9.2 Blood Pressure (NIBP/IBP Transducers) & SpO2 Pulse Oximetry
Key Takeaways
- Non-Invasive Blood Pressure (NIBP) monitors operate via the oscillometric principle (deflation at 2–3 mmHg/s); the peak oscillation amplitude directly identifies Mean Arterial Pressure (MAP), while systolic (50–60% peak) and diastolic (75–80% peak) values are mathematically derived via empirical envelope algorithms.
- AHA cuff sizing mandates a bladder width equal to 40% of mid-arm circumference and length equal to 80–100% of circumference; an undersized cuff produces falsely elevated blood pressure (by 10–30 mmHg), while an oversized cuff produces falsely low readings.
- Invasive Blood Pressure (IBP) utilizes a piezoresistive silicon Wheatstone bridge transducer with standard sensitivity of 5 µV/V/mmHg at 5 VDC excitation; hydrostatic zero leveling to the phlebostatic axis (4th ICS mid-axillary line) produces an error of 1.86 mmHg per inch (0.74 mmHg per cm) of vertical offset.
- IBP hydraulic frequency response requires an undamped natural frequency fn > 25 Hz and damping coefficient ζ of 0.5–0.7; underdamped systems exhibit catheter whip with falsely high systolic and low diastolic, while overdamped systems (bubbles/clots) cause falsely low systolic and high diastolic.
- Pulse oximetry (SpO2) applies the Beer-Lambert Law across Red (660 nm) and Infrared (940 nm) wavelengths, calculating the Ratio of Ratios (R = [AC_red/DC_red] / [AC_ir/DC_ir]); carboxyhemoglobin (COHb) falsely elevates SpO2 to near 100%, while methemoglobin (MetHb) fixes SpO2 readings near 85%.
Blood Pressure (NIBP/IBP Transducers) & SpO2 Pulse Oximetry
Hemodynamic monitoring provides real-time quantitative assessment of systemic arterial perfusion, cardiac output, and peripheral tissue oxygenation. A Biomedical Equipment Technician (CBET) must master the electro-mechanical, pneumatic, and optical physics governing non-invasive blood pressure (NIBP), invasive arterial blood pressure (IBP), and pulse oximetry ($SpO_2$).
1. Automated Non-Invasive Blood Pressure (NIBP): The Oscillometric Principle
Unlike manual auscultation (which detects acoustic Korotkoff sounds via stethoscope), automated NIBP modules employ the oscillometric measurement principle.
+-----------------------------------------------------------------------------+
| NIBP OSCILLOMETRIC MEASUREMENT CYCLE |
| |
| CUFF PRESSURE (mmHg) |
| 200 +-------------+ (Inflation to ~160-180 mmHg) |
| | / \ |
| 160 | / \ (Controlled Linear Bleed: 2-3 mmHg/sec) |
| | / \ |
| 120 | / \====[ SYSTOLIC POINT ] (Oscillations ~50-60% Max) |
| | / \ |
| 80 | / \========[ MAP POINT ] (MAXIMUM OSCILLATION PEAK) |
| | / \ |
| 40 | / \====[ DIASTOLIC POINT ] (Oscillations ~75-80%) |
| | / \ |
| 0 +---+-------------------+-------------------+-----------------------> |
| (Rapid Dump Valve Open) |
| PNEUMATIC MICRO-OSCILLATIONS (Filtered AC Component): |
| |
| - - - - - - - | | | | | | | | | | | | | | | | | - - - - - - - |
| | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | |
| ^ ^ |
| SYSTOLIC MAP |
| (Rising) (PEAK) |
+-----------------------------------------------------------------------------+
Detailed Measurement Sequence:
- Cuff Inflation: A DC diaphragm air pump inflates the non-distensible cuff around the patient's limb to a pressure above the anticipated systolic pressure ($160\text{--}180\text{ mmHg}$ for adults, $100\text{--}120\text{ mmHg}$ for neonates), completely occluding the underlying brachial artery.
- Controlled Linear Deflation: The monitor opens a micro-bleed proportional solenoid valve, releasing air at a steady rate of $2\text{ to }3\text{ mmHg/second}$ (or using stepwise discrete decrements of $4\text{--}8\text{ mmHg}$ per cardiac cycle).
