Asia-Pacific Journal of Ophthalmology最新文献_第2页

Development and Testing of Artificial Intelligence-Based Mobile Application to Achieve Cataract Backlog-Free Status in Uttar Pradesh, India 开发和测试基于人工智能的移动应用程序，以实现印度北方邦无白内障积压状态。

IF 3.7 3区医学 Q1 OPHTHALMOLOGY

Asia-Pacific Journal of Ophthalmology

Pub Date : 2024-09-01 DOI: 10.1016/j.apjo.2024.100094

Background

Uttar Pradesh (UP), the most populous state in India, has about 36 million people aged 50 years or older, spread across more than 100,000 villages. Among them, an estimated 3.5 million suffer from visual impairments, including blindness due to untreated cataracts. To achieve cataract backlog-free status, UP is required to screen this population at the community level and provide treatment to those suffering from cataracts. We envisioned an AI-powered primary screening app utilizing eye images, deployable to frontline health workers for community-level screening. This paper outlines insights gained from developing the AI mobile app “Roshni” for cataract screening.

Method

The AI-based cataract classification model was developed using 13,633 eye images and finalized after three stages of experiments, detecting cataracts in images focused on the eye, iris, and pupil. Overall, 155 experiments were conducted using multiple deep learning algorithms, including ResNet50, ResNet101, YOLOv5, EfficientNetV2, and InceptionV3. We established a minimum threshold of 90 % specificity and sensitivity to ensure the algorithm’s suitability for field use.

Results

The cataract detection model for eye-focused images achieved 51.9 % sensitivity and 87.6 % specificity, while the model for iris-focused images, using a good/bad iris filter, achieved 52.4 % sensitivity and 93.3 % specificity. The classification model for segmented-pupil images, employing a good/bad pupil filter with UNet-based semantic segmentation model and EfficientNetV2, yielded 96 % sensitivity and 97 % specificity. Field testing with 302 beneficiaries (604 images) showed an overall sensitivity of 86.6 %, specificity of 93.3 %, positive predictive value of 58.4 %, and negative predictive value of 98.5 %.

Conclusion

This paper details the development of an AI mobile app designed to facilitate community screening for cataracts by frontline health workers.

背景介绍北方邦（Uttar Pradesh，UP）是印度人口最多的邦，约有 3600 万人年龄在 50 岁或以上，分布在 10 万多个村庄。其中，估计有 350 万人患有视力障碍，包括因白内障得不到治疗而失明。为了实现无白内障积压状态，UP 需要在社区层面对这部分人群进行筛查，并为白内障患者提供治疗。我们设想了一种利用眼部图像的人工智能初级筛查应用程序，可部署给一线卫生工作者进行社区一级的筛查。本文概述了在开发用于白内障筛查的人工智能移动应用程序 "Roshni "过程中获得的启示：方法：基于人工智能的白内障分类模型是利用 13,633 张眼部图像开发的，经过三个阶段的实验后最终确定，该模型可检测眼球、虹膜和瞳孔图像中的白内障。总体而言，我们使用多种深度学习算法进行了 155 次实验，包括 ResNet50、ResNet101、YOLOv5、EfficientNetV2 和 InceptionV3。我们设定了特异性和灵敏度均达到 90% 的最低阈值，以确保算法适合现场使用：眼球聚焦图像的白内障检测模型达到了 51.9% 的灵敏度和 87.6% 的特异性，而使用好/坏虹膜过滤器的虹膜聚焦图像模型达到了 52.4% 的灵敏度和 93.3% 的特异性。瞳孔分割图像分类模型采用了好/坏瞳孔过滤器、基于 UNet 的语义分割模型和 EfficientNetV2，灵敏度为 96%，特异度为 97%。对 302 名受益人（604 幅图像）进行的现场测试表明，总体灵敏度为 86.6%，特异性为 93.3%，阳性预测值为 58.4%，阴性预测值为 98.5%：本文详细介绍了一款人工智能移动应用程序的开发过程，该应用程序旨在为一线卫生工作者开展社区白内障筛查提供便利。

