Pupillography is a technique for recording and measuring pupillary responses. It uses a combination of infrared video cameras and computer software to dynamically and quantitatively evaluate pupillary reactions.
The term “Pupillography” was coined by Lowenstein and Loewenfeld and developed as a dynamic infrared video technique for recording the efferent and afferent pathways of the eye. The introduction of electronic technology has improved accuracy, consistency, and speed compared to traditional manual measurements, and the term “Pupillometry” is now also commonly used.
At the 32nd International Pupil Colloquium (IPC, Morges, Switzerland), experts gathered to establish the first edition of international standards for data collection, processing, and reporting. In recent years, desktop and portable pupillometers have been commercialized, and further performance improvements are expected through combination with AI.
Application fields include ophthalmology, neurology, neuroscience, psychology, and chronobiology.
Originally, Lowenstein and Loewenfeld named the dynamic infrared video technique “Pupillography.” Later, with the development of electronic technology, the term “Pupillometry” also became widespread, and today the two are often used almost synonymously.
The following are representative parameters obtained in a single measurement.
Constriction Phase
Latency (T₁): Time from light stimulus to onset of constriction. Approximately 200 ms with sufficiently bright stimuli.
Constriction velocity (VC): Speed of constriction. Visual input impairment leads to prolonged T₁ and decreased VC.
Maximum constriction amplitude (D₃): Diameter at maximum constriction. Decreased in visual input impairment.
Constriction ratio (CR): Constriction amount / baseline diameter. Decreased in visual input impairment.
Dilation Phase
Dilation velocity (VD): Speed of pupil dilation after light removal. Reflects sympathetic nervous system function.
Mydriasis delay time (T₅): Time to onset of mydriasis. Marked prolongation of T₅ is characteristic of Horner syndrome.
PIPR (Post-illumination Pupillary Response): Sustained miosis after high-intensity, short-wavelength stimulation. Reflects activation of ipRGC and melanopsin.
The pupil is never completely still; it continuously oscillates about ±0.5 mm (pupillary unrest/hippus). The constriction phase mainly reflects parasympathetic function, while the dilation phase mainly reflects sympathetic function.
Retinal photoreceptors → retinal ganglion cells → optic nerve → optic chiasm → optic tract → branches off from the visual pathway just before the lateral geniculate body → pretectal area → ipsilateral Edinger-Westphal (EW) nucleus and contralateral EW nucleus via the posterior commissure.
In humans, the ratio of crossed to uncrossed fibers is approximately 1:1, so the direct and consensual light reflexes are nearly equal in strength.
EW nucleus → oculomotor nerve → cavernous sinus → superior orbital fissure → orbit → inferior branch of oculomotor nerve → synapse at ciliary ganglion → short ciliary nerves → into the eyeball.
95% of parasympathetic fibers from the EW nucleus project to the ciliary muscle (accommodation), and 5% to the sphincter pupillae. This ratio is related to the mechanism of light-near dissociation.
Intrinsically photosensitive retinal ganglion cells (ipRGCs) contain melanopsin and form the main afferent pathway for the pupillary light reflex. Strong short-wavelength blue light stimulation (around 470 nm) induces a slow, sustained miosis (PIPR).
Melanopsin function remains relatively stable throughout life, declining after the 80s, but slower than age-related changes in rods and cones.
Yes, it does. Mental tension, startle, and pain stimuli cause dilation via the sympathetic nervous system, while fatigue and drowsiness induce miosis via the central nervous system. Emotionally, fear is associated with dilation and comfort with miosis. Drugs (caffeine, nicotine, antihistamines) also affect pupil diameter.
Perform in an environment with low ambient light to minimize external influences on pupil diameter. The patient sits in front of a dedicated device, with the camera and both eyes aligned. After recording baseline pupil diameter, various stimuli (light, visual patterns) are presented to record pupillary responses.
Considering diurnal variation, measurements should be started and completed between 10:00 AM and 2:00 PM (avoid within 1 hour after lunch). In closed-type devices, the pupil diameter is larger than in open-type devices, and monocular vision results in a pupil diameter about 1.0 mm larger in bright conditions and about 0.2 mm larger in dark conditions compared to binocular vision. Therefore, standardization of conditions is essential.
Segmentation algorithms are used to track pupil size. The simplest Hough Transform method is effective when the pupil is circular, but accuracy is insufficient for misalignment or abnormal shapes. Using high-precision pattern recognition models enables highly accurate detection of abnormally shaped pupils.
Pupillography enables accurate measurement through standardized recording, allowing quantitative assessment of additional parameters such as latency, constriction amplitude, and dilation velocity that cannot be evaluated with a penlight. It also allows documentation of results, longitudinal comparison, and elimination of examiner bias.
Compared to the conventional swinging flashlight test, standardized, quantitative, and reproducible measurements are possible. RAPD appears in a wide range of retinopathies, optic neuropathies, optic tract lesions, and pretectal lesions.
