Chapter 02: Health Monitoring Sensors

WIA-SENIOR-007 Senior Wearable Standard | Sensor Technology

The Sensor Foundation

At the heart of every senior wearable device lies a sophisticated array of sensors that continuously monitor vital signs, detect anomalies, and provide the raw data that powers health analytics. Unlike consumer fitness devices where approximate measurements suffice, senior health wearables demand clinical-grade accuracy and reliability. A missed cardiac event or false fall alarm can have life-or-death consequences, making sensor technology the most critical component of the WIA-SENIOR-007 standard.

This chapter explores the sensor platform in depth, covering the physics of each sensor type, age-specific challenges in signal acquisition, calibration techniques for older adult physiology, and the signal processing pipelines that transform noisy raw data into actionable health insights.

Core Sensor Types

WIA-SENIOR-007 defines a multi-modal sensor platform that combines optical, inertial, thermal, and electrical sensors to create a comprehensive picture of user health status.

Photoplethysmography (PPG) Sensors

PPG sensors form the workhorse of continuous health monitoring, providing non-invasive measurement of multiple cardiovascular parameters. The basic principle involves shining light (typically green or red LEDs) into the skin and measuring the amount reflected back. Blood absorbs more light than surrounding tissue, so changes in blood volume during the cardiac cycle create measurable variations in reflected light intensity.

Age-Specific PPG Challenges

Older adults present unique challenges for PPG measurement that require specialized signal processing and hardware design:

• Reduced Skin Perfusion: Aging reduces capillary density and blood flow to extremities, weakening the PPG signal. WIA-SENIOR-007 devices use higher LED intensity (within safety limits) and more sensitive photodiodes to compensate.
• Increased Skin Thickness Variation: Age-related changes create more variability in optical path length. Adaptive algorithms adjust LED intensity based on baseline signal quality.
• Motion Artifacts: Hand tremor and less stable device positioning create noise. Advanced motion cancellation using accelerometer data improves signal quality during movement.
• Arrhythmias: Irregular heartbeats are more common in seniors, requiring sophisticated peak detection algorithms that don't assume regular intervals.

PPG Metric	What It Measures	Clinical Accuracy Required	Senior-Specific Consideration
Heart Rate	Beats per minute (BPM)	±2 BPM or ±2%, whichever is greater	Handle irregular rhythms (AFib common in seniors)
Heart Rate Variability (HRV)	Beat-to-beat interval variation	±5ms RMSSD	Reduced HRV normal in aging; adjust baseline
SpO2 (Blood Oxygen)	Arterial oxygen saturation	±2% for SpO2 70-100%	Critical for COPD, heart failure monitoring
Blood Pressure (Estimated)	Systolic/Diastolic pressure	±5 mmHg systolic, ±3 mmHg diastolic	Requires individual calibration; arterial stiffness affects accuracy
Pulse Wave Velocity	Arterial stiffness indicator	±0.5 m/s	Increases with age; cardiovascular risk marker
Respiratory Rate	Breaths per minute	±1 breath/minute	Derived from PPG amplitude modulation

PPG Signal Processing Pipeline

// PPG Signal Processing for Senior Wearables
class SeniorPPGProcessor {
  constructor(config) {
    this.sampleRate = config.sampleRate || 125; // Hz
    this.ageAdjustment = config.userAge >= 65;
    this.motionCancellation = true;
  }

  // Step 1: Pre-processing and Noise Reduction
  preprocess(rawPPG, accelerometer) {
    // Band-pass filter: 0.5-5 Hz for heart rate
    let filtered = this.bandPassFilter(rawPPG, 0.5, 5.0);

    // Motion artifact removal using accelerometer
    if (this.motionCancellation) {
      const motionSignal = this.extractMotionArtifact(accelerometer);
      filtered = this.adaptiveFilter(filtered, motionSignal);
    }

    // Baseline wander removal (respiratory modulation)
    filtered = this.removeBaselineWander(filtered);

    return filtered;
  }

  // Step 2: Peak Detection with Arrhythmia Handling
  detectPeaks(ppgSignal) {
    const peaks = [];
    const threshold = this.calculateAdaptiveThreshold(ppgSignal);

    for (let i = 1; i < ppgSignal.length - 1; i++) {
      // Local maximum detection
      if (ppgSignal[i] > ppgSignal[i-1] &&
          ppgSignal[i] > ppgSignal[i+1] &&
          ppgSignal[i] > threshold) {

