Biomagnetism - UW-Madison Department of Physicsuw.physics.wisc.edu/~rzchowski/phy208/HonorsLectures/RWakaiBiomag.pdfBiomagnetism Ron Wakai Professor Department of Medical Physics Magnetism

1

Biomagnetism

Ron WakaiProfessor

Department of Medical Physics

Magnetism in Medicine: A bad beginning

Franz Mesmer (1734-1815)• “discoverer” of “animal

magnetism”

Some current “applications” of magnetism:

Current Medical Applications of Magnetism

Magnetic resonance imaging (MRI)

Transcranial magnetic stimulation (TMS)

Magnetoencephalography (MEG)

Nerve and muscle cells act like tiny

batteries that drive ionic currents

• current dipole, Q (current source)

• volume current, JV

! magnetic field, B

Ionic currents produce electric and magnetic signals:• electrocardiogram (ECG) and magnetocardiogram (MCG)

• electroencephalogram (EEG) and magnetoencephalogram (MEG)

**MEG provides improved source localization

Origin of Surface Electric vs. Magnetic Signals

Extracellular current (volume current) surface electric potentials• topography is distorted by inhomogenous conductivity

Intracellular current (primary current) surface magnetic fields• topography of magnetic signals is distorted much less, allowing accurate source localization

Neuromagnetism

• Dendritic activity gives rise to EEG/MEG• Neurons organized in columnar arrangement• Need >1000 neurons to get detectable signal

• Focal activity produces dipolar spatial pattern

• Source lies below phase inversion

2

! geofield

! typical urban magnetic noise

! magnetic field of magnetic contaminants in lung

! magnetocardiogram

! fetal magnetocardiogram

! spontaneous magnetoencephalogram

! evoked magnetic brain responses

Instrumentation Requirements

1. extremely sensitivity detector: SQUID(superconducting quantum interference device)

2. noise rejection: magnetically shielded room (high-magnetic permeability alloy, aluminum)

Whole-head system148 channels

Dual-system system74 channels

SQUID (Superconducting Quantum Interference Device)

Superconducting loopinterrupted by two “weak-links” (Josephson tunnel

junctions)

Current-voltage characteristic• current at zero voltage (supercurrent, Ic)

Modulation of I-V Characteristic by Magnetic Signal

•supercurrent show interferencepattern

Ehrenberg-Siday-Aharonov-Bohm Effect

SQUID1. particle beam divides

2. paths enclose region of flux

3. vector potential, A, alters phase

(B= ∇xA)

phase advancesphase retards

3

Current-Biased SQUID: Periodic V vs. Φ

High Magnetic Permeability Shielding

mu-metal

Eddy-Current Shielding

Approximate weight: 7,000 kgApproximate cost: $350,000

Magnetically ShieldedRoom

time-varying magneticinterference

eddy-currents

Essential for low frequency shielding

Highly cost-effective

Magnetic source imaging (MSI)= MEG+MRI

whole-head mapping ofauditory evoked response

isofield contourmap

dipole overlay

Nakasato et al., 1995

Statement of problem: To compute the distributionof brain currents based on measurement ofexternal magnetic field

**Has no unique solution**

Modeling Assumptions• Dipole approximation: signal due to collection of

current dipoles—localized current elements—thatlie on convoluted cortical surface

• Homogeneous sphere model: model head ashomogenous conducting sphere

Inverse Problem

Positron Emission Tomography (PET) Functional Magnetic Resonance Imaging (fMRI)Functional brain imaging techniques

spatial temporal brainmodality resolution resolution region cost

1. positron emission 1 cm minutes whole $3Mtomography (PET) brain

2. functional magnetic 1 cm seconds whole $2Mresonance imaging (fMRI) brain

3. electric source imaging >1cm msec cortex $100k

4. magnetic source imaging 1 cm msec cortex $2M(MEG + MRI)

MSI: records cortical activity directly and with high temporal resolution, butsuffers from an ill-posed inverse problem

4

Clinical applications

• presurgical functional mapping– somatosensory, auditory– language

• epilepsy– localization of epileptogenic foci– most useful for neocortical epilepsy

approved for clinical use in 2002

Magnetic Localization of Somatosensory Cortex

sites of mechanical stimulation

sites of cortical response

Nakamura et al., Neuroimage 7(4 Pt 1):377 (1998)

Presurgical mapping of central sulcus

Central sulcus, containing somatosensory cortex, wasat A, not B, allowing surgical resection of tumor

Orrison et al., Perspect Neurol Surg 4:141 (1993)

Language Mapping: Response to Visually Presented Words

Simos et al., J Clin Exp Neuropsych 1998; 5:706

language activity sensory response

Sensory andlanguage responsesare temporallydistinct

Validated viaintraoperativemapping

Magnetoencephalographic mapping of theMagnetoencephalographic mapping of thelanguage-specific cortex.language-specific cortex.Papanicolaou, SimosPapanicolaou, Simos et al. J et al. J NeurosciNeurosci 1997 1997

Language Lateralization

NRNL

NRNLLI

+

!=

Picture Naming

Salmelin et al., Nature 368:463-8 (1994)

5

Application of MEG to Epilepsy

• Incidence of about 1%• 10-20% of epilepsy is intractable to drug therapy

Presurgical evaluation• Focal or generalized?• Number and relative timing of foci (primary vs.

secondary focus)• Location with respect to eloquent cortex

Mirror focus

Primary focus

Transmission of electric and magnetic fetal signals

primary currents→ fMCG,fMEGvolume currents → fECG,fEEG

•fECG is attenuated strongly by thepoor conductivity of vernix caseosa

•fMCG is affected much less

Fetal Magnetocardiography (fMCG)

