Tiarnán D L Keenan - research works

In the progression cohort, over 2.1±1.6 years, 66/1,082 high-intensity lipophilic statin-exposed participants (6.10%) and 85/1,082 unexposed participants (7.86%) progressed to exudative AMD (sHR=0.62, 95%-CI, 0.45-0.87; P=0.006; IPCW HR=0.61, 95%-CI, 0.43-0.87; P=0.006).

Any statin use was not associated with AMD.

For cataract, the sHR was 1.18 (95%-CI, 1.03-1.34; P=0.015), and the IPCW HR was 1.20 (95%-CI, 1.05-1.38; P=0.007).

For PVD, the sHR was 1.00 (95%-CI, 0.81-1.22; P=0.96), and the IPCW HR was 1.00 (95%-CI, 0.81-1.24; P=0.99).

Artificial intelligence (AI) offers potential solutions.

However, advancing MLLMs for ophthalmology is hindered by the lack of unified, comprehensive benchmark datasets for development and evaluation.

In this work, we present LMOD+, a large-scale multimodal ophthalmology benchmark dataset comprising 32,633 instances with multi-granular annotations across 12 common ophthalmic conditions and 5 imaging modalities.

First, we expanded the dataset by nearly 50% (from 21,933 to 32,633 instances).

Among the cohorts reporting predictability outcomes, 16 of 19 achieved more than 60% of eyes within ±0.50 diopter (D) of the intended correction.

Likewise, 16 of 21 cohorts reported more than 90% of eyes within ±1.00 D of target, and all cohorts reached at least 80% within this range.

The safety index (ratio of postoperative CDVA to preoperative CDVA) ranged from 0.97 to 1.05.

Worse stability outcomes were associated with higher hyperopia treatments, slower pulse rates (< 500 Hz), older laser models, and smaller optical zones (≤ 6 mm).