        // Age-specific validation
        if (this.ageAdjustment) {
          // Don't assume regular intervals for seniors
          if (this.isValidPeak(ppgSignal, i, peaks)) {
            peaks.push({
              index: i,
              timestamp: i / this.sampleRate,
              amplitude: ppgSignal[i]
            });
          }
        } else {
          peaks.push({
            index: i,
            timestamp: i / this.sampleRate,
            amplitude: ppgSignal[i]
          });
        }
      }
    }

    return peaks;
  }

  // Step 3: Heart Rate Calculation with AFib Detection
  calculateHeartRate(peaks) {
    if (peaks.length < 2) return null;

    const intervals = [];
    for (let i = 1; i < peaks.length; i++) {
      intervals.push(peaks[i].timestamp - peaks[i-1].timestamp);
    }

    // Check for irregular rhythm (AFib indicator)
    const irregularity = this.calculateIrregularity(intervals);
    const isIrregular = irregularity > 0.25; // 25% variation

    // Median-based heart rate (robust to outliers)
    const medianInterval = this.median(intervals);
    const heartRate = 60 / medianInterval;

    // HRV metrics
    const rmssd = this.calculateRMSSD(intervals);
    const sdnn = this.calculateSDNN(intervals);

    return {
      heartRate: heartRate,
      confidence: this.assessConfidence(peaks, intervals),
      irregularRhythm: isIrregular,
      irregularityIndex: irregularity,
      hrv: {
        rmssd: rmssd,
        sdnn: sdnn
      },
      alertFlags: this.checkAlertConditions(heartRate, isIrregular)
    };
  }

  // Step 4: SpO2 Calculation
  calculateSpO2(redPPG, infraredPPG) {
    // AC and DC components
    const redAC = this.calculateAC(redPPG);
    const redDC = this.calculateDC(redPPG);
    const irAC = this.calculateAC(infraredPPG);
    const irDC = this.calculateDC(infraredPPG);

    // Ratio of ratios
    const R = (redAC / redDC) / (irAC / irDC);

    // Empirical calibration curve (age-adjusted)
    let SpO2 = 110 - 25 * R;

    // Age-specific calibration offset
    if (this.ageAdjustment) {
      SpO2 = this.applySeniorCalibration(SpO2, R);
    }

    return {
      spO2: Math.round(SpO2),
      confidence: this.assessSpO2Confidence(redPPG, infraredPPG),
      perfusionIndex: (irAC / irDC) * 100
    };
  }

  // Alert condition checking
  checkAlertConditions(heartRate, isIrregular) {
    const alerts = [];

    if (heartRate < 50) {
      alerts.push({
        type: 'BRADYCARDIA',
        severity: 'MEDIUM',
        message: 'Heart rate below 50 BPM'
      });
    }

    if (heartRate > 100) {
      alerts.push({
        type: 'TACHYCARDIA',
        severity: 'MEDIUM',
        message: 'Heart rate above 100 BPM'
      });
    }

    if (isIrregular) {
      alerts.push({
        type: 'IRREGULAR_RHYTHM',
        severity: 'HIGH',
        message: 'Possible atrial fibrillation detected'
      });
    }

    return alerts;
  }
}

Inertial Measurement Unit (IMU)

The 3-axis accelerometer and 3-axis gyroscope form the inertial measurement unit that enables fall detection, activity recognition, gait analysis, and balance assessment. For senior wearables, IMU data is particularly critical for safety features.

Fall Detection Algorithm

Falls are the leading cause of injury-related death in adults over 65. Accurate fall detection requires sophisticated analysis of acceleration patterns to distinguish true falls from activities of daily living like sitting down quickly or bending over.

// Advanced Fall Detection for Seniors
class FallDetector {
  constructor() {
    // Thresholds calibrated for senior population
    this.impactThreshold = 2.5; // g (gravitational acceleration)
    this.postFallImmobilityTime = 30000; // ms
    this.preImpactWindow = 500; // ms
  }

  async processIMUData(accelerometer, gyroscope, timestamp) {
    const totalAcceleration = Math.sqrt(
      accelerometer.x ** 2 +
      accelerometer.y ** 2 +
      accelerometer.z ** 2
    ) / 9.81; // Convert to g

    // Phase 1: Pre-impact detection (free fall)
    const isFreeFall = totalAcceleration < 0.6; // Weightlessness

    if (isFreeFall) {
      await this.trackFreeFallDuration(timestamp);
    }

    // Phase 2: Impact detection
    if (totalAcceleration > this.impactThreshold) {
      const impactEvent = {
        timestamp: timestamp,
        magnitude: totalAcceleration,
        vector: accelerometer,
        rotation: gyroscope,
        hadFreeFall: this.recentFreeFall(timestamp)
      };

