Automatic Temperature Controlled Retinal Photocoagulation Facilitates Four Predictable Lesion Strengths Including Sub-threshold Lesions

Abstract

 

Purpose

 

Conventional photocoagulation relies on retinal whitening for power adjustment. The method produces inconstant lesions and does not allow to control the effect of very mild or sub-threshold lesions. We applied automatic, temperature controlled retinal photocoagulation, that uses temperature data to adjust irradiation times, and aimed at lesions of five different strengths in rabbits.

 

Methods

 

We used a modified 532nm photocoagulator that facilitated real-time retinal temperature measurements based on optoacoustics. A feedback loop allowed automatic exposure time control onto predefined time / temperature dependent characteristics for different lesion strengths. We applied lesions of 133 µm diameter to six rabbit eyes and varied laser powers (20 - 66 mW) and intensities. N = 225 control lesions were applied conventionally with fixed, 200 ms, exposure time, and n = 794 lesions with the new, automatic method in five consecutive characteristic groups. After 1 hour to 12 weeks, we examined fundus color and Spectralis® optical coherence tomographic (OCT) images and classified lesion morphologies according to a sixfold OCT classifier.

 

Results

 

Control lesion effects depended on the laser power and achieved morphological classes 2 - 6. Automatically controlled characteristic groups 1 - 4 correlated to increasing morphological lesion intensities (incidence peaks in OCT classes 1 to 4). These groups showed increasing median funduscopic diameters (0 [IQR 0-0] µm - 170 [149-195] µm) and OCT diameters (129 [0-144] µm - 282 [261-302] µm). Group 5 did not differ significantly from group 4. Visibility rates in funduscopy (OCT) were 17 % (68 %) for group 1, 38 % (90 %) for group 2 and > 94 % (> 98%) for all consecutive groups.

 

Conclusions

 

Automatic, temperature controlled photocoagulation allows to apply predictable lesions. A reliable reproducibility for mild and subthreshold lesions can be obtained for characteristic groups 1 - 4, which is not possible with conventional power control. Only strong lesions (characteristic group 5) do not reproduce as accurately owing to technical reasons. Overtreatment, however, can be avoided in any case. The automatic control facilitates standardized low intensity photocoagulation independently of the treating physician and of spot-individual transmission and pigmentation variation.

 

Keywords: 578 laser • 688 retina • 551 imaging/image analysis: non-clinical