{"title":"Development and Testing of Artificial Intelligence-Based Mobile Application to Achieve Cataract Backlog-Free Status in Uttar Pradesh, India","authors":"","doi":"10.1016/j.apjo.2024.100094","DOIUrl":"10.1016/j.apjo.2024.100094","url":null,"abstract":"<div><h3>Background</h3><div>Uttar Pradesh (UP), the most populous state in India, has about 36 million people aged 50 years or older, spread across more than 100,000 villages. Among them, an estimated 3.5 million suffer from visual impairments, including blindness due to untreated cataracts. To achieve cataract backlog-free status, UP is required to screen this population at the community level and provide treatment to those suffering from cataracts. We envisioned an AI-powered primary screening app utilizing eye images, deployable to frontline health workers for community-level screening. This paper outlines insights gained from developing the AI mobile app “Roshni” for cataract screening.</div></div><div><h3>Method</h3><div>The AI-based cataract classification model was developed using 13,633 eye images and finalized after three stages of experiments, detecting cataracts in images focused on the eye, iris, and pupil. Overall, 155 experiments were conducted using multiple deep learning algorithms, including ResNet50, ResNet101, YOLOv5, EfficientNetV2, and InceptionV3. We established a minimum threshold of 90 % specificity and sensitivity to ensure the algorithm’s suitability for field use.</div></div><div><h3>Results</h3><div>The cataract detection model for eye-focused images achieved 51.9 % sensitivity and 87.6 % specificity, while the model for iris-focused images, using a good/bad iris filter, achieved 52.4 % sensitivity and 93.3 % specificity. The classification model for segmented-pupil images, employing a good/bad pupil filter with UNet-based semantic segmentation model and EfficientNetV2, yielded 96 % sensitivity and 97 % specificity. Field testing with 302 beneficiaries (604 images) showed an overall sensitivity of 86.6 %, specificity of 93.3 %, positive predictive value of 58.4 %, and negative predictive value of 98.5 %.</div></div><div><h3>Conclusion</h3><div>This paper details the development of an AI mobile app designed to facilitate community screening for cataracts by frontline health workers.</div></div>","PeriodicalId":8594,"journal":{"name":"Asia-Pacific Journal of Ophthalmology","volume":null,"pages":null},"PeriodicalIF":3.7,"publicationDate":"2024-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142071889","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Practice Patterns and Challenges in Managing Inherited Retinal Diseases Across Asia-Pacific: A Survey from the APIED Network 亚太地区治疗遗传性视网膜疾病的实践模式和挑战：APIED 网络调查。

IF 3.7 3区医学 Q1 OPHTHALMOLOGY

Asia-Pacific Journal of Ophthalmology

Pub Date : 2024-09-01 DOI: 10.1016/j.apjo.2024.100098

Purpose

The objective of this paper is to shed light on the current landscape of genotyping practices, phenotyping practices and availability of essential vision rehabilitation management for inherited retinal diseases (IRD) in the Asia-Pacific (APAC) Region.

Methods

The 62-item questionnaire was distributed electronically via email. The questions covered five domains: (1) structure of the IRD service and registry/database; (2) genotyping practices; (3) genetic counselling; (4) deep phenotyping practices; (5) low-vision rehabilitation services.

Results

The survey was completed by 36 of 45 centres in twelve countries and regions in APAC. Among these centres, 42 % reported managing more than 1000 patients. Notably, 39 % of centres lack an IRD database or registry, and 44 % of centres have tested less than one-quarter of their IRD patients. The majority of centres (67 %) do not have genetic counsellors. While there was consistency in the imaging-based investigations, there was marked heterogeneity for functional testing using electrophysiology and formal perimetry. Only 34 % of centres confirmed the availability of access to low-vision assistive devices.

Conclusions

This study reveals several critical gaps in managing IRDs in the APAC region. These include the lack of IRD database/registry in one-third of centres, a substantial proportion of patients remaining genetically undiagnosed, and limited availability of genetic counsellors. The findings also underscore a need to harmonise investigations for evaluating retinal function and identify areas for improvement in the provision of low-vision rehabilitation services.

目的：本文旨在揭示亚太地区（APAC）遗传性视网膜疾病（IRD）基因分型实践、表型实践和基本视力康复管理的现状：通过电子邮件以电子方式分发了 62 个项目的调查问卷。问题涉及五个方面：(1) IRD 服务和登记/数据库的结构；(2) 基因分型实践；(3) 遗传咨询；(4) 深度表型实践；(5) 低视力康复服务：亚太地区 12 个国家和地区的 45 个中心中有 36 个完成了调查。在这些中心中，42%的中心称管理着 1000 多名患者。值得注意的是，39% 的中心缺乏 IRD 数据库或登记册，44% 的中心对不足四分之一的 IRD 患者进行了测试。大多数中心（67%）没有遗传咨询师。虽然基于影像学的检查具有一致性，但使用电生理学和正规周视测量法进行的功能检测却存在明显差异。只有 34% 的中心确认可以使用低视力辅助设备：这项研究揭示了亚太地区在管理 IRD 方面存在的几个关键差距。结论：这项研究揭示了亚太地区在管理 IRD 方面存在的几个关键差距，其中包括三分之一的中心缺乏 IRD 数据库/登记簿，相当一部分患者仍未得到遗传学诊断，以及遗传学顾问的可用性有限。调查结果还强调，有必要统一视网膜功能评估调查，并确定在提供低视力康复服务方面需要改进的领域。