- Oscillation Sensing: As cuff pressure drops below arterial systolic pressure, blood begins to jet through the partially compressed artery during systole. The pulsatile arterial expansion exerts minute cyclic pressure pulses against the cuff bladder. A high-sensitivity piezoresistive pressure transducer senses total cuff pressure, split into two channels:
- DC Channel (Low-Pass Filter $<0.1\text{ Hz}$): Measures overall static cuff deflation pressure.
- AC Channel (Bandpass Filter $0.5\text{--}5.0\text{ Hz}$, High Gain): Extracts the minute pressure oscillations ($0.5\text{ to }5.0\text{ mmHg}$ amplitude).
- Mean Arterial Pressure (MAP) Identification: As deflation progresses, oscillation amplitudes grow to a distinct global maximum before diminishing. The cuff pressure corresponding precisely to the point of Maximum Oscillation Amplitude is the Mean Arterial Pressure (MAP).
- Algorithmic Derivation of Systolic & Diastolic: Because MAP is the only directly measured physical point, Systolic Blood Pressure (SBP) and Diastolic Blood Pressure (DBP) are mathematically calculated using empirical manufacturer ratios:
- Systolic: Identified when rising oscillation amplitudes reach approximately $50%\text{ to }60%$ of the maximum peak amplitude.
- Diastolic: Identified when falling oscillation amplitudes decline to approximately $75%\text{ to }80%$ of the maximum peak amplitude.
+-----------------------------------------------------------------------------+
| NIBP PNEUMATIC HARDWARE SCHEMATIC |
| |
| +-------------------+ +--------------------+ +-------------+ |
| | DC MOTOR AIR PUMP | ----> | PROPORTIONAL BLEED | ----> | RAPID DUMP | |
| | (Inflation Driver)| | SOLENOID VALVE | | VALVE (NO) | |
| +-------------------+ +--------------------+ +------+------+ |
| | | | |
| +---------------------------+-------------------------+ |
| | |
| v (Pneumatic Hose Manifold) |
| +-------------------+ +--------------------+ +-------------+ |
| | HARDWARE OVER- | | PIEZORESISTIVE | | NIBP CUFF & | |
| | PRESSURE SWITCH | <---> | PRESSURE SENSOR | <---> | LIMB | |
| | (>300 mmHg Adult) | | (Strain Gauge) | | BLADDER | |
| +-------------------+ +--------------------+ +-------------+ |
+-----------------------------------------------------------------------------+
Pneumatic Safety Standards:
- Hardware Overpressure Cutoff: An independent analog pneumatic/electronic comparator switch forcibly dumps cuff pressure if it exceeds $>300\text{--}330\text{ mmHg}$ for adults or $>150\text{--}165\text{ mmHg}$ for neonates, independent of software CPU control.
- Maximum Cycle Timeout: A hardware watchdog timer vents the cuff if a single measurement cycle exceeds $120\text{ seconds}$ in adults or $60\text{ seconds}$ in neonates to prevent limb ischemia and nerve damage.
2. NIBP Cuff Sizing Standards & Clinical Artifacts
Proper cuff dimensioning is critical for diagnostic accuracy. The American Heart Association (AHA) defines strict geometric ratios relative to limb circumference:
+-----------------------------------------------------------------------------+
| AHA NIBP CUFF DIMENSIONING GUIDELINES |
| |
| <--------------------- BLADDER LENGTH -------------------------> |
| +--------------------------------------------------------------+ ^ |
| | | | |
| | BLADDER WIDTH = 40% OF ARM CIRCUMFERENCE | | WIDTH |
| | BLADDER LENGTH = 80% TO 100% OF ARM CIRCUMFERENCE | | |
| | | v |
| +--------------------------------------------------------------+ |
+-----------------------------------------------------------------------------+
| Cuff Size Category | Arm Circumference Range | Ideal Bladder Width ($40%$) | Ideal Bladder Length ($80%\text{--}100%$) |
|---|---|---|---|
| Infant | $8.0\text{--}14.0\text{ cm}$ | $3.5\text{ cm}$ | $10.0\text{ cm}$ |
| Child / Small Adult | $14.0\text{--}22.0\text{ cm}$ | $7.0\text{ cm}$ | $16.0\text{ cm}$ |
| Adult Standard | $22.0\text{--}32.0\text{ cm}$ | $12.0\text{ cm}$ | $24.0\text{ cm}$ |
| Large Adult | $32.0\text{--}42.0\text{ cm}$ | $15.0\text{ cm}$ | $32.0\text{ cm}$ |
| Thigh / Extra Large | $42.0\text{--}50.0\text{ cm}$ | $20.0\text{ cm}$ | $42.0\text{ cm}$ |
Clinical Sizing Errors & Mechanical Faults:
- "Cuff Too Small" (Under-cuffing): An undersized cuff fails to transmit pneumatic pressure efficiently through subcutaneous tissue to compress the artery. The pump must generate excessive pressure to occlude the vessel, causing a falsely high blood pressure reading (often elevating systolic by $10\text{ to }30\text{ mmHg}$).