6

FMCG FROM 17-38 WEEKS FROM SAME SUBJECT Information Content of fMCG

RR interval fetal heart rate (FHR), FHR variabilityQRS amplitude variations fetal (trunk) movementP-, QRS-, T-wave timing and morphology fetal rhythm

•R

RR

QRSamp

PR

QT

QRS

Fetal Heart Rate (FHR) Monitoring

• Primary clinical method• Specific for hypoxia

Reassuring FHR patterns• reactivity: FHR accelerations associated with movement• normal FHR variability

Ominous FHR patterns• decelerations that are late or variable with respect to

uterine contractions• absent variability

Movement-related Signal Amplitude Variation:fMCG Actography

• fetal trunk movement results in prominent movement artifact• instantaneous QRS amplitude serves as actogram tracing

Assessing fetal activity from the fMCG:fMCG Actocardiography

• 2 pumps: left (systemic), right (pulmonary)• 4 chambers: 2 atria, 2 ventricles

valves at inlet and outlet of ventricles

The Heart

7

Unique Electrical Properties of Cardiac Tissue

!threshold

1. rhythmic self-excitation - pacemaker tissue

3. continuous conducting surface- myocardial syncytium

2. long refractory period- 250 ms (vs. 2 ms for neurons)

repolarizationdepolarization

refractory period

Mechanisms of Arrhythmia

1. Abnormal impulse formation– ectopy– altered automaticity

2. Abnormal impulse conduction

– accessory connection

– conduction block

– reentry

Ventricular Fibrillation (V-fib)chaotic, ineffective beating (heart quivers)

• most common immediate cause of sudden death• due to abnormal automaticity and/or “reentry” (impulsepropagates through same tissue more than once)

Initiation of V-fib• QT dispersion (inhomogeneous repolarization)• impulses divide around patches of refractory tissue

ECG during transition from normalsinus rhythm to V-tach to V-fib

• ~1-2% incidence, but most are benign• ~10% of these are serious, sustained arrhythmias

Sustained fetal arrhythmia• rare, but high mortality and morbidity• most common types:

1) sustained tachyarrhythmia, usually SVT2) complete congenital heart block (CCHB)

Fetal Arrhythmia

Drug Therapy for Fetal Arrhythmia

• Aggressive drug therapy is used often

• Adverse side effects: toxic, proarrhythmic

Accurate diagnosis and follow-up are critical, especiallywhen drug therapy is used

Rationale for fMCG:• MCG/ECG is gold standard for rhythm assessment

• Ultrasound is less precise and indirect for rhythm assessment,but can also assess heart function and blood flow

Echo/Doppler Ultrasound

fMCG

echo/Doppler

8

145 146 147 148 149 150

-2500

-1500

-500

500

1500

2500

P

R

P TP

R

P

T

P

R

PT

time (sec)

amplitude (fT) 115 116 117 118 119 120

-2000

-1000

0

1000

time (sec)

amplitude (fT)

Long QT Syndrome (LQTS)

fMCG diagnosed LQTS and torsade due to proarrhythmic drug effect

polymorphic ventriculartachycardia: torsade depointes

• prolonged QT interval

• 2nd-degree AV block

• LQTS: prolonged repolarization, undetectable with ultrasound

P-wave: atria depolarization

QRS complex: ventricular depolarization

T-wave: ventricular repolarization

Depolarization of Heart

• atria and ventricles are isolated except via the AV node

Atrioventricular Block• impulse is slowed or blocked at AV node• second most common serious fetal arrhythmia

60% maternal lupus antibodies, 40% structural disease

264.0 265.0 266.0 267.0 268.0

CCAVB

congenital complete AV block• SA node paces atria• AV node paces ventricles at much lower rate• atrial and ventricular rhythms are dissociated

P-wave

Supraventricular Tachycardia (SVT)

109 110 111 112 113 114

Initiation of supraventricular tachycardia:

• “accessory” pathway (extra connection) exists between atria andventricles• most common serious fetal arrhythmia

circusmovement

Initiation and Termination of SVT:Association with Fetal Movement

300 400 500 600

-2000

0

2000

4000

0

50

100

150

200

250

300

350subject 2

time (sec)

fetal QRS amplitude (fT)

fetal heart rate (bpm)

0 100 200 300

0

2000

4000

0

50

100

150

200

250

time (sec)

fetal QRS amplitude (fT)

fetal heart rate (bpm)

subject 3

SVT SVTSVT

SVT

quiescence activity

quiescence

activity

0 100 200 300 400 500 600

50

100

150

atrial

ventricular

time (sec)

fetal heart rate (bpm)

Subject 2: GA= 33 wks, ventricular rate= 75 bpm

Heart Rate Patterns in Congenital AV Block

0 100 200 300

50

100

150

ventricular

atrial

time (sec)

fetal heart rate (bpm)

High ventricular rate (>60 bpm)atrial and ventricular rates arereactive and highly correlated

Low ventricular rate (< 55 bpm):ventricular tracings become flat

Subject 3: GA= 28 wks, ventricular rate= 52 bpm

9

Subject 3: GA= 25-4/7 weeks; mean ventricular rate= 71 bpm

Tachycardia in Congenital AV Block

Subject 4: GA= 34 weeks; ventricular rate= 50 bpmbigeminy tachycardia

Subject 4: GA= 34 weeks; ventricular rate= 50 bpm

• 20% show tachycardia; generally not detected by ultrasound

• predictive of need for with neonatal pacing