{"title":"Practice Patterns and Challenges in Managing Inherited Retinal Diseases Across Asia-Pacific: A Survey from the APIED Network","authors":"","doi":"10.1016/j.apjo.2024.100098","DOIUrl":"10.1016/j.apjo.2024.100098","url":null,"abstract":"<div><h3>Purpose</h3><div>The objective of this paper is to shed light on the current landscape of genotyping practices, phenotyping practices and availability of essential vision rehabilitation management for inherited retinal diseases (IRD) in the Asia-Pacific (APAC) Region.</div></div><div><h3>Methods</h3><div>The 62-item questionnaire was distributed electronically via email. The questions covered five domains: (1) structure of the IRD service and registry/database; (2) genotyping practices; (3) genetic counselling; (4) deep phenotyping practices; (5) low-vision rehabilitation services.</div></div><div><h3>Results</h3><div>The survey was completed by 36 of 45 centres in twelve countries and regions in APAC. Among these centres, 42 % reported managing more than 1000 patients. Notably, 39 % of centres lack an IRD database or registry, and 44 % of centres have tested less than one-quarter of their IRD patients. The majority of centres (67 %) do not have genetic counsellors. While there was consistency in the imaging-based investigations, there was marked heterogeneity for functional testing using electrophysiology and formal perimetry. Only 34 % of centres confirmed the availability of access to low-vision assistive devices.</div></div><div><h3>Conclusions</h3><div>This study reveals several critical gaps in managing IRDs in the APAC region. These include the lack of IRD database/registry in one-third of centres, a substantial proportion of patients remaining genetically undiagnosed, and limited availability of genetic counsellors. The findings also underscore a need to harmonise investigations for evaluating retinal function and identify areas for improvement in the provision of low-vision rehabilitation services.</div></div>","PeriodicalId":8594,"journal":{"name":"Asia-Pacific Journal of Ophthalmology","volume":null,"pages":null},"PeriodicalIF":3.7,"publicationDate":"2024-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142375001","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Effect of childhood atropine treatment on adult choroidal thickness using sequential deep learning-enabled segmentation 利用序列深度学习分割技术研究儿童期阿托品治疗对成人脉络膜厚度的影响。

IF 3.7 3区医学 Q1 OPHTHALMOLOGY

Asia-Pacific Journal of Ophthalmology

Pub Date : 2024-09-01 DOI: 10.1016/j.apjo.2024.100107

Purpose

To describe choroidal thickness measurements using a sequential deep learning segmentation in adults who received childhood atropine treatment for myopia control.

Design

Prospective, observational study.

Methods

Choroidal thickness was measured by swept-source optical coherence tomography in adults who received childhood atropine, segmented using a sequential deep learning approach.

Results

Of 422 eyes, 94 (22.3 %) had no previous exposure to atropine treatment, while 328 (77.7 %) had received topical atropine during childhood. After adjusting for age, sex, and axial length, childhood atropine exposure was associated with a thicker choroid by 32.1 μm (95 % CI, 9.2–55.0; P = 0.006) in the inner inferior, 23.5 μm (95 % CI, 1.9–45.1; P = 0.03) in the outer inferior, 21.8 μm (95 % CI, 0.76–42.9; P = 0.04) in the inner nasal, and 21.8 μm (95 % CI, 2.6–41.0; P = 0.03) in the outer nasal. Multivariable analysis, adjusted for age, sex, atropine use, and axial length, showed an independent association between central subfield choroidal thickness and the incidence of tessellated fundus (P < 0.001; OR, 0.97; 95 % CI, 0.96–0.98).

Conclusions

This study demonstrated that short-term (2–4 years) atropine treatment during childhood was associated with an increase in choroidal thickness of 20–40 μm in adulthood (10–20 years later), after adjusting for age, sex, and axial length. We also observed an independent association between eyes with thicker central choroidal measurements and reduced incidence of tessellated fundus. Our study suggests that childhood exposure to atropine treatment may affect choroidal thickness in adulthood.

目的：描述使用连续深度学习分割法测量儿童期阿托品治疗近视的成年人脉络膜厚度的情况：前瞻性观察研究：通过扫源光学相干断层扫描测量接受过儿童阿托品治疗的成年人的脉络膜厚度，并使用序列深度学习方法进行分割：结果：在422只眼睛中，94只（22.3%）以前没有接受过阿托品治疗，328只（77.7%）在儿童时期接受过局部阿托品治疗。在对年龄、性别和轴长进行调整后，童年时期接触过阿托品与脉络膜增厚有关，内下侧增厚 32.1 μm（95% CI，9.2 至 55.0；P = 0.006），内上侧增厚 23.5 μm (95% CI, 1.9 to 45.1; P = 0.03)，鼻内侧减少 21.8 μm (95% CI, 0.76 to 42.9; P = 0.04)，鼻外侧减少 21.8 μm (95% CI, 2.6 to 41.0; P = 0.03)。根据年龄、性别、阿托品使用情况和轴长进行调整后进行的多变量分析表明，中央叶下脉络膜厚度与网状眼底发生率之间存在独立关联（P < 0.001；OR，0.97；95% CI，0.96-0.98）：本研究表明，在调整了年龄、性别和轴长之后，儿童期短期（2-4 年）阿托品治疗与成年期（10-20 年后）脉络膜厚度增加 20-40 μm 有关。我们还观察到，中央脉络膜测量值较厚的眼睛与网状眼底发生率降低之间存在独立关联。我们的研究表明，儿童时期接受阿托品治疗可能会影响成年后的脉络膜厚度。

