INSTRUMENTATION ESSENTIALS: DIFFERENTIAL AMPLIFICATION

Transcription

1 INSTRUMENTATION ESSENTIALS: DIFFERENTIAL AMPLIFICATION Daniel Dumitru, M.D., Ph.D. University of Texas Health Science Center San Antonio, Texas 1

2 ELECTRODE DESIGNATIONS E-1: Active Electrode (G-1) Located Over Active Tissue To Be Recorded E-2: Reference Electrode (G-2) Located At Some Distance From E-1 Ground: Sets The Relative Zero Value For Instrument AMPLIFIERS 2

3 AMPLIFIER DEFINITIONS Gain: Amplifier Output/Amplifier Input Input: 10 uv; Output: 10,000 uv = Gain 1000 Sensitivity: Ratio Of Amplifier Input To CRT Deflection 1 cm Deflection For 10 uv Input: 10uV/Div 3

4 DIFFERENTIAL AMPLIFICATION DIFFERENTIAL AMPLIFICATION Two Amplifiers With Essentially But Not Quite Identical Characteristics Positive Or Non-inverting Amplifier Amplifies But Does Not Alter Response Polarity Negative Or Inverting Amplifier Amplifies And Inverts Response Polarity The Two Inputs Are Electrically Summted Net Result Is Magnification Of Dissimilar Signals And Cancellation Of Similar Signals 4

5 DIFFERENTIAL AMPLIFICATION AMPLIFIERS The Biologic Signal Of Interest Is Split Between The Electrode And The Amplifier. What Is Not Lost At The Electrode Passes On To The Amplifier For Us To Observe Etotal = Eelec + Eamp The Current From The Signal Is The Same For Both The Electrode And Amplifier But The Voltage Drop Is Different For Each Depending On Their Respective Impedances (Ohms Law) 5

6 MATH X 2 SERIES CIRCUIT 6

7 AMPLIFIERS OHMS LAW: E = IR or E=IZ E= Voltage Of The Biologic/Environmental Signals I= Current Associated With The Signal Z= Impedance of Electrode/Amplifier Signal Voltage Forms Series Circuit With I & Z Etotal = Eelec + Eamp Eelec=I X Zelec Eamp=I X Zamp SERIES CIRCUIT 7

8 AMPLIFIERS Etotal = Eelec + Eamp Can Use Ohm s Law To Help Understand What The Instrument Is Doing And How That Can Help Us Solve Interference Issues AMPLIFIERS Etotal = Eelec + Eamp Etotal = I X Zelec + I X Zamp Etotal = I (Zelec + Zamp) We Are Interested In The Amplifier Eamp = I X Zamp; or I =Eamp/Zamp Etotal = Eamp/Zamp X (Zelec + Zamp) Eamp = (Etotal) (Zamp)/Zelec + Zamp 8

9 RELATIONSHIPS BETWEEN ELECTRODES/AMPLIFIERS E amp1 = E t ( Z amp1 ) Z elec1 + Z amp1 E amp2 = E t (Z amp2 ) Z elec2 + Z amp2 E amp1 E amp2 = E total RELATIONSHIPS BETWEEN ELECTRODES/AMPLIFIERS E amp1 = E t ( Z amp1 ) ; What If Z elec1 is tiny Z elec1 + Z amp1 E amp1 = E t (Z amp1 )/Z amp1 E amp1 = E t 9

10 RELATIONSHIPS BETWEEN ELECTRODES/AMPLIFIERS E amp1 = E t ( Z amp1 ) ; What If Z elect1 Is Big (= amp) Z elec1 + Z amp1 Z elec1 + Z amp1 = 2 Z amp1 E amp1 = E t (Z amp1 )/2 Z amp1 = Et/2 Cuts Our Signal In Half; False Positive Axonal Loss? RELATIONSHIPS BETWEEN ELECTRODES/AMPLIFIERS E amp1 = E t ( Z amp1/ Z elec1 + Z amp1 ) ; What If Z elect1 Is Big (= amp) Our Desired Signal Is Cut In Half But, What About The Common Signal If The Two Electrodes See Half Each Others Common Signal IT IS MULTIPLIED NOT SUBTRACTED OUT THROUGH DIFFERENTIAL AMPLIFICATION Result: Big 60Hz & Signal Is Lost In The Noise 10

11 RELATIONSHIPS BETWEEN ELECTRODES/AMPLIFIERS RELATIONSHIPS BETWEEN ELECTRODES/AMPLIFIERS E amp1 -E amp2 = E t ( Z amp1 ) - E t (Z amp2 ) Z elec1 + Z amp1 Z elec2 + Z amp2 Consider Common vs Difference Signal 11

