CN-121983327-A - Artificial crystal model selection and rotation implantation prediction method and electronic equipment

CN121983327ACN 121983327 ACN121983327 ACN 121983327ACN-121983327-A

Abstract

The application discloses an artificial crystal model selection and rotation implantation prediction method and electronic equipment, which relate to the technical field of computers and are characterized in that by acquiring eye biological measurement parameters of a user and constructing an intraocular geometric constraint area, the intraocular lens is abstracted into a rectangular footprint geometry and a set of candidate size-axis positions is generated such that each candidate meets the geometric constraints of the patient's intraocular anatomy in terms of implantation position and rotation angle. On the basis, the geometrical adaptation indexes of the candidate schemes and the individuation ocular biology parameters are combined, the postoperative camber, the camber abnormal risk probability and the postoperative refractive result are predicted, and the output with confidence interval or risk probability is provided through post-processing calibration, so that the operation planning is not dependent on experience or rough estimation any more, the risk of camber deviation or postoperative refractive accidents is effectively reduced, the most preferable scheme of the artificial lens is provided, and individuation, quantifiable and predictable implantation planning is realized.

Inventors

NIU LINGLING
WANG XIAOYING
ZHOU XINGTAO
QIAN YISHAN
SHEN YANG
WANG LIN
CHEN XUN
JIANG YINJIE
LIU ZESHENG
ZHAO JING
LIU LIU

Assignees

复旦大学附属眼耳鼻喉科医院
上海智光瞳芯科技有限公司

Dates

Publication Date: 20260505
Application Date: 20260409

Claims (10)