{"title":"Effect of childhood atropine treatment on adult choroidal thickness using sequential deep learning-enabled segmentation","authors":"","doi":"10.1016/j.apjo.2024.100107","DOIUrl":"10.1016/j.apjo.2024.100107","url":null,"abstract":"<div><h3>Purpose</h3><div>To describe choroidal thickness measurements using a sequential deep learning segmentation in adults who received childhood atropine treatment for myopia control.</div></div><div><h3>Design</h3><div>Prospective, observational study.</div></div><div><h3>Methods</h3><div>Choroidal thickness was measured by swept-source optical coherence tomography in adults who received childhood atropine, segmented using a sequential deep learning approach.</div></div><div><h3>Results</h3><div>Of 422 eyes, 94 (22.3 %) had no previous exposure to atropine treatment, while 328 (77.7 %) had received topical atropine during childhood. After adjusting for age, sex, and axial length, childhood atropine exposure was associated with a thicker choroid by 32.1 μm (95 % CI, 9.2–55.0; <em>P</em> = 0.006) in the inner inferior, 23.5 μm (95 % CI, 1.9–45.1; <em>P</em> = 0.03) in the outer inferior, 21.8 μm (95 % CI, 0.76–42.9; <em>P</em> = 0.04) in the inner nasal, and 21.8 μm (95 % CI, 2.6–41.0; <em>P</em> = 0.03) in the outer nasal. Multivariable analysis, adjusted for age, sex, atropine use, and axial length, showed an independent association between central subfield choroidal thickness and the incidence of tessellated fundus (<em>P</em> < 0.001; OR, 0.97; 95 % CI, 0.96–0.98).</div></div><div><h3>Conclusions</h3><div>This study demonstrated that short-term (2–4 years) atropine treatment during childhood was associated with an increase in choroidal thickness of 20–40 μm in adulthood (10–20 years later), after adjusting for age, sex, and axial length. We also observed an independent association between eyes with thicker central choroidal measurements and reduced incidence of tessellated fundus. Our study suggests that childhood exposure to atropine treatment may affect choroidal thickness in adulthood.</div></div>","PeriodicalId":8594,"journal":{"name":"Asia-Pacific Journal of Ophthalmology","volume":null,"pages":null},"PeriodicalIF":3.7,"publicationDate":"2024-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142387539","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Gain of chromosome 8q and high expression of EZH2 may predict poor prognosis in Chinese patients with uveal melanoma 8q染色体的增益和EZH2的高表达可预测中国葡萄膜黑色素瘤患者的不良预后。

IF 3.7 3区医学 Q1 OPHTHALMOLOGY

Asia-Pacific Journal of Ophthalmology

Pub Date : 2024-09-01 DOI: 10.1016/j.apjo.2024.100108

Purpose

To explore risk factors predicting poor prognosis of uveal melanoma in a Chinese population, with specific emphasis on monosomy 3, 8q gain, and EZH2 staining.

Methods

Eighty-nine patients with uveal melanoma from 2012 to 2021 were reviewed. Clinical and pathological records were collected and analyzed. Immunohistochemical staining of EZH2, monosomy 3 and 8q gain were respectively conducted in 45, 54, and 57 cases. Survival was evaluated by Kaplan–Meier analysis and log-rank test. Cox proportional hazard regressions were employed to predict risk factors of distant metastasis.

Results

The median follow-up was 44 months. Altogether, 16 % of patients developed distant metastases and died from disease-related causes. Disease-specific survival at one and three years was 96.6 % and 88.4 % while distant metastasis rates were 7.9 % and 12 %. Univariate Cox regression analysis revealed that age (HR: 1.04), tumor largest basal diameter (HR: 1.21), tumor thickness (HR: 1.21), ciliary body involvement (HR: 3.50), AJCC stage (HR: 5.68), epithelioid cell type (HR: 7.71), 8q gain (HR: 7.48), and high expression of EZH2 (HR: 6.09) were associated with distant metastasis. 8q gain was associated with epithelioid cell type and thicker tumor while EZH2 was correlated with epithelioid cell type. Monosomy 3 lacked a significant correlation with other factors.

Conclusion

EZH2 and 8q gain could be taken into consideration when calculating poor prognosis in Chinese patients with uveal melanoma. Monosomy 3 showed no significance in distant metastasis, but this may be due to a small sample size.