12 DIFFERENCE SIGNAL CLEAN ELECTRODES E amp1 -E amp2 = E t ( Z amp1 ) - 0(Z amp2 ) Z elec1 + Z amp1 Z elec2 + Z amp2 E amp1 E t COMMON SIGNAL CLEAN ELECTRODES E amp1 -E amp2 = E t ( Z amp1 ) - E t (Z amp2 ) Z elec1 + Z amp1 Z elec2 + Z amp2 E t is =at E 1 and E 2 So E amp1 E amp2 = 0; No 60Hz 12

13 DIFFERENCE SIGNAL DIRTY ELECTRODES E amp1 -E amp2 = E t ( Z amp1 ) - 0(Z amp2 ) Z elec1 + Z amp1 Z elec2 + Z amp2 Z elec1 = Z amp1 E amp1 E amp2 = E t (1/2) 0 = E t /2 COMMON SIGNAL DIRTY ELECTRODES E amp1 -E amp2 = E t ( Z amp1 ) - E t (Z amp2 ) Z elec1 + Z amp1 Z elec2 + Z amp2 Z elec1 & Z elec2 Don t See 60Hz The Same Therefore, That Difference Is No Longer Eliminated But AMPLIFIED 13

14 THE NET RESULT The Signal Of Interest Is Cut In Half And The Common Signal Is No Longer Eliminated But Amplified Therefore: The Signal Is Lost Into 60 Hz SO, WHY NOT PUT REFERENCE ON THE BIG TOE? Maybe OK For Signal Of Interest BUT Not For The Common Signal WHY DO WE SEE INTERFERENCE Whenever There Is A Difference Between The Active And Reference Electrodes Recording The Common Signal Dirty Electrodes Broken Electrode Wire Generator Closer To One Electrode Than The Other 14

15 RELATIONSHIPS BETWEEN ELECTRODES/AMPLIFIERS Consider Common vs Difference Signal 60 HZ INTERFERENCE E amp1 -E amp2 = E t ( Z amp1 ) - E t (Z amp2 ) Z elec1 + Z amp1 Z elec2 + Z amp2 E amp1 -E amp2 {for common signal} O This Means E amp1 & E amp2 Are Not Identical Your Job To Get Rid Of 60Hz Is To Make E amp1 & E amp2 As Equal As Possible If You Do, Then 60 Hz Is Eliminated As A Common Mode Signal 15

16 60 HZ INTERFERENCE E amp1 -E amp2 = 0 {For The Common Signal} Start With Electrode Lead Inspection All Three Electrodes Connected: Active/Reference/Ground All Electrodes Connected To The Proper Ports Continuity: No Breaks, Frayed/Loose Wires or Connectors (Gently Tug On Connector Both Ends) All Metal Electrode Interfaces Clean: No Rust, Old Caked-On Paste, Pitted Surfaces If 60 Hz Persists Go Ahead And Change Electrodes 60 HZ INTERFERENCE E amp1 -E amp2 = 0 {For The Common Signal} If 60 Hz Persists Go Ahead And Change Electrodes If 60 Hz Still Persists: Move Pre-Amp As Close As Possible To The Pt Intertwine The Electrode Leads & Bunch Them Up Next To The Patient With The Pre-Amp On Top Look For A Nearby Generator Source 16

17 60 HZ INTERFERENCE E amp1 -E amp2 = 0 {For The Common Signal} If 60 Hz Persists Go Ahead And Change Electrodes If 60 Hz Still Persists: Move Pre-Amp As Close As Possible To The Pt Intertwine The Electrode Leads & Bunch Them Up Next To The Patient With The Pre-Amp On Top 60 HZ INTERFERENCE E amp1 -E amp2 = 0 {For The Common Signal} If 60 Hz Persists Go Ahead And Change Electrodes If 60 Hz Still Persists: Move Pre-Amp As Close As Possible To The Pt Intertwine The Electrode Leads & Bunch Them Up Next To The Patient With The Pre-Amp On Top 17

18 60 HZ INTERFERENCE E amp1 -E amp2 = 0 {For The Common Signal} If 60 Hz Persists Go Ahead And Change Electrodes If 60 Hz Still Persists: Move Pre-Amp As Close As Possible To The Pt Intertwine The Electrode Leads & Bunch Them Up Next To The Patient With The Pre-Amp On Top 60 HZ INTERFERENCE E amp1 -E amp2 = 0 {For The Common Signal} If 60 Hz Persists Go Ahead And Change Electrodes If 60 Hz Still Persists: Move Pre-Amp As Close As Possible To The Pt Intertwine The Electrode Leads & Bunch Them Up Next To The Patient With The Pre-Amp On Top 18