The procedure for the swinging flashlight test involves alternately stimulating the left and right eyes with light for about 2 seconds in a dimly lit room. In unilateral optic nerve damage, both eyes constrict when the healthy eye is stimulated, but the constriction decreases when the light is moved to the affected eye (observed as a “dilating change”). RAPD does not become positive in conditions such as cataract (media opacity), bilateral visual dysfunction, or lesions posterior to the optic chiasm.
In patients with head trauma, it may be the only sign of traumatic optic nerve injury, and is useful for evaluating trauma patients in a comatose state.
Dilation lag is a delay in pupillary relaxation and dilation after light offset. In Horner syndrome, dilation takes up to 15–20 seconds (normal: about 5 seconds). Marked prolongation of T₅ is a characteristic finding. In bilateral Horner syndrome (where relative anisocoria is unclear), measurement of dilation lag may be the only reliable diagnostic method.
Evaluation of anisocoria requires measurement under both light and dark conditions. Miosis on the affected side becomes clearer in the dark, and mydriasis on the affected side becomes clearer in the light. Approximately 20% of normal individuals have physiologic anisocoria (difference ≤1 mm, no difference between light and dark, normal light reflex). However, it should be noted that afferent defects (optic nerve diseases) cause abnormal light reflex but原则上 do not cause anisocoria.
Pupil diameter measurement under low illumination determines the optimal ablation diameter for LASIK and other procedures. If the ablation diameter is smaller than the dilated pupil diameter, there is a risk of nighttime halos. It is also used to assess indications for multifocal intraocular lenses and refractive surgery.
PLR allows individual assessment of rod, cone, and melanopsin pathways. In a family with Jalili syndrome (CNNM4 mutation), cone-mediated PLR was recordable despite undetectable photopic electroretinography 1). Under scotopic conditions, CNNM4 patients showed large PLR, while under photopic conditions, it was near the lower limit of normal or slightly reduced 1).
Typical uses include differentiation of optic neuropathy and retinopathy by RAPD assessment, diagnosis of Horner syndrome by measurement of dilation lag, and evaluation of Adie syndrome in combination with dilute pilocarpine instillation test. It is also used for objective assessment of neurodegenerative diseases and autonomic disorders.
The pupillary light reflex is controlled by integration of the following three pathways.
| Pathway | Photoreceptor | Characteristics |
|---|---|---|
| Rod pathway | Rhodopsin | High sensitivity under low luminance and dark conditions |
| Cone pathway | Opsin | High luminance and photopic conditions. Short latency and fast constriction velocity |
| ipRGC pathway | Melanopsin | Long latency and slow speed. Sustained after stimulation (PIPR) |
Cone input controls sustained constriction to stimulus contrast variations, while melanopsin input sets the light-adapted pupil diameter during prolonged light exposure. This integration of outer and inner retinal signals enables precise modulation of the pupillary response.
Gain control of the pupillary control pathway exists at the level of the Edinger-Westphal nucleus. Regarding cortical input, even with localized ischemic infarction of the insular cortex or frontal eye field, pupil diameter and constriction velocity remain within normal physiological ranges, suggesting that absence of cortical input does not directly affect pupil diameter or constriction velocity.
Pupillary changes induced by cognitive or emotional events are smaller than the light reflex (usually less than 0.5 mm) and are known to be closely correlated with locus coeruleus activity.
This technique aims to separate outer retinal (rod- and cone-mediated) and inner retinal (melanopsin-mediated) responses in a single non-invasive measurement. With technological optimization, it is expected to develop into a highly sensitive and accurate clinical biomarker.
Hyde et al. (2022) compared three sisters with Jalili syndrome (CNNM4 mutation, aged 5, 14, and 15 years) with 10 normal controls. They performed rod pathway measurements after dark adaptation (465 nm, 1 second) and cone pathway measurements after light adaptation (642 nm, 1 second, with a 6 cd/m² blue rod-suppressing field). The Naka-Rushton function was used to calculate Pmax (maximum saturated PLR response) and s (PLR half-saturation constant) 1). Although the light-adapted electroretinogram was undetectable, cone-mediated PLR could be recorded, suggesting that PLR may be a useful tool for assessing cone function 1).
The first edition of the international standard established at the 32nd International Pupil Colloquium includes recommendations on the minimum set of variables required for data collection, processing, and reporting. It aims to improve comparability between studies and is expected to serve as a foundation for future clinical and multicenter research.
Integration of AI and devices is expected to improve performance and objective quantification. Key research topics include detection and progression monitoring of neurodegenerative diseases (Alzheimer’s disease, Parkinson’s disease), objective assessment of autonomic nervous system activity in clinical trials, and evaluation of melanopsin retinal ganglion cell function in patients with sleep and circadian rhythm disorders.
