, and capacitive transducers (14).
The most common commercial application of accelerometer sensors includes airbag deployment in automobiles, inertial guidance mechanisms in rockets and aircraft etc (14).
2.5.1. Principle of Operation
Accelerometer sensor uses acceleration as input signal and gives voltage as output signal. The basic principle of this sensor runs behind Newton’s second law of motion
F = ma
Where ‘F’ is force, ‘m’ is mass and ‘a’ is acceleration.
Accelerometer sensor measures the displacement of the mass which is suspended by spring, the mass spring principle is shown in figure 2.1 below.
Figure 2.3 Mass-Spring principle of Accelerometer External acceleration force, damping force and the restorative force of the spring proportional to position are the respective forces that are acting on the proof mass. The mathematical equation of all the combined forces is
F = maexternal = Md2x/dt2 + B(x) dx/dt + k(x) x
In equilibrium condition i.e. when the mass is in stable position, the restorative force produced by the spring is equal to the external acceleration force on the proof mass.
The displacement of the spring x, is a parameter which is converted to a voltage or electrical signal by many different methods. For example: By measuring a change in resistance of a piezo‐resistive material, and also by measuring a change in capacitance exerted between reference and movable electrical elements (21).
2.5.2. Pin diagram of Accelerometer
The Free scale accelerometer is based on a micro‐machined capacitance technology, the device consists of two surface micro machined capacitive sensing cells or g‐cells and a signal conditioning ASIC contained in a single integrated circuit package.
A bulk micro‐machined cap wafer is used in order to hermetically seal the sensing elements at wafer level. MMA7455L is the three dimensional MEMS based Accelerometer used in this project (8). The pin diagram of the accelerometer sensor used in the present design is given below in figure 2.2. It has 16 pins in which only 7 pins are used, 2 for gravity selection i.e. 1.5g to 6g range, 3 for output in 3 mutually perpendicular axis and 2 for power supply and ground. Figure 2.2 pin diagram of accelerometer (8)
2.5.3. Pin description of Accelerometer
The accelerometer used in the present project has 16 pins; the description of each and every pin of the sensor is given in below table 2.4.
Table2.4. Pin description of the accelerometer sensor (8).
2.6. Microcontroller Circuit
The features of the micro controller used in this project are to be advanced, so I used a 16 bit controller from RENASAS, R8C/1B series. Main use is to facilitate the need of three ports for getting three analogue outputs from accelerometer corresponding to x, y, z axes acceleration signals and minimum memory requirement for software program implementation.
2.6.1. Pin Diagram of R8C/1B MCU
The microcontroller has 20 pins and its assignments vary according to the use. Among all the available pins, 6 are unidirectional in which data or communication flows towards the controller and the remaining 14 pins are bidirectional to the device that are connected to it. Top view pin assignment of R8C/1B series is shown in below figure 2.3.
Figure 2.3 Pin Diagram of R8C/1B MCU (10).
The microcontroller has to be used to collect the data for the analog to digital conversion. Several different types of microcontrollers are available in the market from different manufacturers with different capabilities. Such as in built random access memory (RAM), read on‐ ly memory (ROM), input/output (I/O) ports, Timers, ADC, digital to analog converter (DAC) etc. block diagram of MCU is shown below in figure 2.4.
Figure 2.4 Internal block diagram of R8C/1B MCU (10).
To write the program into the micro controller, a specific tool known as a writer board is needed. The writer board is not a common gadget for all the micro controllers. Even to develop the program is a Herculean job. Normally the programs can be written in assembly and are entered into a PC, via the software provided by the micro controller manufacturer; the assembly code is converted into mnemonics (Assembler).
This mnemonic code is the one, which is to be written into the controller. This is done by a writer board, which is hooked up to the PC, by using the RS‐232 serial interface. Sometimes, it is highly difficult to write the programs in assembly, in those circumstances, the programs are written in ‘C’ and the ‘C’ code is converted into the machine code using a compiler, the code thus generated can be dumped into the micro controller using a PC.
To speed up the process of development of gadget, an emulator is needed, which can be used to develop the programs by using what are called break points. The assembler and compilers, are software tools, while emulator is both hardware and software tool. Without these basic tools, no matter how good the micro controller might be, one cannot use it to its optimum performance.
