Objective Pain Assessment with PPG: Measuring Pain from Photoplethysmography Signals

Pain assessment remains one of medicine's most persistent challenges: it is inherently subjective, difficult to communicate for many patients, and impossible to self-report for those who are anesthetized, sedated, cognitively impaired, or preverbal. The quest for objective pain measurement has led researchers and clinicians to photoplethysmography, which captures autonomic nervous system responses to nociceptive stimuli through changes in heart rate, heart rate variability, and peripheral vascular tone. PPG-derived pain indices are now commercially available in surgical monitoring systems and under active investigation for broader clinical application.

This article examines the physiological basis of PPG-based pain assessment, the established clinical indices (SPI, ANI), emerging machine learning approaches, validation evidence, clinical utility, and the fundamental limitations inherent in attempting to objectively measure a subjective experience. For background on PPG technology, see our introduction to photoplethysmography.

The Physiology of Pain and Autonomic Responses

Nociception versus Pain

A critical distinction underpins all work in this field: nociception is not pain. Nociception is the neural process of detecting and encoding noxious (tissue-damaging or potentially damaging) stimuli through specialized sensory neurons called nociceptors. Pain is the subjective, conscious experience that may or may not result from nociception. Pain involves cortical processing, emotional evaluation, memory, and cognitive interpretation.

This distinction matters because PPG measures peripheral autonomic responses that correlate with nociceptive input, not with the subjective experience of pain. An anesthetized patient receiving a surgical incision exhibits nociceptive autonomic responses (tachycardia, vasoconstriction) without experiencing pain. Conversely, a patient with chronic neuropathic pain may experience severe pain with minimal autonomic activation due to autonomic habituation.

Autonomic Responses to Nociception

Acute nociceptive stimulation triggers a sympathetic stress response through spinal and brainstem reflex pathways:

Cardiovascular responses:

• Heart rate increase of 5-20 bpm (via sympathetic cardiac acceleration and vagal withdrawal)
• Blood pressure increase of 10-30 mmHg (via sympathetic vasoconstriction and increased cardiac output)
• Peripheral vasoconstriction (sympathetically mediated reduction in skin and muscle blood flow)
• Decreased heart rate variability (parasympathetic withdrawal)

PPG-detectable changes:

• Decreased pulse wave amplitude (PPA reduction of 15-50% during noxious stimulation)
• Increased heart rate (shortened inter-beat intervals)
• Reduced HRV (decreased RMSSD, HF power)
• Altered pulse waveform morphology (steeper systolic rise, reduced dicrotic notch)
• Decreased pulse transit time (reflecting blood pressure increase)

These responses are mediated by the same sympatho-adrenal pathway involved in the general stress response, which is why PPG-based nociception indices overlap substantially with stress indices. For a detailed discussion of PPG-based stress detection, see our article on PPG stress detection methods.

Modulation by Analgesia and Anesthesia

Analgesic drugs (opioids, regional anesthetics, non-opioid analgesics) reduce nociceptive autonomic responses in a dose-dependent manner. This creates the clinical rationale for PPG-based nociception monitoring during anesthesia: if the PPG signal shows significant autonomic activation during surgery, the current analgesia may be insufficient.

Opioids attenuate nociceptive autonomic responses through multiple mechanisms: direct suppression of spinal nociceptive transmission, inhibition of brainstem sympathetic outflow, and vagal enhancement. At adequate analgesic doses, the PPG pulse wave amplitude remains stable and heart rate variability is maintained during surgical stimulation. At inadequate doses, noxious stimulation produces the vasoconstriction and tachycardia described above.

Established PPG-Based Pain/Nociception Indices

Surgical Pleth Index (SPI)

The Surgical Pleth Index (SPI), formerly known as the Surgical Stress Index (SSI), was developed by GE Healthcare and introduced by Huiku et al. (2007) (DOI: 10.1097/ALN.0b013e3180a76a69). SPI is calculated from two PPG-derived parameters obtained from finger pulse oximetry:

Components:

• Heartbeat interval (HBI): The pulse-to-pulse interval derived from the PPG waveform, inversely related to heart rate.
• Photoplethysmographic pulse wave amplitude (PPGA): The AC amplitude of the PPG signal, reflecting peripheral vascular tone.

