Laser retinal hazard region: 400–1400 nm

At the same wavelength and duration, a beam’s power alone does not tell you how concentrated the exposure will be on the retina. You also need the power reaching the retinal image and the area over which that power is distributed. This lesson develops that distinction, then applies it to source size, optical aids and alignment controls.

Follow the radiation to the retina

In the retinal hazard region, approximately 400–1400 nm, radiation can pass through the eye’s optical media to the retina. The cornea and lens can concentrate it into a small image. A beam spread over an area at the front of the eye can therefore deposit its transmitted power over a much smaller retinal area. The relevant injury mechanisms also depend on wavelength and exposure duration; focusing does not select a single universal limit.

Read the diagram from left to right: incident rays enter the eye and converge towards a retinal image. It illustrates concentration, not a measured image diameter or a permissible exposure. Actual transmission and image formation depend on wavelength, source geometry and the eye.

Near-IR within this region can reach the retina without appearing bright. An invisible beam therefore lacks the visible cue that might prompt someone to look away. Begin an assessment from the credible exposure path and duration, rather than from how bright the beam seems.

Separate collected power from image area

For a simple uniform-image model, let Pᵣ be the optical power arriving at the retinal image and Aᵣ its area. Its mean irradiance is Eᵣ = Pᵣ/Aᵣ. For a circular image of radius r, Aᵣ = π r². Doubling the radius spreads the same power over four times the area, so mean irradiance falls to one quarter. Doubling power at fixed area instead doubles irradiance. If both change, compare the power ratio with the area ratio.

This is a model of mean irradiance. A real image can have a nonuniform profile and a higher local peak. Tissue absorption, duration and the adopted exposure-limit method still need separate treatment; this model is not a retinal exposure limit.

Connect apparent source size to the retinal image

The angular subtense α describes the apparent source’s angular size as viewed from the eye’s position. For a simple resolved object of diameter D at distance R, the small-angle approximation gives α ≈ D/R, in radians. An ideal imaging model with effective focal length f then gives image diameter d ≈ fα.

The apparent source need not be a physical exit aperture: optics can change its apparent location and size. Beam diameter at the cornea, beam divergence and apparent source angle describe different aspects of the exposure. Do not substitute one for another without establishing the geometry.

Exposure methods distinguish effectively point-like sources from extended sources. A larger apparent source can spread transmitted power over a larger retinal image, changing the applicable limit form. Classification uses the adopted method’s angular bounds and duration rules; it does not follow simply from whether the physical laser aperture looks large.

The ICNIRP laser guidelines, including the 2020 erratum define apparent source size and the conditions for its use in retinal assessment. A geometric image calculation alone cannot classify a source for a formal limit, particularly when diffraction and aberrations matter.

Reassess the path when optical aids are possible

A microscope, telescope or other optical attachment can change the collection aperture, transmission, apparent source and retinal image. More collected power does not by itself specify the change in retinal irradiance: the resulting image area also matters. Nor does a larger image by itself prove acceptable exposure.

If an assessment assumes unaided viewing, establish whether aided viewing is credible during normal use, alignment or service. Assess the actual aid and path, or implement an enforceable restriction that removes it from the task. Where the adopted method requires an aided-viewing hazard distance, calculate that boundary rather than carrying over the unaided one. The NOHD lesson develops distance-based assessment.

Use a credible viewing duration

An aversion response is a conditional time-base assumption in methods that permit it for specified visible exposures. It is not a guarantee that a person will blink or turn away before injury. Invisible near-IR gives no visible trigger; deliberate inspection, constrained posture and optical-aid use also require their own scenario assessment.

For each task, state how an eye could enter the direct or reflected path and how long exposure could credibly continue. Use the duration rules of the adopted method. Do not borrow a visible accidental-viewing assumption for invisible alignment, or use eyewear as a reason to place an eye in the beam plane.

Match controls to the exposure path

Start with engineering controls: enclose accessible beam segments, terminate beams in suitable blocks or dumps, and use interlocks or guarded viewing arrangements where required. Include specular reflections: a reflective tool can redirect a concentrated beam towards an eye even when the primary beam points away.

Administrative controls make the task assumptions operational: an approved alignment procedure specifies permitted output states, access, optical aids and who may perform the work. Any residual eyewear requirement must match the wavelength band and assessed attenuation need. Eyewear supports the controlled task; it does not remove direct and reflected paths.

For the implementation details, see beam enclosures and interlocks, alignment hazards and eyewear optical density.

Recover the assessment chain

Reconstruct the task in this order: wavelength and tissue endpoint; accessible direct and reflected paths; unaided or aided geometry; credible duration; applicable limit method; controls that enforce those assumptions. Keep the power reaching the image separate from its area. “Invisible”, “small beam” and “larger image” are observations to investigate, not exposure-limit conclusions.

Knowledge check

Exercises