2.7. MAX 232
In order to convert the TTL logic levels of microcontroller output into RS‐232 logic levels of serial port input, a converter is needed. So, Max 232 is connected between the microcontroller and the serial port of the PC. MAX232 is chosen in this project, as it is compatible with RS‐232 standard, and consists of dual transceiver. Each receiver converts telecommunication Industry association (TIA)/Electronics Industry alliance (EIA)‐232‐E levels into 5V transistor‐transistor logic (TTL) levels.
Each driver converts TTL levels into TIA/EIA‐232‐E levels. The MAX232 is characterized for operation from ‐40°C to +85°C for all packages.MAX232 is purposed for application in high‐performance information processing systems and control devices of wide application (11).
2.7.1. Features
Input voltage levels are compatible with standard СMOS levels. Output voltage levels are compatible with EIA/TIA‐232‐E levels. Features of MAX232 IC are listed in table 2.5 below.
Sl.no Parameter Specification 1. Single Supply voltage 5V 2. Low input current 0.1μA at ТA= 25 ° 3. Output current 24mA 4. Latching current >450mA at ТA= 25°С Table2.5. Features of MAX232 IC
The transmitter outputs and receiver inputs are protected to + (or)‐ 15kV Air ESD (11).
2.7.2. Pin Diagram of MAX232
Max 232 integrated circuit (IC) has 16 pins with ground, power supply and remaining input and output pins. Pin diagram of the IC is shown in below figure 2.5.
Figure 2.5 Pin Diagram of MAX 232(11) 2.7.3. Max 232 Logic diagram
2.7.4. MAX 232 pin description
MAX232 integrated circuit used in the present project has 16 pins; description of each and every pin of the IC is given in below table 2.5.
Table2.5. Pin description of MAX232 Integrated circuit (11)
2.8. RS 232 Serial port
Normally, modern computers are not equipped with a parallel port. Therefore the present system is designed to use the serial port. Although the serial port is also about to be obsolete, devices connected to the serial port can still be used since many relatively cheap USB‐RS232 interfaces exists in the market.
Figure 2.7.RS232 Serial Port functional diagram
Functional diagram of RS232 serial port is shown in above figure 2.7.and the advantages of using serial data transfer rather than parallel data transfer is given below (17).
Serial Cables can be longer than Parallel cables. The serial port transmits a ʹ1ʹ as ‐3 to ‐ 25 volts and a ʹ0ʹ as +3 to +25 volts where as a parallel port transmits a ʹ0ʹ as 0V and a ʹ1ʹ as 5V. Therefore the serial port can have a maximum swing of 50V compared to the parallel port which has a maximum swing of 5 Volts. Therefore cable loss is not going to be as much of a problem for serial cables as they are for parallel. Serial Transmission doesnʹt need as many wires as parallel transmission. If the device needs to be mounted a far distance away from the comput‐ er then 3 core cable are needed which is a lot cheaper that running 19 or 25 core cable. Microcontrollers have also proven to be quite popular recently. Many of these have in built SCI (Serial Communications Interfaces) which can be used to talk to the outside world. Serial Communication reduces the pin count of the micro processor units (MPUʹs) in which only two pins are commonly used, Transmit Data (TXD) and Receive data(RXD) (17).
2.8.1. Pin Description of DB9 Pin
RS232 serial port or DB9 pin used in this project has 9 pins; description of each and every pin of the pin is given in below table 2.6.
Pin 1 Received Line Signal Detector (Data Carrier Detect). Pin 2 Received Data.
Pin 3 Transmit Data.
Pin 4 Data Terminal Ready. Pin 5 Signal Ground. Pin 6 Data Set Ready. Pin 7 Request to Send. Pin 8 Clear to Send. Pin 9 Ring Indicator.
3 Methodology
Project method is derived from the aim of measuring the involuntary motion of human limbs or body joints especially hands, legs. Involuntary is nothing but unintentional movement due to imperfections in nerve cells, which are responsible for biomechanical feedback system.
Some of the examples of such nervous faults are of Parkinson’s tremor, essential tremor etc that are discussed in above sections. Acceleration of such tremor is measured using accelerometer in terms of frequency. In principle gravity is the fundamental unit of vibration, which is less in normal healthy people i.e. 15mg, where as ±1g in involuntary hand tremors. Table 3.1 below gives different characteristic frequencies of tremors depending on which clinical diagnosis follows (7).