目的：探讨在中国人群中预测葡萄膜黑色素瘤不良预后的风险因素，特别强调单体3、8q增益和EZH2染色：研究回顾了2012年至2021年的89例葡萄膜黑色素瘤患者。收集并分析了临床和病理记录。分别对45、54和57例患者进行了EZH2、3单体和8q增殖的免疫组化染色。生存率通过 Kaplan-Meier 分析和对数秩检验进行评估。结果显示，中位随访时间为44个月：中位随访时间为 44 个月。结果：中位随访时间为44个月，共有16%的患者出现远处转移并死于疾病相关原因。一年和三年的疾病特异性生存率分别为96.6%和88.4%，远处转移率分别为7.9%和12%。单变量Cox回归分析显示，年龄（HR：1.04）、肿瘤最大基底直径（HR：1.21）、肿瘤厚度（HR：1.21）、睫状体受累（HR：3.50）、AJCC分期（HR：5.68）、上皮样细胞类型（HR：7.71）、8q增益（HR：7.48）和EZH2高表达（HR：6.09）与远处转移相关。8q增益与上皮样细胞类型和肿瘤较厚有关，而EZH2与上皮样细胞类型相关。结论：EZH2和8q增益与上皮样细胞类型和肿瘤厚度相关，而EZH2与上皮样细胞类型相关，单体3与其他因素无明显相关性：结论：在计算中国葡萄膜黑色素瘤患者的不良预后时，可将EZH2和8q基因增殖考虑在内。单体3对远处转移无显著影响，但这可能是由于样本量较小。

{"title":"Gain of chromosome 8q and high expression of EZH2 may predict poor prognosis in Chinese patients with uveal melanoma","authors":"","doi":"10.1016/j.apjo.2024.100108","DOIUrl":"10.1016/j.apjo.2024.100108","url":null,"abstract":"<div><h3>Purpose</h3><div>To explore risk factors predicting poor prognosis of uveal melanoma in a Chinese population, with specific emphasis on monosomy 3, 8q gain, and EZH2 staining.</div></div><div><h3>Methods</h3><div>Eighty-nine patients with uveal melanoma from 2012 to 2021 were reviewed. Clinical and pathological records were collected and analyzed. Immunohistochemical staining of EZH2, monosomy 3 and 8q gain were respectively conducted in 45, 54, and 57 cases. Survival was evaluated by Kaplan–Meier analysis and log-rank test. Cox proportional hazard regressions were employed to predict risk factors of distant metastasis.</div></div><div><h3>Results</h3><div>The median follow-up was 44 months. Altogether, 16 % of patients developed distant metastases and died from disease-related causes. Disease-specific survival at one and three years was 96.6 % and 88.4 % while distant metastasis rates were 7.9 % and 12 %. Univariate Cox regression analysis revealed that age (HR: 1.04), tumor largest basal diameter (HR: 1.21), tumor thickness (HR: 1.21), ciliary body involvement (HR: 3.50), AJCC stage (HR: 5.68), epithelioid cell type (HR: 7.71), 8q gain (HR: 7.48), and high expression of EZH2 (HR: 6.09) were associated with distant metastasis. 8q gain was associated with epithelioid cell type and thicker tumor while EZH2 was correlated with epithelioid cell type. Monosomy 3 lacked a significant correlation with other factors.</div></div><div><h3>Conclusion</h3><div>EZH2 and 8q gain could be taken into consideration when calculating poor prognosis in Chinese patients with uveal melanoma. Monosomy 3 showed no significance in distant metastasis, but this may be due to a small sample size.</div></div>","PeriodicalId":8594,"journal":{"name":"Asia-Pacific Journal of Ophthalmology","volume":null,"pages":null},"PeriodicalIF":3.7,"publicationDate":"2024-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142456904","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A review of ophthalmology education in the era of generative artificial intelligence 人工智能时代的眼科教育回顾。

IF 3.7 3区医学 Q1 OPHTHALMOLOGY

Asia-Pacific Journal of Ophthalmology

Pub Date : 2024-07-01 DOI: 10.1016/j.apjo.2024.100089

Purpose

To explore the integration of generative AI, specifically large language models (LLMs), in ophthalmology education and practice, addressing their applications, benefits, challenges, and future directions.

Design

A literature review and analysis of current AI applications and educational programs in ophthalmology.

Methods

Analysis of published studies, reviews, articles, websites, and institutional reports on AI use in ophthalmology. Examination of educational programs incorporating AI, including curriculum frameworks, training methodologies, and evaluations of AI performance on medical examinations and clinical case studies.

Results

Generative AI, particularly LLMs, shows potential to improve diagnostic accuracy and patient care in ophthalmology. Applications include aiding in patient, physician, and medical students’ education. However, challenges such as AI hallucinations, biases, lack of interpretability, and outdated training data limit clinical deployment. Studies revealed varying levels of accuracy of LLMs on ophthalmology board exam questions, underscoring the need for more reliable AI integration. Several educational programs nationwide provide AI and data science training relevant to clinical medicine and ophthalmology.