Both components are normalized to a 0-100 scale using historical data, and SPI is computed as:

SPI = 100 - (0.33 x HBI_norm + 0.67 x PPGA_norm)

The weighting emphasizes PPGA because pulse wave amplitude is more specifically affected by nociceptive sympathetic activation than heart rate, which responds to many non-nociceptive stimuli.

Interpretation: SPI ranges from 0 to 100:

• SPI 0-20: Deep analgesia, minimal sympathetic activation
• SPI 20-50: Adequate analgesia for most surgical stimulation
• SPI 50-80: Possible inadequate analgesia, consider supplemental dosing
• SPI 80-100: Strong nociceptive response, likely inadequate analgesia

Validation evidence: Gruenewald et al. (2009) (DOI: 10.1093/bja/aen374) demonstrated that SPI detected noxious stimulation (skin incision, intubation) under general anesthesia with an AUC of 0.84 (n = 30, sevoflurane-remifentanil anesthesia). SPI predicted hemodynamic responses to surgical stimulation significantly better than heart rate alone (AUC 0.67) or blood pressure alone (AUC 0.71).

In a randomized controlled trial (n = 120), Bergmann et al. (2013) (DOI: 10.1093/bja/aet018) compared SPI-guided analgesia management with standard practice. SPI-guided patients received 30% less remifentanil, had fewer hemodynamic events requiring intervention (12% vs. 28%, p = 0.03), and had shorter extubation times (8.2 vs. 11.5 minutes, p = 0.01) with similar pain scores in recovery.

Analgesia Nociception Index (ANI)

The Analgesia Nociception Index (ANI), developed by MDoloris Medical Systems, is derived from the parasympathetic component of PPG or ECG heart rate variability (Jeanne et al., 2012; DOI: 10.1093/bja/aes034). ANI quantifies the respiratory influence on heart rate, which reflects parasympathetic (vagal) tone:

Principle: During adequate analgesia, parasympathetic tone is preserved and respiratory sinus arrhythmia (RSA) produces clear HF modulation of the heart rate. During nociceptive stimulation with inadequate analgesia, sympathetic activation suppresses parasympathetic tone, reducing RSA amplitude.

Calculation: ANI analyzes the HF component (0.15-0.50 Hz) of the RR interval series over a moving 64-second window, normalizing it to a 0-100 scale:

• ANI 50-100: Adequate analgesia, preserved parasympathetic tone
• ANI 30-50: Borderline, possible nociception
• ANI 0-30: Inadequate analgesia, significant nociceptive response

Validation: Boselli et al. (2013) (DOI: 10.1093/bja/aet017) evaluated ANI in 128 surgical patients and found that pre-incision ANI below 50 predicted hemodynamic reactivity to skin incision with sensitivity of 85% and specificity of 73%. ANI decreased from a mean of 72 (pre-incision) to 48 (post-incision) during inadequate analgesia (p < 0.001).

Ledowski et al. (2014) (DOI: 10.1093/bja/aeu300) compared SPI and ANI head-to-head in 48 patients and found comparable performance for detecting noxious stimulation (SPI AUC = 0.81, ANI AUC = 0.78, p = 0.42 for comparison).

Newborn Infant Parasympathetic Evaluation (NIPE)

A specialized application of the ANI algorithm, the NIPE index (MDoloris Medical Systems) is designed for neonatal pain assessment. Neonates cannot self-report pain, and behavioral pain scales (NIPS, PIPP) have significant inter-observer variability.

Faye et al. (2010) demonstrated that NIPE correlated inversely with behavioral pain scores in premature neonates undergoing painful procedures (heel lance, venipuncture), with NIPE values below 60 associated with PIPP scores indicating moderate-to-severe pain (r = -0.58, p < 0.001, n = 42 neonates).

Machine Learning Approaches to PPG-Based Pain Assessment

Feature Engineering for Pain Classification

Beyond SPI and ANI, researchers have explored richer feature sets for PPG-based pain classification:

Study example: Werner et al. (2014) (DOI: 10.1109/TAFFC.2014.2332031) used 43 PPG-derived features for pain intensity classification in 90 healthy volunteers receiving calibrated heat pain stimuli (four intensity levels). Features included:

• Time-domain HRV metrics (12 features)
• Frequency-domain HRV metrics (8 features)
• Pulse wave morphology features (15 features)
• Nonlinear dynamics features (8 features)

Using a random forest classifier with leave-one-subject-out cross-validation, they achieved:

• Binary classification (pain vs. no pain): Accuracy 76.4%, AUC 0.82
• Four-level classification (no pain, low, medium, high): Accuracy 52.3%
• Baseline vs. high pain: Accuracy 83.1%, AUC 0.89

The most discriminative features were pulse wave amplitude (information gain = 0.31), RMSSD (0.28), mean heart rate (0.26), and systolic upslope steepness (0.22).