S.N
o Tremor Type Frequency(Hz)
1 Normal Hand Tremor 9‐25 2 Essential Tremor 4‐12 3 Parkinson’s disease 3 – 8 4 Cerebellar Lesions 1.5 – 4 5 Physiological tremor 7‐12 Table3.1. Hand Tremors with varying frequencies (7)
The block diagram of the system is shown in figure 3.1 below. The system consists of the accelerometer sensor, the Micro controller, a micro controller program to acquire data and transfer it, a level converter to connect the micro controller to the PC via RS232 or DB9 connection. Post processing to display data and extract parameters of interest is carried out by Mat lab installed on the PC.
Figure 3.1 Block Diagram of the Tremor analysis system
The accelerometer in the system is fabricated by free scale semiconductors and the remaining parts of the signal conditioning system are made from commercially available components as specified earlier. According to the requirement of the goal to measure hand shake, specific parameters are selected and listed in table 3.2.
Sl.no Parameter range
1 Number of Axes 3 2 Frequency 0.1 ‐25 HZ 3 Maximum Acceleration +/‐ 5g 4 Maximum Acceleration without Damage +/‐ 25 g 5 Acceleration Resolution 0.001 g 6 Mass 5 grams 7 Size 7 mm³ Table3.2. Accelerometer sensor Parameters (8)
In order to reconstruct the total acceleration of the complex hand motion occurring in three dimensions, sensor must be active in three mutual perpendicular axes. Normal hand shake is of ±1g as discussed in above sections; accordingly required sensor has the range between 0.1g‐25g. In order to overcome effects due to higher accelerations, the physical sensor should have a gravity limit of ±50g. But a normal human hand has a very small involuntary acceleration for which a resolution of 1 mg is appropriate.
Accurate measurement of involuntary human hand motion needs specific physical requirements of the sensor. Such as for every gram of additional mass, peak frequency of finger tremor decreases by 0.85 Hz and hand tremor by 0.018 Hz. Sensor must be small enough to avoid the interference with normal finger motion i.e. touching an adjacent finger. So, sensor mass of 5 gm and volume of 7mm3 is reasonable (14).
3.1. Relevant work using accelerometer in the past
The use of accelerometer sensors in biomedical field is relatively new until the beginning of twenty first century. Usage of these sensors from the past is explained below (21).
In 1967, Gage used accelerometers to determine the vertical and horizontal accelerations of the trunk as well as the angular acceleration of the shank analyzing human gait.
In 1972, Prokop used accelerometers mounted in the shoe sole of spike shoes on various track surfaces.
In 1973, Morris used five accelerometers to quantify the three‐ dimensional movement of the shank assuming that the transverse rotations of the shank are small and can be neglected, and in same year Nigg used accelerometers mounted at the head, hip, and shank during alpine skiing
In 1974, Unold used accelerometers mounted at the head, hip, and shank during walking and running with different footwear on various
In 1977, Saha studied the effect of soft tissue in vibration tests using skin‐ mounted accelerometers.
During the year 1978, Chao proposed an experimental protocol for the quantification of joint kinematics using systems of accelerometers.
In 1979, Light measured skeletal accelerations at the tibia using bone mounted accelerometers and Ziegert studied the effect of soft tissue on skin mounted accelerometer measurements.
In 1980, Denoth used acceleration measurement in an effective mass model to determine bone‐to‐bone impact forces in the ankle and knee joint and he also used accelerometers to determine in vivo force deformation diagrams of human heels. In the same year Light compared results from bone and skin mounted accelerometers to find a loss of high frequency content and a phase shift for skin mounted accelerometers.
In 1983, Voloshin studied the shock (impact force) absorbing capacity of the leg using skin‐mounted accelerometers in conjunction with force plate measurements.
In 1987, Valiant estimated a magnification factor for skin‐mounted accelerometers using a linear spring damper model.
In 1991, Lafortune described the contribution of angular motion and gravity to tibial acceleration during walking and running.
3.1.1. Measurement of Involuntary Hand Motion in the past
Involuntary hand motion is a significant disability. It prevents many normal activities involving hand such as writing, eating, and drinking. People with severe tremor are socially embarrassed by their condition. Initially tremor starts to a single part of the body like a finger with infrequent and small amplitude and slowly spreads to entire limbs and even to head with large amplitude.
One of the most common neurological disorders is essential tremor; the patients of this movement disorder vary widely, ranging from 0.08 to 220 cases per 1000 people (14).
Research has been carried over on hand tremor analysis in order to find the cause of involuntary hand motion and also stopping it. The tremor is cured by symptomatic treatment. Earlier methods to analyze tremors quantitatively include electromyography (EMG) and accelerometer sensors.
