Ortho-K parameters: best measured manually or with software?

Paper title: Manual and software-based measurements of treatment zone parameters and characteristics in children with slow and fast axial elongation in orthokeratology

Authors: Guo, Biyue (1); Wu, Huihuan (2); Cheung, Sin Wan (1); Cho, Pauline (1)

1. Centre for Myopia Research, School of Optometry, The Hong Kong Polytechnic University, Hong Kong SAR, China
2. Department of Electrical Engineering, The Hong Kong Polytechnic University, Hong Kong, SAR, China

Date: Apr 2022

Reference: Guo B, Wu H, Cheung SW, Cho P. Manual and software-based measurements of treatment zone parameters and characteristics in children with slow and fast axial elongation in orthokeratology. Ophthalmic Physiol Opt. 2022 Jul;42(4):773-785.

[Link to abstract]

Orthokeratology (Ortho-K) is thought to provide a myopia controlling effect through changes in the flatter central treatment zone (TZ) and the steeper mid-peripheral zone (PSZ). These changes can include the size of the TZ, central and peripheral corneal power changes and the width and decentration of PSZ.^1-4   Measurement of these parameters can be manual or software-based.

This study aimed to compare TZ measurements obtained using both methods and to assess the relationship between the decentration and characteristics of TZ and axial elongation rates after 2yrs of Ortho-K wear.

This retrospective study analyzed topographical data from 69 participants across 3 previous studies: ROMIO⁵, TO-SEE⁶ and OKIC⁷. They were aged between 7 and 13yrs old with myopia of up to 5.00D and astigmatism up to 3.50D and wore either Z Night or Z Night toric lenses (Menicon) for 2yrs.

TZ parameters along the horizontal and vertical axes were manually measured from topography maps for fast progressors (>0.54mm, n = 38) and slow progressors (<0.18mm, n = 31).  The distances measured were converted to refractive power differences and the rate of power change in the periphery for nasal, temporal, inferior and superior directions was calculated. The measurements were repeated using Python-based software and compared.

• Differences in the measurement of refractive power changes in the PSZ and most of the TZ parameters were found to be significantly different. However, both methods showed that slow progressors typically had more TZ decentration, smaller TZ size and higher rate of power changes inferiorly.
• The TZ decentration was mostly temporal (manual -0.36mm, software-based -0.41mm) and inferior (manual -0.18mm, software-based -0.16mm).
• Clinically insignificant differences were seen for the decentration, size and width of PSZ.

This study showed that although statistically significant differences were found between manual and software-based measurements for TZ size and decentration and PSZ width, they may still be clinically acceptable values.

• The values for TZ size and decentration and for inferior power changes were different for the fast and slow progressors, suggesting they play a role in myopia control with Ortho-K

Software techniques allow for quick and easy measuring, particularly for inexperienced users, for complicated measurements or research purposes.

• However, manual measuring is adequate if no software is available.

Due to how software algorithms locate the TZ centre and allow for missing or incomplete information, it means manual and software measuring methods are not interchangeable

The differences between the manual and software-based measurements may be due to:

• Software algorithms using a best-fit ellipse approach to locate the center of the TZ, whereas manual measuring involves approximately choosing the center before measuring outwards
• Refractive power change differences are likely to be due to error-correcting algorithms within the software program. These differences may cause variations in other TZ characteristics.
• The algorithm is unknown for the Medmont topographer used in the studies. This and a lack of studies into manual v software measurements means it is not possible to say how much more accurate software-based measuring is than manual measuring.

Further research into this could establish how accurate manual measurements can be.

Angle kappa (angle between visual and pupillary axes) may be relevant to measuring lens decentration. However, it was not measured in the 3 original studies and therefore not included in the data. The authors state the impact of angle kappa will be included in a future study.

The combination of the smaller TZ size, greater TZ decentration and the higher rate of inferior power change for slow progressors may be influencing myopia progression with Ortho-K.

• One study highlighted a relationship between pupil diameter and TZ size, suggesting that pupils larger than the TZ size experienced slower axial growth and reduced myopic progression.⁸
• Another found a positive correlation between the horizontal TZ size and axial elongation after 1yr of Ortho-K wear.⁹

Careful evaluation of TZ characteristics and further studies into decentration of TZ will explain their role in myopia control and axial length slowing in Ortho-K treatment. 

Title: Manual and software-based measurements of treatment zone parameters and characteristics in children with slow and fast axial elongation in orthokeratology

Authors: Guo, Biyue; Wu, Huihuan; Cheung, Sin Wan; Cho, Pauline

Purpose: To compare the treatment zone (TZ) measurements obtained using manual and software-based methods in orthokeratology (ortho-k) subjects and explore the TZ characteristics of children with slow and fast axial elongation after ortho-k.

Methods: Data from 69 subjects (aged 7 to <13years old), who participated in three 24-month longitudinal orthokeratology studies, showing fast (>0.27 mm, n = 38) and slow (<0.09 mm, n = 31) axial elongation, were retrieved. The TZ after ortho-k was defined as the central flattened area enclosed by points with no refractive power change. TZ parameters, including decentration, size, width of the peripheral steepened zone (PSZ), central and peripheral refractive power changes and peripheral rate of power change, were determined manually and using python- based software. TZ parameters were compared between measurement methods and between groups.

Results: Almost all TZ parameters measured manually and with the aid of software were significantly different (p < 0.05). Differences in decentration, size and the PSZ width were not clinically significant, but differences (0.45 to 0.92 D) in refractive power change in the PSZ were significant, although intraclass coefficients (0.95 to 0.98) indicated excellent agreement between methods. Significantly greater TZ decentration, smaller TZ size and greater inferior rate of power change (relative to the TZ centre) were observed in slow progressors using both methods, suggesting a potential role of TZ in regulating myopia progression in ortho-k.

Conclusions: TZ measurements using manual and software-based methods differed significantly and cannot be used interchangeably. The combination of TZ decentration, TZ size and peripheral rate of power change may affect myopia control effect in ortho-k.

[Link to abstract]