From Empiricism to Simulation: A New Era in Cataract Surgery Planning

For centuries, “successful” cataract surgery simply meant removing an opaque lens. Even early techniques such as couching – displacing the lens rather than replacing it – were considered victories against inevitable blindness. Today, that definition feels almost anachronistic. Modern cataract surgery is remarkably safe, and the absence of severe complications is increasingly viewed not as the goal, but as the baseline.

The real benchmark has shifted: patients judge success by whether their postoperative vision matches the agreed refractive target – distance, near, or a tailored compromise.

Refractive accuracy is no longer optional

Once intraocular lens (IOL) implantation became routine, accurate IOL power calculation transformed cataract surgery into a genuine refractive procedure. Early IOL selection was largely empirical until mathematical approaches – tracing back to Fyodorov’s foundational work – introduced a structured scientific framework. The progress is measurable: in many contemporary series, approximately 90% of patients now achieve outcomes within ±0.50 D of target (1, 2).

Artificial intelligence now represents the next structural shift in IOL power calculation. By refining effective lens position (ELP) prediction, AI-driven models are advancing rapidly, and their influence on refractive accuracy is poised to expand significantly in the immediate future.

Yet even this degree of precision fails to address a persistent cultural issue: refractive “near-misses” are still too often accepted by some surgeons – especially in older patients.

Such complacency is increasingly difficult to defend. A residual refractive error after cataract surgery is not comparable to an imprecise spectacle prescription – it may require corneal laser enhancement, piggyback implantation, or even IOL exchange. Cataract surgery is a once-in-a-lifetime optical intervention, replacing one of the eye’s two principal refractive components. Approaching it as anything less than a high-stakes refractive opportunity ultimately underserves our patients.

Premium expectations, premium complexity

Patients’ expectations have evolved alongside surgical reliability and modern lifestyles. “Clearer” vision is no longer sufficient; patients increasingly demand spectacle independence. In response, presbyopia-correcting technologies – particularly EDOF and multifocal IOLs – have proliferated, transforming the marketplace into a rapidly expanding and heavily marketed arena.

With new presbyopia-correcting IOL models introduced almost every year, selecting an IOL has become at least as complex – and as consequential – as determining its power. Yet here the field reveals an uncomfortable inconsistency: while IOL power calculation is rigorously formula-driven, IOL type selection often remains shaped by availability, economic pressures, local practice patterns, and persuasive manufacturer narratives.

The future of IOL model selection lies in optical simulation and computer-based analysis. At present, we remain in a kind of pre-Fyodorov era – this time in optical design selection rather than dioptric calculation. A simulation-driven approach may help explain why approximately 10% of patients report dissatisfaction with presbyopia-correcting lenses (3). Unlike simple spherical error, dissatisfaction related to higher-order aberrations rarely resolves with spectacles. For the affected minority, the consequences can be substantial – glare, halos, reduced contrast sensitivity, and, perhaps most damaging, erosion of trust in the surgical outcome.

Optical simulation is not merely an academic refinement; it is poised to become an essential step toward more predictable, rational, and truly personalized refractive outcomes.

Why optics still matters in a neuro-visual world

A common counterargument is that optical simulations can never fully predict real-world vision because the brain is not a linear sensor. It adapts, suppresses information, and continuously recalibrates perception. This is true – and clinically relevant, particularly for IOL designs that rely on so-called “static accommodation” principles.

But acknowledging neuroadaptation does not relegate optics to a secondary role. On the contrary, a higher-quality optical system reduces the burden placed on neural processing. The brain can compensate – but it should not be forced to compensate unnecessarily. High-quality retinal image formation remains the foundation upon which neural interpretation is built.

Clinical judgment must integrate both physiology and physics. Yet physics is the domain we can model, quantify, simulate, and compare objectively. Neuroadaptation is powerful, but it does not absolve us from optimizing the optical system itself. 

Moving beyond paraxial approximations

In engineering optics – whether in microscopes, telescopes, or camera systems – computer-based ray tracing is the standard. Complex optical systems are designed using comprehensive models, not simplified paraxial approximations. By contrast, cataract planning still frequently relies on assumptions that struggle to capture the true complexity of modern IOL optics interacting with real corneal geometries and dynamically changing pupils.

Ray tracing provides a logical evolution. By simulating light propagation through the cornea, pupil, and a specific IOL design – applying fundamental physical laws such as Snell’s law, and in advanced platforms incorporating diffraction and light scattering – the surgeon can anticipate not only the refractive endpoint but also the quality of vision across focus. This is precisely the territory in which presbyopia-correcting lenses either succeed or fail.

The value of this approach becomes even more apparent in post-refractive surgery eyes. Corneal aberrations are altered in predictable yet clinically significant ways: myopic ablations typically induce positive spherical aberration, whereas hyperopic treatments tend to induce negative spherical aberration. These altered corneal signatures interact directly with the intrinsic aberration profile of a given IOL design – particularly in EDOF optics – where the interaction may be synergistic or, conversely, visually detrimental.

A glimpse of decision support: simulation enters the clinic

Platforms such as AIOLsci, developed by Prof. Damien Gatinel and myself, illustrate the direction in which cataract planning is evolving: accessible, clinician-oriented simulation tools that enable comparison of IOL designs under defined corneal power, corneal asphericity, and pupil diameter conditions, using objective metrics such as through-focus modulation transfer function (MTF).

Importantly, these platforms are not medical recommendation engines. They require broader clinical validation and larger datasets correlating simulated predictions with both patient-reported outcomes and objective performance measures. Nevertheless, they provide something cataract surgeons have historically lacked: a physics-based preview of optical behavior that complements – and disciplines – clinical intuition.

The next definition of “planning”

Cataract surgery has already evolved from simple lens removal to fully fledged refractive surgery. The next evolution must occur not in the operating room, but in the planning process itself – from empiricism and marketing-influenced selection to simulation-informed decision-making.

Ray tracing will not replace clinical judgment. It will refine and discipline it – allowing surgeons to understand how the cornea, pupil, and IOL design function together as a unified optical system rather than as isolated variables.

In an era in which patients expect spectacle independence and seamless functional vision across distances, moving “from empiricism to simulation” is not an academic refinement. It is the logical maturation of refractive cataract surgery – and an overdue step toward truly personalized, physics-informed outcomes.