EMG records the electrical action or contraction of muscles that causes tremor, where as accelerometer sensors capture the tremor acceleration. Correlation between the analysis of tremor EMG signals from the muscle and signals from accelerometer sensors shows very good response. As EMG recordings are generally considered as “gold standard” in tremor analysis, the good correlation of accelerometer sensors to EMG analysis indicates that these sensors are a simple and non‐invasive way to study tremor.
Many studies have been made to analyze the origins of hand tremor and how to identify its symptoms. Some of the studies are the effect of neurologically active chemicals on hand tremor and the effectiveness of pharmaceuticals targeted to decrease the amplitude of tremor (14). 3.1.2. Tracking of human limb movement using accelerometer by Morris
Earlier accelerometers were large and had poor resolution, though they were used for the study of vibrations. During the last quarter of twentieth century Morris is the first to use accelerometers for tracking the motion of a human limb. In particular he studied the movement of the shank or lower leg. Morris aim was to develop a system of measurement that is simple to operate and suitable for both experimental and clinical use (14).
Morris used more than three accelerometers to estimate the vibration of a shank. He did not attempt to measure transverse rotations of the shank as they require larger dimension of the platform in the plane normal to the perpendicular axis of accelerometer platform. As the rotations were considered to be small, they were assumed to zero.
His study took special care in minimizing the mechanical damping between the bone and the transducer. They did not attach the transducer directly to the bone, but they did choose a site that minimized the effects of soft tissue movements (14).
3.1.3. Measuring Chemical Effects on Involuntary Hand Tremor Nicotinic receptors are the pre‐ganglionic receptors in the parasympathetic nervous system as they are stimulated by nicotine in addition to the normal parasympathetic neurotransmitter acetylcholine. The parasympathetic and sympathetic nervous system are the two halves of autonomic nervous system, so increasing the concentration of a neurotransmitter such as nicotine affects only one‐half of the system and causes an imbalance in the nervous system.
Caffeine is a neurotransmitter that is normally injected through coffee and other caffeinated beverages. Caffeine stimulates many neuro receptors including adenosine receptors in the central nervous system. One common side effect of caffeine is shaking in hands or an increase in hand tremor. Research have been done quantitatively on correlating the hand tremor with caffeine intake, among which a study measured finger tremor in healthy people with an accelerometer sensor and found that 150 mg of caffeine i.e. equivalent to 3 cups of coffee consumed while fasting caused finger tremor to increase significantly, while the same amount of caffeine taken with a normal diet didn’t change the amplitude of tremor.
A very common symptom with alcohol or narcotics addiction withdrawal is hand tremor, which has been quantitatively analyzed with an accelerometer sensor. Hand tremor was measured for several weeks during verified abstinence. The results showed alcohol‐ dependent patients had a very high hand tremor while doing a pointing task i.e. trying to hold their hand steady but the tremor amplitude decreased with continued absence.
The cocaine‐dependent patients had abnormal hand tremor while resting which is very similar to Parkinson’s disease, but when doing a pointing task the tremor temporarily went away; the resting tremor did not improve with continued abstinence. The earlier results were interpreted as showing alcohol and cocaine affect different parts of the nervous system: alcohol temporarily affects the cerebellum, and cocaine more permanently affects the extra‐pyramidal nervous system.
Hand tremor is a common side effect of many pharmaceuticals. One study analyzed the hand tremor side effects with two drugs that are inhaled to improve lung function in asthmatic patients, they are salmeterol and salbutamol. The study showed that salbutamol significantly improved in lung function after just two minutes, while salmeterol took seven minutes for significant improvement. Similarly, hand tremor as measured with a linear accelerometer sensor had a much more rapid onset with salbutamol than salmeterol (14).
3.1.4. Measurement of Hand Acceleration using Accelerometer Accelerometer sensors have an important application in the measurement of voluntary and involuntary hand motion. Involuntary hand tremors are quite common and accelerometer sensors are frequently used for quantitative measurement of tremor in medical research.
In applications measuring either voluntary or involuntary hand movements, the ideal sensor should be as small and light as possible to avoid interfering with the motion. Although some movements are still measured using heavy sensors, the measurements are not an accurate representation of movement in the unencumbered system.
In involuntary movement, loading the hand increase the amplitude of tremor while decreasing the frequency. So, the sensor to measure hand motion has ideal parameters like transparency with negligible mass to the user, small size for not interfering with motion, and light wires with high flexibility.
In the present design, the sensor is small and light for measuring human hand tremor; the connecting wires are also flexible with very fine gauge. The sensor used in this project can measure both voluntary and involuntary hand motion as the requirements are so similar (14).