1. An intraocular lens selection and rotation implantation prediction method, comprising: Acquiring eye biological measurement parameters of a user, and constructing an intraocular geometric constraint area according to the eye biological measurement parameters; Abstracting an artificial crystal into a rectangular space occupying geometrical body, and generating a candidate size-axis position set according to the size of the artificial crystal and the rotation angle around the center of the artificial crystal; determining geometric adaptation indexes meeting that the rectangular occupation geometric body is positioned in the geometric constraint area in the eye according to the candidate size-axis set to form a plurality of candidate schemes; predicting postoperative camber, camber abnormal risk probability and postoperative refractive result of each candidate scheme according to the ocular biology measurement parameters and the geometric adaptation indexes of the candidate schemes, performing post-processing calibration, and outputting prediction results of confidence intervals or risk probabilities of each candidate scheme; And sequencing the candidate schemes according to the prediction result, and outputting at least one recommended scheme according to the sequencing order and the geometric adaptation index of each candidate scheme.
2. The method according to claim 1, wherein the obtaining the biological measurement parameters of the eye of the user and constructing the geometric constraint area of the eye according to the biological measurement parameters of the eye comprises: Acquiring eye biological measurement parameters of a user including one or more of horizontal ciliary sulcus diameter, vertical ciliary sulcus diameter, cornea horizontal diameter, anterior chamber depth, eye axis length, lens thickness, cornea curvature, eye axis length, pupil diameter and refractive data; identifying whether an eyeball of a user has an abnormal region in a ditch or not, if so, representing the abnormal region in the ditch by a clock point or an angle interval; judging whether the data of the diameter of the horizontal ciliary sulcus and the diameter of the vertical ciliary sulcus are missing or not; If the data of the horizontal ciliary sulcus diameter and the vertical ciliary sulcus diameter are not missing, respectively taking the horizontal ciliary sulcus diameter and the vertical ciliary sulcus diameter as an ellipse major axis or an ellipse minor axis to construct an intraocular geometric constraint area; If any one of the data of the horizontal ciliary sulcus diameter or the vertical ciliary sulcus diameter is missing, using the data of the vertical ciliary sulcus diameter or the horizontal ciliary sulcus diameter to replace, and adopting the replaced data to construct an intraocular geometric constraint area; And if the data of the horizontal ciliary sulcus diameter and the vertical ciliary sulcus diameter are both missing, calculating an intraocular geometric constraint area based on the cornea horizontal diameter, wherein the length of the long axis or the short axis of the ellipse is not limited.
3. The method of claim 1, wherein abstracting the intraocular lens into a rectangular footprint geometry having a diagonal length equal to a nominal size of the intraocular lens, generating a candidate size-axis set from the size of the intraocular lens and the angle of rotation about its center, comprises: Abstracting an artificial lens into a rectangular space-occupying geometrical body, and setting the diagonal length of the rectangular space-occupying geometrical body to be equal to the nominal size of the artificial lens; and rotating the rectangular occupied geometric body around the center, scanning a plurality of candidate placement angles within the range of 0-180 degrees, and associating each candidate angle with the size of the rectangular occupied geometric body to form a candidate size-axis position set.
4. The method of claim 1, wherein associating each candidate angle with a size of the rectangular footprint geometry forms a candidate size-axis set, further comprising: Setting a rotation drift interval for each candidate angle, sampling geometric adaptation allowance of rectangular space occupying geometric bodies with a plurality of sizes in the rotation drift interval, and counting geometric adaptation allowance of the rectangular space occupying geometric bodies with each size under each candidate angle; Simulating a rotational drift range after the artificial lens operation according to geometric adaptation allowance of the rectangular occupying geometric body with each size under each candidate angle; the rotational drift range is added when forming the candidate size-axis set.
5. The method of claim 1, wherein determining, from the set of candidate size-axis positions, a geometric fit index that satisfies the rectangular footprint geometry within the intraocular geometric constraint area forms a plurality of candidate solutions, comprising: Constructing an elliptic equation corresponding to the geometric constraint area in the eye, and acquiring a normalized position in the elliptic equation; Judging whether four vertexes of a rectangle are all at normalized positions in the elliptic equation for each size of rectangular space occupying geometric bodies of any candidate angle in the candidate size-axis position set; When four vertexes of the rectangle are all positioned at the normalized position in the elliptic equation, determining that the rectangular occupation geometric body with the corresponding size under the candidate angle can be positioned in the intraocular geometric constraint area; And acquiring geometric adaptation indexes of the rectangular occupation geometric body capable of being positioned in the intraocular geometric constraint area to form a plurality of candidate schemes.
6. The method according to claim 1, wherein predicting the post-operative crown, the abnormal risk probability of crown and the post-operative refractive result of each candidate according to the ocular biometrical measurement parameter and the geometric adaptation index of the candidate, and performing post-processing calibration, and outputting the prediction result of the confidence interval or risk probability of each candidate comprises: stacking and integrating the artificial lens implantation historical data into a prediction model through a multi-base learner, and training the prediction model through adding a missing indicating variable; predicting postoperative arches, risk probability of abnormal arches and postoperative refractive results of all candidate schemes through the prediction model according to the ocular biological measurement parameters and the geometric adaptation indexes of the candidate schemes; Determining the risk of the arch height deviating from the safety value according to the predicted postoperative arch heights of the candidate schemes, and converting the risk into abnormal risk probability of the arch height; vectorizing modeling is carried out on the postoperative refractive result to obtain a postoperative refractive value, wherein the postoperative refractive value comprises sphere lens equivalence, astigmatism vector J0 and astigmatism vector J45; Constructing a residual target according to the postoperative refractive value and the expected postoperative refractive value, and predicting a refractive residual by the residual target to determine a postoperative refractive result; And carrying out post-processing calibration on the post-operation camber, camber abnormal risk probability and post-operation refraction result of each predicted candidate scheme, and outputting the predicted result of each candidate scheme with confidence interval or risk probability.
7. The method for intraocular lens selection and rotational implant prediction according to claim 4, the method is characterized in that the step of sorting the candidate schemes according to the prediction result comprises the following steps: Calculating the probability that the rectangular occupation geometric body with each size meets the upper and lower limits of the allowance in the corresponding drift interval under a given angle, the worst allowance in the middle, the rotation sensitivity, the asymmetry of four corners and vertexes of the rectangular occupation geometric body and the width of a feasible angle interval to form scoring parameters; And grading each candidate scheme according to the grading parameters, sorting each candidate scheme according to grading results, and outputting the feasible size, the feasible angle interval corresponding to each size, the optimal recommended angle, the adaptation probability and the stability grading.
8. The method for intraocular lens selection and rotational implant prediction according to claim 7, the method is characterized in that the step of grading each candidate scheme according to the grading parameters comprises the following steps: calculating a geometric score, an arch height safety score and a refractive accuracy score according to the scoring parameters; Setting an uncertainty punishment item, wherein rewarding or punishment is given when the difference value between the candidate angle and the horizontal or vertical angle is smaller than the first angle, the angle is adjusted in a rotation angle mode when the difference value between the candidate angle and the horizontal or vertical angle is larger than or equal to the first angle, and punishment is carried out according to the angle quantity deviating from the horizontal or vertical angle; setting a placement direction preference item; And determining comprehensive scores of the candidate schemes according to the geometric scores, the doming security scores, the refractive accuracy scores, the uncertainty penalty items and the placing direction preference items, and grading the candidate schemes according to the comprehensive scores.
9. The method according to claim 1, wherein the outputting at least one recommended solution according to the sorting order and the geometric adaptation index of each candidate solution comprises: Acquiring a recommended scheme of implanting the candidate angle which is a horizontal angle or a vertical angle according to the sorting order as an optimal recommended scheme; Acquiring a recommended scheme which has the same size as the rectangular occupation geometry of the optimal recommended scheme and has the smallest candidate angle of rotation from the horizontal or vertical direction according to the sorting sequence as a first alternative scheme; acquiring a recommended scheme which is implanted with the candidate angle being a horizontal angle or a vertical angle and is adjacent to the rectangular space occupying geometric body size of the optimal recommended scheme according to the sorting sequence as a second alternative scheme; and acquiring a recommended scheme with the candidate angle being the smallest rotation angle from the horizontal or vertical direction according to the sorting order as a third alternative scheme.
10. An electronic device, comprising: A memory for storing a computer program; a processor for implementing the steps of the intraocular lens selection and rotational implantation prediction method according to any one of claims 1 to 9 when executing said computer program.