Conclusions

Generative AI and LLMs offer promising advancements in ophthalmology education and practice. Addressing challenges through comprehensive curricula that include fundamental AI principles, ethical guidelines, and updated, unbiased training data is crucial. Future directions include developing clinically relevant evaluation metrics, implementing hybrid models with human oversight, leveraging image-rich data, and benchmarking AI performance against ophthalmologists. Robust policies on data privacy, security, and transparency are essential for fostering a safe and ethical environment for AI applications in ophthalmology.

目的：探讨生成式人工智能，特别是大型语言模型（LLMs）在眼科教育和实践中的整合，探讨其应用、益处、挑战和未来方向：设计：对当前人工智能在眼科领域的应用和教育项目进行文献综述和分析：分析已发表的有关人工智能在眼科中应用的研究、评论、文章、网站和机构报告。检查包含人工智能的教育项目，包括课程框架、培训方法以及人工智能在医学考试和临床案例研究中的表现评估：生成式人工智能，尤其是 LLM，显示出提高眼科诊断准确性和患者护理的潜力。其应用包括帮助病人、医生和医科学生接受教育。然而，人工智能的幻觉、偏差、缺乏可解释性以及训练数据过时等挑战限制了临床应用。研究显示，LLM 对眼科医学考试题的准确性参差不齐，这突出表明需要更可靠的人工智能集成。全国有多个教育项目提供与临床医学和眼科学相关的人工智能和数据科学培训：结论：生成式人工智能和 LLM 为眼科教育和实践带来了充满希望的进步。通过包括基本人工智能原则、道德准则和最新、无偏见的培训数据在内的综合课程来应对挑战至关重要。未来的发展方向包括制定与临床相关的评估指标、在人工监督下实施混合模型、利用图像丰富的数据以及以眼科医生为基准来衡量人工智能的性能。健全的数据隐私、安全和透明度政策对于为眼科领域的人工智能应用营造一个安全、道德的环境至关重要。

{"title":"A review of ophthalmology education in the era of generative artificial intelligence","authors":"","doi":"10.1016/j.apjo.2024.100089","DOIUrl":"10.1016/j.apjo.2024.100089","url":null,"abstract":"<div><h3>Purpose</h3><p>To explore the integration of generative AI, specifically large language models (LLMs), in ophthalmology education and practice, addressing their applications, benefits, challenges, and future directions.</p></div><div><h3>Design</h3><p>A literature review and analysis of current AI applications and educational programs in ophthalmology.</p></div><div><h3>Methods</h3><p>Analysis of published studies, reviews, articles, websites, and institutional reports on AI use in ophthalmology. Examination of educational programs incorporating AI, including curriculum frameworks, training methodologies, and evaluations of AI performance on medical examinations and clinical case studies.</p></div><div><h3>Results</h3><p>Generative AI, particularly LLMs, shows potential to improve diagnostic accuracy and patient care in ophthalmology. Applications include aiding in patient, physician, and medical students’ education. However, challenges such as AI hallucinations, biases, lack of interpretability, and outdated training data limit clinical deployment. Studies revealed varying levels of accuracy of LLMs on ophthalmology board exam questions, underscoring the need for more reliable AI integration. Several educational programs nationwide provide AI and data science training relevant to clinical medicine and ophthalmology.</p></div><div><h3>Conclusions</h3><p>Generative AI and LLMs offer promising advancements in ophthalmology education and practice. Addressing challenges through comprehensive curricula that include fundamental AI principles, ethical guidelines, and updated, unbiased training data is crucial. Future directions include developing clinically relevant evaluation metrics, implementing hybrid models with human oversight, leveraging image-rich data, and benchmarking AI performance against ophthalmologists. Robust policies on data privacy, security, and transparency are essential for fostering a safe and ethical environment for AI applications in ophthalmology.</p></div>","PeriodicalId":8594,"journal":{"name":"Asia-Pacific Journal of Ophthalmology","volume":null,"pages":null},"PeriodicalIF":3.7,"publicationDate":"2024-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2162098924000902/pdfft?md5=c732bd60e43fb37cd5168b2517837774&pid=1-s2.0-S2162098924000902-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141970533","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Upholding artificial intelligence transparency in ophthalmology: A call for collaboration between academia, industry, and government for patient care in the 21st century 维护眼科人工智能的透明度：呼吁学术界、产业界和政府合作，促进 21 世纪的患者护理。

