目的：建立和验证一个涉及多级临床场景的白内障协作通用的人工智能(artificial intelligence,AI)管理平台，探索基于AI的医疗转诊模式，以提高协作效率和资源覆盖率。方法：训练和验证的数据集来自中国AI医学联盟，涵盖多级医疗机构和采集模式。使用三步策略对数据集进行标记: 1)识别采集模式；2)白内障诊断包括正常晶体眼、白内障眼或白内障术后眼；3)从病因和严重程度检测需转诊的白内障患者。此外，将白内障AI系统与真实世界中的居家自我监测、初级医疗保健机构和专科医院等多级转诊模式相结合。结果：通用AI平台和多级协作模式在三步任务中表现出可靠的诊断性能: 1)识别采集模式的受试者操作特征(receiver operating characteristic curve,ROC)曲线下面积(area under the curve,AUC)为99.28%~99.71%)；2)白内障诊断对正常晶体眼、白内障或术后眼，在散瞳-裂隙灯模式下的AUC分别为99.82%、99.96%和99.93%，其他采集模式的AUC均 > 99%；3)需转诊白内障的检测(在所有测试中AUC >91%)。在真实世界的三级转诊模式中，该系统建议30.3%的人转诊，与传统模式相比，眼科医生与人群服务比率大幅提高了10.2倍。结论：通用AI平台和多级协作模式显示了准确的白内障诊断性能和有效的白内障转诊服务。建议AI的医疗转诊模式扩展应用到其他常见疾病和资源密集型情景当中。

Objective: To establish and validate a universal artificial intelligence (AI) platform for collaborative management of cataracts involving multilevel clinical scenarios and explored an AI-based medical referral pattern to improve collaborative efficiency and resource coverage. Methods: The training and validation datasets were derived from the Chinese Medical Alliance for Artificial Intelligence, covering multilevel healthcare facilities and capture modes. The datasets were labelled using a three step strategy: (1)capture mode recognition; (2) cataract diagnosis as a normal lens, cataract or a postoperative eye and (3) detection of referable cataracts with respect to aetiology and severity. Moreover, we integrated the cataract AI agent with a real-world multilevel referral pattern involving self-monitoring at home, primary healthcare and specialised hospital services. Results: The universal AI platform and multilevel collaborative pattern showed robust diagnostic performance in three-step tasks: (1) capture mode recognition (area under the curve (AUC) 99.28%–99.71%), (2) cataract diagnosis (normal lens, cataract or postoperative eye with AUCs of 99.82%, 99.96% and 99.93% for mydriatic-slit lamp mode and AUCs >99% for other capture modes) and (3)detection of referable cataracts (AUCs >91% in all tests). In the real-world tertiary referral pattern, the agent suggested 30.3%  of people be ’referred’, substantially increasing the ophthalmologist-to-population service ratio by 10.2-fold compared with the traditional pattern. Conclusions: The universal AI platform and multilevel collaborative pattern showed robust diagnostic performance and effective service for cataracts. The context of our AI-based medical referral pattern will be extended to other common disease conditions and resource-intensive situations.

