Mao Grace

Hood Donald C, Lau Wai Tak, Stowman Arin L, Gebhardt Tayna, Mao Grace, La Bruna Sol et al.

A deep learning model detected glaucoma from OCT reports, but its "heat maps" often didn't precisely highlight clinically relevant damage locations, suggesting current AI guidance isn't optimal for clinicians.

Sun Anna J, Gomide Gabriel, Tsamis Emmanouil, Mao Grace, Leshno Ari, La Bruna Sol et al.

OCT revealed diverse macular GCL damage patterns in advanced glaucoma, consistently following a temporal-to-nasal progression model. This aids clinical monitoring and trial design for preserving central vision.

La Bruna Sol, Rai Anvit, Mao Grace, Kerr Jennifer, Amin Heer, Zemborain Zane Z et al.

OCT RNFL maps frequently show glaucoma-like artifacts in healthy eyes. A "vertical midline" rule helps distinguish true glaucoma, but other OCT data are crucial for high sensitivity, especially in early disease.