IF 3.7 3区医学 Q1 OPHTHALMOLOGY

Asia-Pacific Journal of Ophthalmology

Pub Date : 2024-07-01 DOI: 10.1016/j.apjo.2024.100093

引用次数: 0

Comment on “Update on coronavirus disease 2019: Ophthalmic manifestations and adverse reactions to vaccination” 关于 "2019 年冠状病毒疾病更新：眼科表现和疫苗接种不良反应 "的评论

IF 3.7 3区医学 Q1 OPHTHALMOLOGY

Asia-Pacific Journal of Ophthalmology

Pub Date : 2024-07-01 DOI: 10.1016/j.apjo.2024.100079

引用次数: 0

The role of saliency maps in enhancing ophthalmologists’ trust in artificial intelligence models 突出图在增强眼科医生对人工智能模型的信任方面的作用。

IF 3.7 3区医学 Q1 OPHTHALMOLOGY

Asia-Pacific Journal of Ophthalmology

Pub Date : 2024-07-01 DOI: 10.1016/j.apjo.2024.100087

Purpose

Saliency maps (SM) allow clinicians to better understand the opaque decision-making process in artificial intelligence (AI) models by visualising the important features responsible for predictions. This ultimately improves interpretability and confidence. In this work, we review the use case for SMs, exploring their impact on clinicians’ understanding and trust in AI models. We use the following ophthalmic conditions as examples: (1) glaucoma, (2) myopia, (3) age-related macular degeneration, and (4) diabetic retinopathy.

Method

A multi-field search on MEDLINE, Embase, and Web of Science was conducted using specific keywords. Only studies on the use of SMs in glaucoma, myopia, AMD, or DR were considered for inclusion.

Results

Findings reveal that SMs are often used to validate AI models and advocate for their adoption, potentially leading to biased claims. Overlooking the technical limitations of SMs, and the conductance of superficial assessments of their quality and relevance, was discerned. Uncertainties persist regarding the role of saliency maps in building trust in AI. It is crucial to enhance understanding of SMs' technical constraints and improve evaluation of their quality, impact, and suitability for specific tasks. Establishing a standardised framework for selecting and assessing SMs, as well as exploring their relationship with other reliability sources (e.g. safety and generalisability), is essential for enhancing clinicians' trust in AI.

Conclusion

We conclude that SMs are not beneficial for interpretability and trust-building purposes in their current forms. Instead, SMs may confer benefits to model debugging, model performance enhancement, and hypothesis testing (e.g. novel biomarkers).

目的：显著性图（Saliency maps，SM）通过可视化负责预测的重要特征，让临床医生更好地理解人工智能（AI）模型中不透明的决策过程。这最终会提高可解释性和可信度。在这项工作中，我们回顾了 SM 的使用案例，探讨了 SM 对临床医生理解和信任人工智能模型的影响。我们以以下眼科疾病为例：(1）青光眼；（2）近视；（3）老年性黄斑变性；（4）糖尿病视网膜病变：方法：使用特定关键词在 MEDLINE、Embase 和 Web of Science 上进行多领域检索。结果：研究结果表明，SMs 在青光眼、近视、AMD 或 DR 中的应用非常普遍：结果：研究结果表明，人工智能模型经常被用于验证人工智能模型并倡导采用人工智能模型，这可能会导致有偏见的说法。研究发现，人们忽视了SMs的技术局限性，并对其质量和相关性进行了肤浅的评估。关于显著性地图在建立人工智能信任方面的作用，仍然存在不确定性。加强对突出显示图的技术限制的了解，改进对其质量、影响和对特定任务的适用性的评估至关重要。建立选择和评估SMs的标准化框架，以及探索它们与其他可靠性来源（如安全性和普遍性）的关系，对于增强临床医生对人工智能的信任至关重要：我们的结论是，目前形式的 SMs 对可解释性和建立信任并无益处。相反，SMs 可为模型调试、模型性能提升和假设检验（如新型生物标记物）带来益处。