- "Cuff Too Large" (Over-cuffing): An oversized cuff compresses a broader arterial zone than necessary, occluding the vessel at lower pressures and yielding a falsely low blood pressure reading (lowering systolic by $5\text{ to }15\text{ mmHg}$).
- Loose Wrapping: Loosely wrapped cuffs require extra air volume to snug against the skin, slowing cycle times and attenuating oscillation transmission, causing false hypertensive readings or pneumatic timeout errors.
- Pneumatic Leak Troubleshooting: BMET testing requires connecting the module to a calibrated digital pressure gauge and rigid metal calibration cylinder ($500\text{ mL}$). Pressurize to $250\text{ mmHg}$; the system must not leak more than $<6\text{ mmHg/minute}$ per AAMI/ANSI SP10.
3. Invasive Blood Pressure (IBP) Transducers & Wheatstone Bridge Physics
Invasive Blood Pressure (IBP) monitoring provides continuous, beat-to-beat arterial, central venous, or pulmonary artery pressure waveforms via an indwelling catheter coupled through a fluid-filled column to a piezoresistive transducer.
+-----------------------------------------------------------------------------+
| IBP PIEZORESISTIVE WHEATSTONE BRIDGE CIRCUIT |
| |
| +V_EXC (+5.000 VDC) |
| | |
| v |
| o |
| / \ |
| / \ |
| (Strain Gauge 1) / \ (Strain Gauge 2) |
| [ R + ΔR ] / \ [ R - ΔR ] |
| / \ |
| / Vm \ |
| -SIGNAL OUT o----o-------------o----o +SIGNAL OUT |
| (To Bioamp) \ / (To Bioamp) |
| \ / |
| (Strain Gauge 4)\ / (Strain Gauge 3) |
| [ R - ΔR ] \ / [ R + ΔR ] |
| \ / |
| \ / |
| o |
| ^ |
| | |
| -V_EXC (GROUND / -5 VDC) |
+-----------------------------------------------------------------------------+
Piezoresistive Transducer Specifications:
- Piezoresistive Effect: Silicon diaphragms micro-machined with four diffused piezoresistors. Applied fluid pressure deflects the diaphragm, subjecting opposite strain gauges to tensile stress ($R + \Delta R$) and compressive stress ($R - \Delta R$).
- Standard Sensitivity ($S$): The global medical standard (AAMI/ANSI BP22) is precisely $5\text{ }\mu\text{V per Volt of excitation per mmHg of pressure}$ ($5\text{ }\mu\text{V/V/mmHg}$):
- Example Calculation: With a standard excitation of $V_{exc} = 5.0\text{ VDC}$ and an arterial pressure $P = 120\text{ mmHg}$:
- Continuous Flush Device: A continuous micro-infusion valve delivers $3\text{ mL/hour}$ of heparinized saline under $300\text{ mmHg}$ pressure bag compression to prevent thrombus formation at the catheter tip without affecting pressure transmission.
4. IBP Hydrostatic Leveling & Zero Calibration
Accurate IBP measurement requires establishing an exact atmospheric pressure reference (zeroing) and eliminating hydrostatic fluid column pressure offsets (leveling).