目的：观察依镜有晶状体眼后房屈光晶体(phakic refractive lens,PRL)植入术矫正超高度近视的疗效和安全性。方法：纳入自2018年1月—2020年9月在深圳市眼科医院行依镜PRL植入术的超高度近视患者共24例39眼，进行自身对照研究。其中，男8例13眼，女16例26眼，平均年龄(31.15±6.33)岁。观察术后屈光度、视力[裸眼视力(uncorrected visual acuity,UCVA)、最佳矫正视力(best corrected visual acuity,BCVA)]、眼压、角膜内皮细胞计数、拱高、手术并发症等指标。结果：中位随访时间为5.5(3.0，11.0)月。屈光度从术前(?22.29±4.96) D降低至术后(?0.28±1.01) D(t=24.421，P<0.001)，其中术后球镜度数±0.5 D占28眼(82.4%)，±1.0 D占31眼(91.2%)。LogMAR UCVA从术前1.40(1.30，1.70)改善至术后(0.28±0.20) D，LogMAR BCVA由术前0.40(0.22，0.70)改善至术后0.15(0.00，0.30)，差异均有统计学意义(均P<0.001)。术后BCVA较术前提高3.00(1.00，5.00)行，所有患者均无BCVA丢失。术前、术后1 d、末次随访眼压值比较差异有统计学意义(F=8.779 P=0.012)，其中术后1 d较术前眼压增高(Z=-3.401，P=0.001)，而末次随访较术后1 d眼压降低(Z=-2.685，P=0.007)，末次随访与术前眼压之间比较差异无统计学意义(Z=-0.894，P=0.371)。角膜内皮细胞计数从术前(2 782.20±296.30)个/mm²降低至术后(2 472.54±394.32)个/mm² (t=-5.437，P<0.001)，平均丢失角膜内皮细胞数11.2%。末次随访33眼平均拱高值为(379.00±283.27)μm，其中0~250 μm占12眼(36.4%)，250~750 μm占19眼(57.6%)，大于750 μm占2眼(6%)。21眼PRL术后3个月拱高值为(269.81±194.67)μm，较术后1个月拱高值(373.62±195.75)μm降低(t=?2.917，P=0.009)。术后并发症包括激素性青光眼(1例2眼)、PRL光学面裂痕(1眼)、黄斑出血(1眼)、PRL偏位(2例3眼)。结论：对于超高度近视患者，依镜PRL植入术是一种可供选择的安全、有效的眼内屈光手术方式，但其远期疗效及安全性仍需更长时间和更大样本量进一步观察。

Objective: To assess the efficacy and safety of phakic refractive lens (PRL) implantation for the correction of ultra-high myopia. Methods: This self-controlled case series study included 39 eyes of 24 patients with ultra-high myopia who underwent PRL implantation at Shenzhen Eye Hospital between January 2018 and September 2020. The study comprised 13 eyes in 8 males and 26 eyes in 16 females, with a mean age of (31.15 ± 6.33) years. Postoperative parameters, including refraction, visual acuity (UCVA, BCVA), intraocular pressure, corneal endothelial cell count,vault, and surgical complication were observed. Results: The median follow-up time was 5.5 (3, 11) months. The refraction significantly decreased from preoperative (-22.29±4.96) D to postoperative (-0.28±1.01) D (t=24.421, P<0.001). Postoperatively, 82.4% of eyes achieved a spherical degree within ±0.5 D, and 91.2% within ±1.0 D. LogMAR UCVA significantly improved from 1.40 (1.30, 1.70) preoperatively to (0.28±0.20) postoperatively. LogMAR BCVA significantly improved from 0.40 (0.22, 0.70) preoperatively to 0.15 (0.00, 0.30) postoperatively (P<0.001 for all). Postoperative BCVA improved by 3.00 (1.00, 5.00) lines compared with preoperative BCVA, with no instances of BCVA loss in any patient. Intraocular pressure values showed significant differences among preoperative, 1 day postoperative and last follow up (F=8.779, P=0.012). Intraocular pressure increased significantly 1 day after surgery compared to before surgery (Z=-3.401, P=0.001), but decreased significantly at the last follow-up compared to 1 day postoperatively(Z=-2.685, P=0.007), with no significant difference in intraocular pressure between preoperative and last follow-up (Z=-0.894, P=0.371). Corneal endothelial cell count decreased significantly from preoperative (2 782.20±296.30)/mm² to postoperative (2 472.54±394.32)/mm² (t=?5.437, P<0.001), with a mean loss of 11.2%. The average vault at the last follow-up was (379.00±283.27) μm, of which 0~250 μm in 12 eyes (36.4%), 250~750 μm in 19 eyes (57.6%), and > 750 μm in 2 eyes (6%). In 21 eyes, the vault at 3 months postoperative (269.81±194.67) μm was significantly lower than that at 1 month postoperative (373.62±195.75) μm (t=?2.917, P=0.009). Postoperative complications included steroid-induced glaucoma (2 eyes in 1 case), PRL optical surface crack (1 eye), macular hemorrhage (1 eye), and PRL decentration (3 eyes in 2 cases). Conclusions: PRL implantation is a safe and effective intraocular refractive surgery for ultra-high myopic patients. Nonetheless, it should be neccessary to observe for long-term efficacy and saff lens; high myopia