Systems and methods for intraocular lens selection using emmetropia zone prediction
Systems and methods for intraocular lens (IOL) selection using emmetropia zone prediction include determining pre-operative measurements of an eye, estimating a post-operative anterior chamber depth (ACD) of an intraocular lens based on the pre-operative measurements, estimating a post-operative manifest refraction in spherical equivalent (MRSE) of the eye with the IOL implanted based on the pre-operative measurements and the estimated post-operative ACD, determining whether the eye with the IOL implanted is likely to be in an emmetropia zone based on the estimated post-operative MRSE, re-estimating the post-operative MRSE of the eye with the IOL implanted using an emmetropia zone prediction model or a non-emmetropia zone prediction model based on the emmetropia zone determining, and providing the re-estimated post-operative MRSE to a user to aid in selection of an IOL for implantation in the eye.
1 . A system comprising: one or more optical biometers configured to obtain one or more pre-operative measurements of an eye of a current patient;
a prediction engine comprising:
one or more processors; and memory containing a plurality of machine-readable instructions which when executed by one or more processors are adapted to cause the one or more processors to:
train a neural network emmetropia zone prediction model using a set of data from a collection of eyes of a plurality of previous patients, wherein the set of data includes an estimated post-operative manifest refraction in spherical equivalent (MRSE) for each of the plurality of previous patients and an actual post-operative MRSE for each of the plurality of previous patients where a post-operative eye was placed in an emmetropia zone of the previous patients;
train a neural network non-emmetropia zone prediction model using a set of data from a collection of eyes of previous patients, wherein the set of data includes an estimated post-operative manifest refraction in spherical equivalent (MRSE) for each of the plurality of previous patients and an actual post-operative MRSE for each of the plurality of previous patients where a post-operative eye was placed in a non-emmetropia zone of the previous patients;
receive, from the one or more optical biometer, the one or more pre-operative measurements of the eye of the current patient;
estimate, based on the one or more pre-operative measurements of the eye of the current patient, a post-operative anterior chamber depth (ACD) of an intraocular lens (IOL);
estimate, based on the one or more pre-operative measurements of the eye of the current patient and the estimated post-operative ACD, a post-operative manifest refraction in spherical equivalent (MRSE) of the eye of the current patient with the IOL implanted;
determine, based on the estimated post-operative MRSE, whether the eye of the current patient with the IOL implanted is likely to be in an emmetropia zone;
selecting the emmetropia zone prediction model based on determining that the eye of the current patient with the IOL implanted has a likelihood greater than or equal to a threshold value as being in the emmetropia zone, re-estimate the post-operative MRSE of the eye with the IOL implanted using the emmetropia zone prediction model, and
selecting the non-emmetropia zone prediction model based on determining that the eye of the current patient with the IOL implanted is not likely to be in the emmetropia zone, re-estimate the post-operative MRSE of the eye with the IOL implanted using the non-emmetropia zone prediction model; and
provide the re-estimated post-operative MRSE to a user to aid in selection of an IOL for implantation in the eye of the current patient;
wherein the one or more optical biometers are further configured to:
obtain one or more postoperative measurements of the eye of the current patient with the IOL implanted; and
update, based on the one or more post-operative measurements of the eye of the current patient, one of the neural network non-emmetropia zone prediction model or the neural network emmetropia zone prediction model.
2 . The system of claim 1 , wherein to estimate the post-operative ACD, the prediction engine is configured to:
generate a first estimate using a geometric model of the eye; and
update the first estimate using a neural network-based prediction model.
3 . The system of claim 1 , wherein to determine whether the eye with the IOL implanted is in the emmetropia zone, the prediction engine is configured to use a classifier and a receiver operator characteristic curve.
4 . The system of claim 1 , wherein the prediction engine is further configured to:
estimate and re-estimate a post-operative MRSE of a second IOL; and
present the re-estimated post-operative MRSE of the IOL and the re-estimated post-operative MRSE of the second IOL to the user to aid the user in the selection of the IOL for implantation in the eye;
wherein the IOL and the second IOL have different IOL powers.