+-----------------------------------------------------------------------------+
| IBP HYDROSTATIC LEVELING PHYSICS |
| |
| |
| TRANSDUCER PLACED TOO HIGH |
| (Above Phlebostatic Axis) PATIENT PHLEBOSTATIC AXIS |
| +--------------------+ (4th Intercostal Space, Mid-Axillary Line)|
| | Transducer at +5" | o (Right Atrium Reference) |
| +--------------------+ /| |
| ^ / | |
| | Δh = +5 inches / | |
| | / | |
| =====================================+================================= |
| / | |
| / | Δh = -5 inches |
| / | |
| v v |
| +--------------------+ |
| | Transducer at -5" | |
| +--------------------+ |
| TRANSDUCER PLACED TOO LOW |
| (Below Phlebostatic Axis) |
+-----------------------------------------------------------------------------+
Hydrostatic Pressure Derivation & Rules:
- Phlebostatic Axis: The anatomical reference level for the right atrium, located at the intersection of the 4th intercostal space (ICS) and the mid-axillary line (halfway between anterior and posterior chest surfaces).
- Hydrostatic Fluid Pressure Formula:
- Converting saline/water column density to mercury ($\text{Hg}$):
- Clinical Directional Errors:
- Transducer Positioned ABOVE the Phlebostatic Axis: Hydrostatic weight of the saline column pulls downward away from the transducer, creating negative gauge pressure. Readings are FALSELY LOW by $1.86\text{ mmHg per inch}$ ($0.74\text{ mmHg/cm}$).
- Transducer Positioned BELOW the Phlebostatic Axis: Weight of the fluid column adds hydrostatic pressure to the sensor diaphragm. Readings are FALSELY HIGH by $1.86\text{ mmHg per inch}$ ($0.74\text{ mmHg/cm}$).
- Zero Calibration Procedure: Open the transducer stopcock to atmospheric air ($0\text{ mmHg}$ gauge pressure), depress the monitor's "Zero IBP" softkey to nullify ambient barometric pressure and sensor electrical offset voltage, then close the stopcock to air and reopen to the patient catheter.
5. IBP Dynamic Frequency Response & The Fast Flush Test
The fluid-filled catheter-tubing-transducer system acts as a mechanical second-order underdamped harmonic oscillator. The fidelity of the displayed pressure waveform depends on the system's natural resonant frequency ($f_n$) and damping coefficient ($\zeta$, zeta).
+-----------------------------------------------------------------------------+
| IBP SQUARE-WAVE FAST FLUSH TEST |
| |
| NORMAL DAMPED SYSTEM (ζ ≈ 0.6 - 0.7, fn > 25 Hz) |
| +300 mmHg +------------+ |
| | Fast Flush | |
| | Snap Pull | |
| + +---+ |
| | 1-2 Ring Oscillations |
| +--\ /\_ |
| \/ +================ (Normal Arterial) |
| |
| UNDERDAMPED SYSTEM (ζ < 0.4: Whip/Resonance -> Falsely High Systolic) |
| +300 mmHg +------------+ |
| | Fast Flush | /\ |
| | Snap Pull | / \ Multiple Sharp Ring Oscillations |
| + +-+ \ /\ /\_ |
| \/ \/ \/=========== (Exaggerated Syst) |
| |
| OVERDAMPED SYSTEM (ζ > 0.8: Clot/Bubble/Soft Line -> Falsely Low Systolic)|
| +300 mmHg +------------+ |
| | Fast Flush | |
| | Snap Pull | |
| + +--\ |
| \____ |
| +==================== (Sluggish Upstroke) |
+-----------------------------------------------------------------------------+
Dynamic Response Parameters:
- Natural Resonant Frequency ($f_n$): Arterial pressure waves have fundamental heart rate frequencies ($1\text{--}2\text{ Hz}$) and contain diagnostic harmonic components up to the 10th harmonic ($10\text{--}20\text{ Hz}$). To prevent resonance, the physical tubing-transducer system must have a natural frequency $f_n > 25\text{ Hz}$ (ideally $>40\text{ Hz}$). Non-compliant (stiff) pressure tubing, short tube lengths ($<48\text{ inches}$), and minimal stopcocks maximize $f_n$.
- Damping Coefficient ($\zeta$): Represents viscous energy dissipation in the fluid column. Optimal clinical damping is $\zeta = 0.5\text{ to }0.7$.
- Underdamped System ($\zeta < 0.4$): Characterized by "catheter whip" and acoustic resonance. Fast-rising systolic wavefronts bounce against the diaphragm, generating multiple sharp ringing oscillations on the fast flush test. Result: Falsely elevated systolic pressure ($+15\text{ to }+30\text{ mmHg}$), falsely low diastolic pressure, and prominent dicrotic notch ringing. (MAP remains accurate).
