SEMIAUTOMATED PATTERNED SCANNING LASER FOR RETINAL PHOTOCOAGULATION

The concept of retinal photocoagulation was introduced by Meyer-Schwickerath for the treatment of diabetic retinopathy in the 1950s and used with some success in the 1960s. The xenon arc photocoagulator utilized for this purpose was large, polychromatic, inefficient, and difficult to operate, prompting a search for a better method of treatment. Further progress was achieved when ruby,1 argon ion,2 and krypton ion3 lasers were coupled to a slit lamp with an articulating arm containing mirrors.4 A contact lens, aiming beam, and movable joystick were used to place the laser beam on the retina. These innovations allowed for creating single laser spots of variable size, power, and duration on the retina with a high degree of precision and ushered in the modern era of retinal laser photocoagulation in the 1970s. The techniques enabled by these devices, termed focal photocoagulation, grid photocoagulation, and panretinal photocoagulation, were refined and shown to be effective in the treatment of proliferative diabetic retinopathy and advanced forms of nonproliferative diabetic retinopathy associated with macular edema in large, prospective, multicenter, randomized trials—the DRS and ETDRS.5,6 These trials validated the efficacy and institutionalized the indications and parameters for treatment that have remained the gold standard since that time. Patients with high risk proliferative diabetic retinopathy who undergo panretinal photocoagulation typically receive between 1,200 and 1,500 laser spots in two to four sessions of 10 minutes to 20 minutes each over the course of 2 weeks to 4 weeks. Because the spots are delivered individually, treatments are time consuming and tedious for the patient and physician alike and can be painful, especially in the retinal periphery. Focal photocoagulation and grid photocoagulation for macular edema are less painful and time consuming, because the spots are applied more posteriorly and are fewer in number, but still are tedious and require a considerable degree of patient cooperation and physician skill to achieve a successful outcome and avoid complications. Until now, little has changed in the general design of the devices used for retinal photocoagulation aside from the substitution of fiber-optic cables for articulating arms and the use of air-cooled solid state lasers rather than water-cooled gas tubes. These innovations have had limited or no impact from the standpoint of the patient or physician on the technique of treatments and clinical outcomes. We reasoned that greater precision and safety in retinal photocoagulation might be achieved by delivering a multiplicity of spots in a pattern created by a scanner rather than as a series of individually placed lesions. We also wondered whether the pattern application time and patient discomfort could be further reduced by using shorter pulses than the conventional 100 milliseconds to 200 milliseconds recommended in the DRS and ETDRS.5,6 Prior efforts toward improvement in retinal photocoagulation systems were principally directed toward (1) fully automated systems with retinal stabilization based on eye tracking7–10 and (2) determination of the optimal dose in each spot using the tissue reflectance– based feedback systems.11 Automated systems required acquisition of an image of the retina before the treatment, planning and aligning all treatment locations with reference to the retinal image, and treating all of these locations automatically. Complex retinal tracking systems were also required in these approaches to ensure alignment between planned treatment locations defined on the acquired image and actual sites on the retina.7–10 The complexity of these fully automated systems hampered the introduction of