{"title":"The role of saliency maps in enhancing ophthalmologists’ trust in artificial intelligence models","authors":"","doi":"10.1016/j.apjo.2024.100087","DOIUrl":"10.1016/j.apjo.2024.100087","url":null,"abstract":"<div><h3>Purpose</h3><p>Saliency maps (SM) allow clinicians to better understand the opaque decision-making process in artificial intelligence (AI) models by visualising the important features responsible for predictions. This ultimately improves interpretability and confidence. In this work, we review the use case for SMs, exploring their impact on clinicians’ understanding and trust in AI models. We use the following ophthalmic conditions as examples: (1) glaucoma, (2) myopia, (3) age-related macular degeneration, and (4) diabetic retinopathy.</p></div><div><h3>Method</h3><p>A multi-field search on MEDLINE, Embase, and Web of Science was conducted using specific keywords. Only studies on the use of SMs in glaucoma, myopia, AMD, or DR were considered for inclusion.</p></div><div><h3>Results</h3><p>Findings reveal that SMs are often used to validate AI models and advocate for their adoption, potentially leading to biased claims. Overlooking the technical limitations of SMs, and the conductance of superficial assessments of their quality and relevance, was discerned. Uncertainties persist regarding the role of saliency maps in building trust in AI. It is crucial to enhance understanding of SMs' technical constraints and improve evaluation of their quality, impact, and suitability for specific tasks. Establishing a standardised framework for selecting and assessing SMs, as well as exploring their relationship with other reliability sources (e.g. safety and generalisability), is essential for enhancing clinicians' trust in AI.</p></div><div><h3>Conclusion</h3><p>We conclude that SMs are not beneficial for interpretability and trust-building purposes in their current forms. Instead, SMs may confer benefits to model debugging, model performance enhancement, and hypothesis testing (e.g. novel biomarkers).</p></div>","PeriodicalId":8594,"journal":{"name":"Asia-Pacific Journal of Ophthalmology","volume":null,"pages":null},"PeriodicalIF":3.7,"publicationDate":"2024-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2162098924000884/pdfft?md5=947c519db2eaf73afca2cbfe4639495c&pid=1-s2.0-S2162098924000884-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141787164","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Natural Language Processing in medicine and ophthalmology: A review for the 21st-century clinician 医学和眼科学中的自然语言处理：面向 21 世纪临床医生的综述》。

IF 3.7 3区医学 Q1 OPHTHALMOLOGY

Asia-Pacific Journal of Ophthalmology

Pub Date : 2024-07-01 DOI: 10.1016/j.apjo.2024.100084

Natural Language Processing (NLP) is a subfield of artificial intelligence that focuses on the interaction between computers and human language, enabling computers to understand, generate, and derive meaning from human language. NLP's potential applications in the medical field are extensive and vary from extracting data from Electronic Health Records –one of its most well-known and frequently exploited uses– to investigating relationships among genetics, biomarkers, drugs, and diseases for the proposal of new medications. NLP can be useful for clinical decision support, patient monitoring, or medical image analysis. Despite its vast potential, the real-world application of NLP is still limited due to various challenges and constraints, meaning that its evolution predominantly continues within the research domain. However, with the increasingly widespread use of NLP, particularly with the availability of large language models, such as ChatGPT, it is crucial for medical professionals to be aware of the status, uses, and limitations of these technologies.

自然语言处理（NLP）是人工智能的一个分支领域，主要研究计算机与人类语言之间的互动，使计算机能够理解、生成人类语言并从中获取意义。NLP 在医疗领域的潜在应用非常广泛，从提取电子健康记录中的数据--这是其最著名、最常用的用途之一--到研究遗传学、生物标记物、药物和疾病之间的关系，从而提出新的药物建议，不一而足。NLP 还可用于临床决策支持、患者监控或医学图像分析。尽管 NLP 潜力巨大，但由于各种挑战和制约因素，其在现实世界中的应用仍然有限，这意味着它的发展主要仍停留在研究领域。然而，随着 NLP 的应用越来越广泛，特别是随着大型语言模型（如 ChatGPT）的出现，医疗专业人员了解这些技术的现状、用途和局限性至关重要。

{"title":"Natural Language Processing in medicine and ophthalmology: A review for the 21st-century clinician","authors":"","doi":"10.1016/j.apjo.2024.100084","DOIUrl":"10.1016/j.apjo.2024.100084","url":null,"abstract":"<div><p>Natural Language Processing (NLP) is a subfield of artificial intelligence that focuses on the interaction between computers and human language, enabling computers to understand, generate, and derive meaning from human language. NLP's potential applications in the medical field are extensive and vary from extracting data from Electronic Health Records –one of its most well-known and frequently exploited uses– to investigating relationships among genetics, biomarkers, drugs, and diseases for the proposal of new medications. NLP can be useful for clinical decision support, patient monitoring, or medical image analysis. Despite its vast potential, the real-world application of NLP is still limited due to various challenges and constraints, meaning that its evolution predominantly continues within the research domain. However, with the increasingly widespread use of NLP, particularly with the availability of large language models, such as ChatGPT, it is crucial for medical professionals to be aware of the status, uses, and limitations of these technologies.</p></div>","PeriodicalId":8594,"journal":{"name":"Asia-Pacific Journal of Ophthalmology","volume":null,"pages":null},"PeriodicalIF":3.7,"publicationDate":"2024-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2162098924000859/pdfft?md5=4d1793e4d147d08d7a1dbc7d3ffdca4e&pid=1-s2.0-S2162098924000859-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141765063","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Diversity, equity and inclusion in curriculum vitae for medical and surgical specialty training college entrance 内科和外科专科培训学院入学简历中的多样性、公平性和包容性。

IF 3.7 3区医学 Q1 OPHTHALMOLOGY

Asia-Pacific Journal of Ophthalmology

Pub Date : 2024-07-01 DOI: 10.1016/j.apjo.2024.100080

引用次数: 0