- Overdamped System ($\zeta > 0.8$): Caused by micro-air bubbles in the transducer dome, fibrin/blood clots at the catheter tip, loose stopcocks, compliant (soft/compliant) IV extension tubing, or kinking. The flush test exhibits a sluggish return to baseline with zero ringing oscillations. Result: Falsely low systolic pressure, falsely high diastolic pressure, loss of the dicrotic notch, and a flattened, blunted waveform.
6. Pulse Oximetry ($SpO_2$) & The Beer-Lambert Law
Pulse oximetry provides non-invasive, continuous measurement of arterial hemoglobin oxygen saturation ($SpO_2$) using photoplethysmography (PPG) and dual-wavelength spectrophotometry.
+-----------------------------------------------------------------------------+
| BEER-LAMBERT TRANSMITTANCE MODEL |
| |
| INCIDENT LIGHT (I_0) |
| (Red 660 nm & Infrared 940 nm LEDs) |
| | |
| v |
| +------------------------------------+ |
| | NON-PULSATILE TISSUE (DC) | (Skin, Bone, Muscle) |
| +------------------------------------+ |
| | VENOUS BLOOD (DC) | (Non-pulsatile deox blood) |
| +------------------------------------+ |
| | PULSATILE ARTERIAL BED (AC) | (Expands with Systolic Pulse) |
| | Path Length = Δd | |
| +------------------------------------+ |
| | |
| v |
| TRANSMITTED LIGHT (I) |
| (To Silicon PIN Photodiode) |
+-----------------------------------------------------------------------------+
The Physics of Optical Absorption:
- Beer-Lambert Law: Total light absorbance ($A$) through an absorbing medium is proportional to the substance's extinction coefficient ($\epsilon$), concentration ($c$), and optical path length ($d$):
- Pulsatile (AC) vs. Non-Pulsatile (DC) Decomposition:
- DC Component: Constant baseline optical attenuation from skin pigmentation, subcutaneous adipose, bone, venous blood, and non-pulsatile arterial blood.
- AC Component: Minute cyclical fluctuation in optical absorbance ($1%\text{ to }5%$ of total light) caused by pulsatile arterial expansion during ventricular systole.
7. Dual-Wavelength Spectrophotometry & The Ratio of Ratios ($R$)
Arterial blood contains two primary hemoglobin variants: Oxygenated Hemoglobin ($HbO_2$) and Deoxygenated / Reduced Hemoglobin ($Hb$).
+-----------------------------------------------------------------------------+
| HEMOGLOBIN OPTICAL ABSORPTION SPECTRUM (600 - 1000 nm) |
| |
| EXTINCTION |
| COEFFICIENT (ε) |
| ^ |
| | (Hb High Absorption at 660 nm) |
| | \ |
| | \ Deoxygenated Hb |
| | \_ / |
| | \_ / (HbO2 High Absorption at 940 nm) |
| | \ / / |
| | X (ISOBESTIC POINT: 805 nm) |
| | / \ / |
| | Oxygenated \_ _/ |
| | HbO2 \________________/ |
| | |
| +--------------+-------------------+---+----------------------------> |
| 660 805 940 WAVELENGTH |
| (RED) (ISO)(IR) (nm) |
+-----------------------------------------------------------------------------+
Optical Wavelength Physics:
- Red LED ($660\text{ nm}$): Deoxygenated hemoglobin ($Hb$) absorbs approximately 10 times more light at $660\text{ nm}$ than oxygenated hemoglobin ($HbO_2$).
- Infrared (IR) LED ($940\text{ nm}$): Oxygenated hemoglobin ($HbO_2$) absorbs substantially more light at $940\text{ nm}$ than reduced hemoglobin ($Hb$).
- Isobestic Point ($805\text{ nm}$): The wavelength where the absorption coefficients of $Hb$ and $HbO_2$ are precisely identical (used in specialized co-oximeters).