Deep Learning on Raw PPG

Recent work has applied deep learning directly to raw PPG waveforms, bypassing manual feature engineering:

Thiam et al. (2020) trained a 1D CNN on 5-second raw PPG segments from the BioVid Heat Pain Database (87 subjects, four pain levels). Their architecture (4 convolutional layers, global average pooling, 2 dense layers) achieved 72.8% accuracy for binary pain classification and 48.1% for four-level classification in a leave-one-subject-out evaluation. While accuracy was modest, the model learned interpretable features corresponding to pulse amplitude changes and inter-beat interval modifications.

Lopez-Martinez and Picard (2018) (DOI: 10.1109/TAFFC.2017.2763882) explored personalized deep learning models for pain detection, finding that fine-tuning a pretrained CNN with as little as 2 minutes of individual calibration data improved binary pain classification from 71% to 82% accuracy.

Multi-Modal Pain Assessment

PPG alone captures only the autonomic component of pain. Combining PPG with other modalities improves pain assessment:

• PPG + EDA (electrodermal activity): EDA captures sympathetic cholinergic sudomotor responses that are independent of the sympathetic adrenergic responses captured by PPG. Combining both modalities improved binary pain classification from 76% (PPG alone) to 84% (PPG + EDA) in Werner et al. (2014).
• PPG + EMG (electromyography): Facial EMG captures pain-related muscle activation (corrugator, zygomaticus) that is mechanistically distinct from autonomic responses.
• PPG + facial video: Computer vision-based facial action unit detection combined with PPG provides complementary behavioral and physiological pain information.

Clinical Applications and Evidence

Intraoperative Nociception Monitoring

The primary clinical application of PPG-based pain assessment is guiding analgesic administration during general anesthesia. The clinical goal is maintaining adequate analgesia without overdosing, which causes hypotension, delayed recovery, and respiratory depression.

A meta-analysis by Ledowski (2019) (DOI: 10.1007/s12325-019-00903-0) reviewed 15 randomized controlled trials (total n = 1,847 patients) comparing nociception index-guided analgesia with standard practice. Key findings:

• Nociception-guided groups received 18% less intraoperative opioid (standardized mean difference -0.42, 95% CI: -0.62 to -0.22)
• Hemodynamic events (hypertension/tachycardia episodes) were reduced by 35% (RR 0.65, 95% CI: 0.48-0.88)
• Recovery room pain scores were not significantly different (SMD -0.08, 95% CI: -0.28 to 0.12)
• Time to extubation was reduced by 2.4 minutes (95% CI: 0.8-4.0 minutes)

These results suggest that PPG-based nociception monitoring can optimize opioid use without compromising pain control, though the clinical significance of the effect sizes is debated.

ICU Pain Assessment

Critically ill patients in the ICU are frequently unable to self-report pain due to sedation, mechanical ventilation, or cognitive impairment. Current practice relies on behavioral pain scales (BPS, CPOT) that require trained nurses and provide only intermittent assessment.

Continuous PPG-based nociception monitoring could supplement behavioral assessment in the ICU. Ben-Israel et al. (2013) (DOI: 10.1007/s10877-013-9418-9) evaluated a PPG-derived Nociception Level (NoL) index in 72 ICU patients and found correlations of r = 0.56 with the CPOT behavioral pain scale. Sensitivity for detecting clinical pain events (procedural pain during suctioning, repositioning) was 78% with specificity of 65%.

Neonatal Pain Assessment

Neonates, particularly premature infants, experience numerous painful procedures during NICU stay (heel lances, IV insertions, intubation) and may be inadequately treated due to difficulty assessing their pain. PPG-based indices offer continuous, objective supplementation of behavioral assessment.

Cremillieux et al. (2018) (DOI: 10.1097/AJP.0000000000000571) evaluated the NIPE index in 50 premature neonates (28-36 weeks gestational age) during heel lance and found that NIPE decreased from 68 plus or minus 12 (pre-procedure) to 42 plus or minus 15 (during procedure), with AUC of 0.82 for detecting procedurally-induced pain compared to the PIPP-R behavioral scale.