19 60 HZ INTERFERENCE E amp1 -E amp2 = 0 {For The Common Signal} If 60 Hz Persists Go Ahead And Change Electrodes If 60 Hz Still Persists: Move Pre-Amp As Close As Possible To The Pt Intertwine The Electrode Leads & Bunch Them Up Next To The Patient With The Pre-Amp On Top 60 HZ INTERFERENCE E amp1 -E amp2 = 0 {For The Common Signal} If 60 Hz Persists Start Turning Off Or UNPLUGGING Equipment Lights, Electric Appliances, Exam Table Heating Pads If In ICU: Bed, Calf Pump, Any Non-Essential Equip If 60 Hz Persists Touch The Patient With A Bit Of Paste On Finger NCV: Average Response With RR Not Divisible Into 60 (e.g. 3.3, 2.9, etc) 19

20 60 HZ INTERFERENCE E amp1 -E amp2 = 0 {For The Common Signal} If 60 Hz Persists Touch The Patient With A Bit Of Paste On Finger NCV: Average Response With RR Not Divisible Into 60 (e.g. 3.3, 2.9, etc) EMG: Use Two Monopolar Needles Instead Of Surface Reference (May Need Needle Ground Too) With Reference Close to Insertion Site But Not Too Close To Active Deeper In Tissue If Using Concentric: Switch To Two Monopolars 60 HZ INTERFERENCE E amp1 -E amp2 = 0 {For The Common Signal} If 60 Hz Persists If Using Concentric: Switch To Two Monopolars 20

21 60 HZ INTERFERENCE E amp1 -E amp2 = 0 {For The Common Signal} If 60 Hz Persists If Using Concentric: Switch To Two Monopolars 60 HZ INTERFERENCE E amp1 -E amp2 = 0 {For The Common Signal} Whenever 60 Hz Is Sufficient To Interfere With The Signal & The Ground Is Intact and Plugged In, Then The Active/Reference Electrodes Are Not Recording The Common Environmental Signals To The Same Degree Our Job Is To Do What It Takes To Make Them Record The Common Signals The Same 21

22 INTERELECTRODE SEPARATION AS IT PERTAINS TO DIFFERENTIAL AMPLIFICATION Why Not Put The Reference On The Big Toe And Forget About It? What Is The Deal With How Close Is Too Close And How Far Is Too Far INTERELECTRODE SEPARATION It Is Important To Consider The Separation Distance Between The Electrodes 22

23 SNAP INTERELECTRODE SEPARATION E-1/E-2: 4 cm Separation Maximize SNAP Amplitude Resolve SNAP Rise Time (~0.8 ms) NCV=D/T 50 M/s=D/0.8ms D=4 cm D<4cm SNAP Amp Not Maximized SNAP INTERELECTRODE SEPARATION E-/E-2: 4 cm Separation Maximize SNAP Amplitude Resolve SNAP Rise Time (~0.8 ms) 23

24 24

25 SNAP INTERELECTRODE SEPARATION When Using Bar Electrode For SNAPs Will Have A Smaller Amplitude Because Of Interelectrode Separation Think About Using Separate Disc Electrodes If Having Trouble Getting Response LFC Superfical Peroneal Sensory Sural Saphenous 25

26 PITFALL: AMPLITUDE 50-80% Amp Reduction; Normal Latency Axonal Loss/No Demyelination Pure Sensory Neuropathy/Ganglionopathy PITFALL: LATENCY 13% Latency Reduction; (Normal Latency) Miss Borderline Long Latency (False Neg) Could Miss A Borderline Neuropathy/CTS 26

27 INTERELECTRODE SEPARATION CMAP CMAP INTERELECTRODE SEPATION E-1: Muscle Motor Point E-2: Away From Muscle Tissue E-1/E-2 On Same Muscle Tissue Results In Reduced CMAP (Erroneously Conclude Axonal Loss, NMJ, or Myopathy) 27

28 28

29 PITFALL: AMPLITUDE ~50% Amplitude Reduction/Normal Latency Pure Axonal Loss Neurogenic/NMJ/Myopathy 29

30 FILTERS 30

31 FILTER A Device That Included Frequencies Of Interest But Excludes Unwanted Frequencies (Noise) FILTERS High Frequency: Low Pass Filter Low Frequency: High Pass Filter 31

32 FILTERS All Biologic Signals Can Be Conceptualized To Be Comprised Of Multiple Subcomponent Waveforms Of Varying Amplitude, Frequencies (High And Low), And Phases 32

33 LOW FREQUENCY FILTER Suppress Low Frequencies ( Hz) Affects Slowly Changing Portion Of Waveform 33

34 WHAT HAPPENS WHEN WE ELEVATE THE LOW FREQUENCY FILTERFROM 10 HZ TO 500 HZ? WAVEFORM COMPOSITION High Frequency Components: Fast Stuff Low Frequency Components: Slow Stuff When Elevate Low FF Affect Slow Stuff: YES Affect Fast Stuff: No So, What Happens To The Waveform s: Onset Lat, Peak Lat, Amp, Total Durat, Phases 34