The Mathematical "Ratio of Ratios" ($R$):
The pulse oximeter normalizes the pulsatile AC signal by the baseline DC signal for each respective wavelength, forming the parameter $R$:
+-----------------------------------------------------------------------------+
| EMPIRICAL CALIBRATION CURVE (R vs SpO2) |
| |
| R RATIO VALUE --> CALCULATED SpO2 PERCENTAGE |
| R = 0.4 --> 100% SpO2 (Pure HbO2) |
| R = 1.0 --> 85% SpO2 (Equal AC modulation at 660 & 940 nm) |
| R = 2.0 --> 0% SpO2 (Pure Deoxygenated Hb) |
+-----------------------------------------------------------------------------+
[!NOTE] Empirical Calibration Table: Pulse oximeters cannot calculate $SpO_2$ from pure theoretical physics because light scattering in living tissue violates idealized Beer-Lambert assumptions. Instead, manufacturers map calculated $R$ values to an internal empirical lookup table generated from human volunteer blood co-oximetry studies ($70%\text{ to }100%\text{ }SaO_2$).
8. Pulse Oximetry Limitations, Artifacts & Dyshemoglobins
Biomedical engineers must diagnose clinical and technical pulse oximetry failure modes:
+-----------------------------------------------------------------------------+
| DYSHEMOGLOBIN ABSORPTION INTERFERENCE |
| |
| 1. CARBOXYHEMOGLOBIN (COHb) - Carbon Monoxide Poisoning |
| - Absorption profile at 660 nm is virtually identical to HbO2. |
| - Pulse oximeter counts COHb as HbO2, calculating R ≈ 0.4. |
| - CLINICAL ERROR: Falsely NORMAL or HIGH SpO2 (e.g., 99%) despite |
| severe, lethal cellular tissue hypoxia. |
| |
| 2. METHEMOGLOBIN (MetHb) - Fe2+ Oxidized to Fe3+ (Benzocaine / Dapsone) |
| - Absorbs equally strongly at BOTH 660 nm (Red) and 940 nm (IR). |
| - Forces AC/DC ratios at both wavelengths to equalize (R -> 1.00). |
| - CLINICAL ERROR: Locks the displayed SpO2 reading near 85%, regardless|
| of whether true arterial saturation is 50% or 100%. |
| |
| 3. INTRAVENOUS DYES (Methylene Blue, Indocyanine Green) |
| - Methylene blue exhibits massive optical absorption peak at 668 nm. |
| - Mimics a massive surge in deoxygenated Hb; triggers acute SpO2 drop |
| to 60-70% lasting several minutes post-injection. |
+-----------------------------------------------------------------------------+
Additional Operational Artifacts:
- Low Perfusion Index ($PI < 0.2%$): Peripheral vasoconstriction, severe hypothermia, or hypovolemic shock diminishes arterial pulsation amplitude ($AC$), reducing the optical signal-to-noise ratio and triggering "Low Perfusion" or "Searching for Pulse" errors.
- Optical Shunting (Light Bypass): When a finger sensor is misaligned or too large, light from the LEDs passes directly to the photodiode without traversing tissue, creating an artificial high DC current that skews $R$ toward $1.0$ ($85%$).
- Ambient Light Interference: Strong OR surgical lamps, fluorescent ballasts, or direct sunlight saturate the PIN photodiode. Standard modules alternate LED firing ($660\text{ nm ON} \rightarrow \text{OFF} \rightarrow 940\text{ nm ON} \rightarrow \text{OFF}$) hundreds of times per second, subtracting the "OFF" ambient background level from the active readings.
A biomedical technician is evaluating an invasive blood pressure (IBP) setup in the ICU. The patient is elevated in bed, and the transducer is positioned 5 inches (12.7 cm) higher than the patient's phlebostatic axis. What measurement error will this height mismatch produce?
A firefighter rescued from a burning structure arrives in the Emergency Department with suspected severe carbon monoxide poisoning. The bedside pulse oximeter indicates an SpO2 reading of 99%. Why is this SpO2 value dangerously misleading?
An adult patient with an arm circumference of 40 cm has an NIBP measurement taken using a standard Adult cuff (designed for arm circumferences 22–32 cm). What impact will this undersized cuff have on the measured blood pressure values?
During a square-wave fast flush test on a radial artery arterial line, the monitor displays an initial sharp pressure step followed by 4–5 continuous rapid ringing oscillations before returning to the arterial waveform. The systolic pressure reads 185 mmHg, while an auscultatory cuff reads 150 mmHg. What hydraulic condition exists in the IBP system?