Chronic Pain Monitoring

PPG-based pain assessment is least validated for chronic pain, where the relationship between nociception and autonomic activation is weakest. Chronic pain patients often show:

• Autonomic habituation (blunted autonomic responses to ongoing pain)
• Central sensitization (pain without peripheral nociceptive input)
• Autonomic dysfunction (baseline autonomic abnormalities confounding interpretation)
• Medication effects (opioids, gabapentinoids, antidepressants alter autonomic function)

Limited evidence suggests that PPG-derived HRV may serve as a biomarker for pain-related autonomic burden in chronic pain populations. Tracy et al. (2018) found reduced resting HRV (RMSSD: 22 ms vs. 38 ms, p < 0.01) in chronic low back pain patients compared to pain-free controls (n = 56 per group), but this likely reflects general autonomic dysregulation rather than pain intensity per se.

For information on autonomic assessment via PPG, see our article on PPG autonomic function testing.

Limitations and Ethical Considerations

Fundamental Measurement Limitations

PPG-based pain indices face several irreducible limitations:

Specificity: Autonomic responses to pain are not distinguishable from autonomic responses to anxiety, stress, fever, hypovolemia, bladder distension, or many other stimuli. Any factor activating the sympathetic nervous system will alter SPI, ANI, and similar indices regardless of whether pain is present.

Sensitivity to medications: Beta-blockers, alpha-agonists (clonidine, dexmedetomidine), anticholinergics (atropine, glycopyrrolate), and vasopressors directly affect the cardiovascular parameters that PPG-based pain indices measure, independently of nociception.

Individual variability: Pain sensitivity, autonomic reactivity, and analgesic requirements vary enormously between individuals. Population-level calibrations provide only rough estimates for any individual patient.

Chronic pain insensitivity: PPG indices primarily capture acute nociceptive responses and have limited sensitivity to chronic pain states where autonomic habituation has occurred.

Ethical Considerations

The pursuit of objective pain measurement raises important ethical questions. Pain is fundamentally a subjective experience, and patients' self-reports should remain the primary basis for pain management decisions when available. PPG-based indices should supplement, not replace, patient communication.

There is a risk that objective pain indices could be used to discredit patient pain reports (if the index reads low, is the patient's complaint dismissed?). Clinical guidelines should ensure that pain indices are used to detect under-reported pain (in non-communicative patients) rather than to override self-reported pain.

Our conditions page provides additional context on how various health conditions affect pain processing and PPG signals.

Future Directions

Multi-Wavelength Pain Assessment

Different PPG wavelengths probe different tissue depths and vascular beds. Multi-wavelength PPG arrays could simultaneously capture superficial cutaneous vasoconstriction and deeper muscular vascular responses to pain, potentially improving specificity. For background on wavelength selection, see our article on green vs. red vs. infrared PPG.

Integration with Other Biomarkers

Combining PPG-derived autonomic pain indices with other objective biomarkers, including pupillometry (pupil dilation reflex), skin conductance (sudomotor response), and EEG-based pain signatures, could create multi-modal pain assessment systems with improved accuracy and specificity.

Personalized Pain Models

Individual calibration using brief standardized pain stimuli (pressure algometry, electrical stimulation) could establish person-specific relationships between nociceptive input and PPG response, enabling personalized pain scoring that accounts for individual autonomic reactivity. This personalized approach, combined with continuous learning from clinical outcomes, represents the most promising path toward clinically useful objective pain assessment through PPG-based health monitoring algorithms.

Conclusion

PPG-based pain assessment provides a window into the autonomic component of nociception, capturing the sympathetic activation and parasympathetic withdrawal that accompany acute noxious stimulation. Established indices like SPI and ANI have demonstrated clinical utility in guiding intraoperative analgesia, reducing opioid consumption, and supplementing pain assessment in non-communicative patients. However, fundamental limitations in specificity, sensitivity to confounders, and individual variability prevent PPG from serving as a standalone pain measurement tool. The technology is most valuable when understood for what it truly measures: not pain itself, but the body's autonomic response to nociceptive input, a physiological correlate that, combined with clinical context and patient communication, can improve pain management decisions.