35 ELEVATING LOW FREQUENCY FILTER Reduce Amplitude Peak Latency Shortens Negative Spike Duration Shortens Total Potential Duration Shortens Onset Latency Remains Unchanged Number Of Phases Increases 35

36 PITFALL: AMPLITUDE ~60% Amplitude Reduction/Normal Latency Sensory Axonal Neuropathy/Ganglionopathy LOW FREQUENCY FILTER EFFECTS CMAP 36

37 PITFALL: AMPLITUDE & PHASES ~80% Amplitude Reduction/Polyphasic Amp: Pure Axonal Neuropathy/Myopathy Phases: Asynchronous Demyelination GBS/Diabetes LOW FREQUENCY FILTER EFFECTS MUAP 37

38 PITFALL: AMPLITUDE & DURATION ~40% Amp Decline/50% Duration Shortening Could Create False + Myopathy HIGH FREQUENCY FILTER Suppresses High Frequencies ,000 Hz Affects The Fast Changing Portion Of The Waveform 38

39 WHAT HAPPENS WHEN YOU LOWER THE HIGH FREQUENCY FILTER? HIGH FREQUENCY FILTER Suppresses High Frequencies Affects Rapidly Changing Portions Of Waveform FAST STUFF VS SLOW STUFF Affects Fast Stuff Therefore What Would You Predict: Amp, Onset Latency, Negative Spike Duration, Total Waveform Duration 39

40 LOWERING HIGH FREQUENCY FILTER Reduces Waveform Amplitude Prolongs Peak Latency Prolongs Onset Latency Prolongs Negative Spike Duration Prolongs Total Waveform Duration Does Not Alter Number Of Phases 40

41 PITFALL: PEAK LATENCY ~15% Increase Peak Latency False Positive: Distal Focal/Generalized Neuropathy (CTS/Diabetes/etc.) STIMULATOR 41

42 NERVE ACTIVATION CATHODE (-) ANODE (+) CATHODE NEGATIVE POLE (-) ATTRACTS CATIONS (+) DEPOLARIZES THE NERVE SUPRAMAXIMAL DEPOLARIZATION 20% > LARGEST CMAP SURFACE OR NEEDLE ELECTRODE 42

43 POSITIVE POLE ANODE ATTRACTS ANIONS (-) CAN DEPOLARIZE THE NERVE ANODAL BLOCK DOES NOT OCCUR LARGE SURFACE DISPERSIVE ELECTRODE WHEN USED WITH A NEEDLE CATHODE STIMULATOR 43

44 CURRENT FLOW HOW CATHODE STIMULATES 44

45 HOW THE ANODE STIMULATES ANODAL BLOCK Theoretically: Anode Hyperpolarizes Nerve And Blocks Action Potentials ANODE BLOCK DOES NOT OCCUR Rather: We Routinely Can Get Anodal Stimulation 45

46 HOW THE ANODE STIMULATES ANODAL STIMULATION PITFALL Anodal Stimulation: Shorten Latency 46

47 ANODAL STIMULATION PITFALL Anodal Stimulation: Shorten Latency ANODAL STIMULATION PITFALL Anodal Stimulation: Shorten Latency 47

48 ANODAL STIMULATION PITFALL Anodal Stimulation: Shorten Latency/Biphasic ANODAL STIMULATION PITFALL Anodal Stimulation: Shorten Latency/Biphasic 48

49 ANODAL STIMULATION PITFALL Anodal Stimulation: Shorten Latency/Biphasic ANODAL STIMULATION PITFALL Anodal Stimulation: Shorten Latency/Biphasic 49

50 ANODAL STIMULATION PITFALL: BLINK REFLEX Anodal Stimulation: Create Unanticipated Responses Mimicking False Pathways ANODAL STIMULATION PITFALL: BLINK REFLEX 50

51 ANODAL STIMULATION PITFALL: BLINK REFLEX ANODAL STIMULATION PITFALL: BLINK REFLEX 51

52 ANODAL STIMULATION PITFALL: BLINK REFLEX ANODAL STIMULATION PITFALL: BLINK REFLEX 52

53 ANODAL STIMULATION PITFALL: BLINK REFLEX STIMULUS ARTIFACT 53

54 STIMULUS ARTIFACT Remove Excess Perspiration Remove Body Cream/Makeup Use Small Amount Electrolyte Paste Place Ground Next To E-1 Use Just Supramaximal Stimulus Reduce Skin Impedance Touch Patient Rotate Anode About Cathode Do Not Elevate LFF 54

55 STIMULUS ARTIFACT TOUCHING PATIENT 55

56 STIMULUS ARTIFACT MODULATION 56