      // Phase 3: Post-impact analysis
      const postImpactData = await this.monitorPostImpact(
        impactEvent,
        this.postFallImmobilityTime
      );

      // Classification: True fall vs. ADL
      const isFall = this.classifyEvent(impactEvent, postImpactData);

      if (isFall) {
        return this.initiateFallResponse(impactEvent, postImpactData);
      }
    }

    return { fallDetected: false };
  }

  classifyEvent(impactEvent, postImpactData) {
    let fallScore = 0;

    // Factor 1: Pre-impact free fall (10 points)
    if (impactEvent.hadFreeFall) fallScore += 10;

    // Factor 2: Impact magnitude (0-20 points)
    fallScore += Math.min(20, (impactEvent.magnitude - 2.0) * 10);

    // Factor 3: Post-impact immobility (0-30 points)
    if (postImpactData.movementLevel < 0.1) {
      fallScore += 30;
    } else if (postImpactData.movementLevel < 0.3) {
      fallScore += 20;
    }

    // Factor 4: Orientation change (0-20 points)
    const orientationChange = this.calculateOrientationChange(
      impactEvent,
      postImpactData
    );
    if (orientationChange > 60) fallScore += 20; // Degrees

    // Factor 5: User position (0-20 points)
    if (postImpactData.finalPosition === 'PRONE' ||
        postImpactData.finalPosition === 'SUPINE') {
      fallScore += 20;
    }

    // Threshold: 60+ points = high confidence fall
    return fallScore >= 60;
  }
}

Gait Analysis & Balance Assessment

Beyond fall detection, continuous IMU monitoring enables assessment of gait patterns and balance that can predict fall risk before accidents occur. Changes in walking speed, stride length, and gait symmetry often precede falls by weeks or months.

Gait Parameter	Normal Range (Seniors)	Fall Risk Indicator	Detection Method
Walking Speed	0.8-1.2 m/s	<0.6 m/s indicates high fall risk	Accelerometer step detection + stride length estimation
Stride Length	1.2-1.5 meters	Decreasing trend over weeks	Double integration of acceleration during swing phase
Gait Symmetry	>90% left-right similarity	<85% suggests neurological/orthopedic issues	Comparison of left vs right step characteristics
Cadence	100-120 steps/minute	<90 steps/min associated with frailty	Step frequency from vertical acceleration peaks
Gait Variability	<5% coefficient of variation	>7% indicates instability	Standard deviation of stride times
Postural Sway	<2cm range while standing	>3cm suggests balance impairment	Integration of acceleration during quiet standing

Temperature Sensors

Continuous skin temperature monitoring serves multiple purposes in senior health monitoring: fever detection for infection screening, circadian rhythm tracking for sleep quality assessment, and environmental exposure monitoring.

Clinical Applications

• Early Infection Detection: Temperature elevation often precedes other symptoms of infection by 24-48 hours. For seniors with weakened immune systems, early antibiotic intervention can prevent serious complications.
• Fever Monitoring in Chronic Conditions: Patients with autoimmune diseases, cancer, or on immunosuppressive medications need continuous temperature monitoring to catch complications early.
• Hypothermia Prevention: Older adults are more susceptible to hypothermia. Alert systems trigger when skin temperature drops below safe thresholds, especially during winter months.
• Sleep Quality Assessment: Core body temperature follows a circadian rhythm that correlates with sleep quality. Disrupted temperature patterns indicate poor sleep, which affects overall health and fall risk.

Optional Advanced Sensors

ECG (Electrocardiogram)

While PPG provides good cardiac rhythm screening, medical-grade ECG offers definitive arrhythmia diagnosis. WIA-SENIOR-007 supports optional single-lead ECG via electrodes integrated into the device body and a touchpoint that the user activates with their opposite hand.

ECG is particularly valuable for seniors with known cardiac conditions like atrial fibrillation, where confirmation of rhythm irregularities is needed before adjusting anticoagulation therapy.

Bioimpedance Sensors

Bioimpedance spectroscopy measures the electrical impedance of body tissues, providing insights into:

• Hydration Status: Dehydration is common and dangerous in seniors, affecting medication levels and organ function.
• Body Composition: Muscle mass tracking helps detect sarcopenia (age-related muscle loss) that increases fall risk and metabolic issues.
• Edema Detection: Fluid accumulation in extremities indicates heart failure exacerbation, kidney problems, or venous insufficiency.