Description

Artificial crystal model selection and rotation implantation prediction method and electronic equipment Technical Field The application relates to the technical field of geometric modeling, in particular to an artificial crystal model selection and rotation implantation prediction method and electronic equipment. Background ICL (Implantable Collamer Lens) is an implantable contact lens, also known as an intraocular lens, simply referred to as an intraocular lens or "intraocular lens". At present, ICL selection mainly depends on an empirical formula and a manufacturer rule table, and is roughly estimated by a small number of parameters such as cornea horizontal diameter (white to white, WTW), anterior Chamber Depth (ACD), eye axis length, comprehensive optometry result and the like. This approach has the following problems: first, individual differences are not adequately modeled, and there may be significant inequality between the horizontal ciliary sulcus diameter (h-STS) and the vertical ciliary sulcus diameter (v-STS) of the patient, and conventional methods that rely on WTW to estimate the actual elliptical space and directionality of the implanted intraocular lens often fail to reflect the actual elliptical space and directionality of the implanted intraocular lens. Secondly, the artificial lens rotation factor is considered to be insufficient, the ICL is approximately rectangular in shape, rotation is possible after implantation, and the final axial position has important influence on adaptability and stability. Third, ICL has only limited dimension specifications, taking staar surgical ICL as an example, and only has 4 dimension specifications, if only horizontal or vertical implantation is adopted, ICL dimension and STS cannot be continuously matched, the prior horizontal or vertical implantation mode can increase one specification to cause high arch, cause high intraocular pressure and even glaucoma, decrease one specification to cause low arch to cause cataract, or the horizontal implantation ICL is modified to cause low arch embarrassment after vertical implantation of ICL, and the operation safety and effectiveness are affected. ICL size is selected according to WTW before operation, ICL diopter is determined according to comprehensive optometry, however, whether proper camber and target diopter are reached or not can be evaluated only after implantation, and closed loop of candidate space generation and postoperative result prediction cannot be completed. Fifth, the risk identification of clinical abnormalities is deficient in that the most clinically significant risks are too low and too high, and residual refractive errors after surgery, but the existing procedures fail to provide probabilistic, calibratable risk information. Sixth, the real world data has noise and different standards, the preoperative/postoperative evaluation equipment has different standards in different technical development mechanisms, equipment operators have different technical proficiency, and the problems of data open-term, equipment precision error, human-caused operation error, non-uniform follow-up time point and the like exist, so that the traditional static formula is difficult to process robustly. Disclosure of Invention The application provides an artificial crystal model selection and rotation implantation prediction method and electronic equipment, which at least solve the problems that the artificial crystal model selection and rotation implantation in the related technology cannot be personalized and accurately determined, and whether the proper camber and target diopter are reached after the operation cannot be predicted before the operation so as to determine the reasonable artificial crystal model selection and rotation implantation angle. The application provides an artificial crystal model selection and rotary implantation prediction method, which comprises the following steps: Acquiring eye biological measurement parameters of a user, and constructing an intraocular geometric constraint area according to the eye biological measurement parameters; Abstracting an artificial crystal into a rectangular space occupying geometrical body, and generating a candidate size-axis position set according to the size of the artificial crystal and the rotation angle around the center of the artificial crystal; determining geometric adaptation indexes meeting that the rectangular occupation geometric body is positioned in the geometric constraint area in the eye according to the candidate size-axis set to form a plurality of candidate schemes; predicting postoperative camber, camber abnormal risk probability and postoperative refractive result of each candidate scheme according to the ocular biology measurement parameters and the geometric adaptation indexes of the candidate schemes, performing post-processing calibration, and outputting prediction results of confidence intervals or risk probabilities of each candida