3.2. Types of accelerometer sensors
Accelerometers are of many types, they include strain gauge, piezoelectric, piezo‐resistive and capacitive transducers.
3.2.1. Strain Gauge Accelerometer
A strain gauge accelerometer consists of four strain‐sensitive wires attached to a cantilevered mass element that is mounted on a fixed base. The wires are connected to an electric Wheatstone bridge circuit. When the base is accelerated, the mass element causes a deformation due to its inertia. This deformation of the mass element causes a change of the strain in the wires changing their resistance and consequently changing the balance of the bridge circuit. The result is an electric output proportional to the acceleration of the base (4).
3.2.2. Piezoelectric Accelerometer
In Piezoelectric materials, a mechanical stress produces electrical charge or an applied electric field generates mechanical strain. Electric polarization occurs in the crystal due to the internal dipoles, as shown in Figure 3.2. The both ends of the piezoelectric material have a net charge from the dipoles, which is apparently neutralized by free charges available in the environment.
Mechanical strain applied to the crystal produces a linear deformation in the crystal structure, temporarily changing the surface charges until it is re‐neutralized. Thus the stress acting on the crystalline material is directly related to the acceleration produced (20).
Figure 3.2.Dipoles in a piezoelectric material.
Piezoelectric materials don’t need external energy; and the piezo electric accelerometer only measures the alternating acceleration. They have a high internal resistance, so the voltage across the ends of the material is inversely proportional to generated charges and directly related to force acting on it. The relation among them is shown below (14).
As the piezo materials measure alternating response, so they are not suitable for direct or DC acceleration. Piezoelectric MEMS accelerometer is made by depositing ferroelectric materials on silicon wafer, these sensors are used preferably for high accelerations and amplitudes like shocks but not for low acceleration applications as the charge created by continuous stress acting on it escapes (20).
3.2.3. Piezo-resistive Accelerometer
Piezo resistive materials are solid state resistors usually act as strain gauges whose resistance changes with the application of mechanical stress. MEMS based Piezo‐resistive accelerometers have a mass suspended by a spring, a cantilever beam with a proof mass, and a Piezo‐resistive patch. The conductance of the Piezo‐resistive material is directly proportional to the force, or the resistance is indirectly proportional to the applied force (14).
The Piezo‐resistive MEMS accelerometer is shown in Figure 3.3.when the device is exerted by acceleration, the gap between the mass and the bulk of the device changes. The voltage produced by the Piezo‐resistive patch and the bulk device due to external stress is proportional to the acceleration of the vibrating object (14).
Two types of Piezo‐resistive models are shown in Figure 3.4. In the Figure 3.4(a), the Piezo‐resistor is made by printing conductive traces on a resistive film. As the silicon substrate below the film changes, the resistance of the resistive film between the traces will also change. Microscopic wire strain gages, as diagrammed in Figure 3.4(b) are the other way to arrange Piezo‐resistors. Mechanical stress applied on this materials change the conductors form and hence the resistance of the Piezo‐resistors is a function of the width of the gap between the Piezo‐ resistors (14). (a) (b) Figure 3.4 (a) Diagram of a Piezo‐resistive layout using resistive film backing. (b) Diagram of a free‐standing Piezo‐resistive strain gage.
In order to measure the displacement of a proof mass in an accelerometer sensor one among these two techniques is used. In general, to increase sensitivity and reduce thermal variations more than one Piezo‐resistor is used with the same mass‐spring system (14).
3.2.4. Capacitive Accelerometer
In capacitive accelerometer, measurement of the capacitance between the moving proof mass and fixed object on the base of the sensor is made.
The displacement of the proof mass changes in relation to the base area, and the distance between the both sides of the capacitor changes. The mathematical equation of capacitance is an inverse function of the gap (d) and in relation with area (A) of the base which is shown below.
The capacitance is either made as single‐sided or differential pair. A differential capacitive pair is shown in Figure 3.5 with conductive proof mass, and a connection between the spring‐proof mass serves as the electrical connection to the floating capacitor plate.
Capacitive sensors are most commonly used in accelerometers fabrication. Single‐sided capacitive accelerometers can be designed with the perpendicular axis of the silicon die plane and the capacitor plates in the plane of the die, sensitivity of the capacitor increase by extending its area.
The non‐linearity between springs, capacitors is avoided by arranging the fixed mass position in differential capacitive accelerometer; this is another advantage of the sensors with electrical feed back system (4).
