Fig. 8: Evaluation of individualized treatment effect (measured by SE) from myopia control interventions using MPPM.
From: AI-guided personalized predictions on myopia progression and interventions

We employed MPPM to assess the individualized treatment effects of four myopia control interventions: 0.01% atropine (Atropine, a, b), peripheral defocus spectacles (PDS, c, d), orthokeratology lenses (Ortho-K, e, f), and repeated low-level red-light therapy (RLRL, g, h). The x-axis indicates time in months, with month 0 marking the initiation of the myopia control intervention. The y-axis represents the change in myopia degree, measured by the difference in SE from baseline (time 0). Each scatter point corresponds to an individual data observation. Orange dots represent pre-intervention measurements, while blue dots correspond to post-intervention follow-up data. Green dots depict the predicted trajectory of myopia progression without intervention, as estimated by NPM. Red dots reflect the predicted changes in myopia after intervention, as estimated by IPM. Solid lines represent fitted trends for each set of data points. These plots illustrate: (1) IPM reliably predicted individual myopia progression after the initiation of myopia control interventions; and (2) compared to the natural progression predicted by NPM, the actual rate of myopia progression was reduced in individuals receiving each of the four interventions. For the RLRL cohort, the orange line is simply an extension of the green line in the opposite direction, as participants in this cohort did not have refractive measurements before timepoint 0. SE spherical equivalent, od: right eye, os: left eye, Atropine: 0.01% atropine eye drops; PDS peripheral defocus spectacles, Ortho-K orthokeratology, RLRL repeated low-level reg light therapy.