Simulating a colour-blind ophthalmologist for diagnosing and staging diabetic retinopathy

AlRyalat, Saif Aldeen; Muhtaseb, Ruba; Alshammari, Taher

doi:10.1038/s41433-020-01232-z

Cited by 4 publications

(5 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The results of our current study go hand-in-hand with a recently published study that assessed the impact of CVD on diabetic retinopathy staging, wherein graders with color vision deficiency had lower staging accuracy, a difference that was more pronounced among protanopic graders [ 18 ].…”

Section: Discussionsupporting

confidence: 66%

The Impact of Color Vision Deficiency on the Capability of Ophthalmologists to Diagnose Benign and Malignant Choroidal Tumors

Elfalah

AlRyalat

Abu-Yaghi

et al. 2023

JCM

Self Cite

View full text Add to dashboard Cite

Background: Color vision deficiency (CVD) is an under-reported problem among medical personnel, and its impact is still not well characterized. We aim to assess the impact of CVD among ophthalmologists on the accuracy of diagnosing different benign and malignant choroidal lesions. Methods: This is a cross-sectional study conducted on ophthalmologists. We used a web-based survey to collect responses through professional ophthalmology society social media. The survey included a set of five images for normal fundus, choroidal nevus, circumscribed choroidal hemangioma, choroidal metastasis, and choroidal melanoma, wherein each image simulated the three main types of CVD: protanopia, deuteranopia, and tritanopia, in addition to a non-simulated image. Results: Forty-one participants were included, with a mean age of 40 (±9.2) years. They were 28 (68%) men and 13 (32%) women. Participants showed significantly low accuracy for definite diagnosis for circumscribed choroidal hemangioma, nevus, melanoma, and metastasis when the images simulated protanopia and deuteranopia, but not tritanopia. Nevertheless, participants maintained the capability to recognize the nature of the lesions for both simulated and non-simulated images if they were benign or malignant, thereby ensuring immediate referral for specialized care. The exception was with simulated choroidal nevi images, wherein participants incorrectly assigned simulated protanopia and deuteranopia nevi images to malignant lesions. Conclusion: Protanopia and deuteranopia affected the accuracy of diagnosing several choroidal lesions; however, ophthalmologists with those two simulated CVDs were still able to discriminate between benign and malignant tumors.

show abstract

Section: Discussionsupporting

confidence: 66%

The Impact of Color Vision Deficiency on the Capability of Ophthalmologists to Diagnose Benign and Malignant Choroidal Tumors

Elfalah

AlRyalat

Abu-Yaghi

et al. 2023

JCM

Self Cite

View full text Add to dashboard Cite

show abstract

“…The implications of CVD on ophthalmologists’ clinical practice extend beyond our study findings. A previously published study demonstrated lower staging accuracy for diabetic retinopathy among graders with CVD, especially among those with protanopia [ 15 ]. Additionally, physicians with CVD have shown reduced confidence in identifying certain clinical signs through colored photographs compared to those with normal color vision [ 16 ].…”

Section: Discussionmentioning

confidence: 99%

Assessing the Impact of Simulated Color Vision Deficiency on Ophthalmologists’ Ability to Differentiate between Choroidal Melanoma and Choroidal Nevus

Yousef,

Alkhatib,

Elfalah

et al. 2024

JCM

Self Cite

View full text Add to dashboard Cite

Background: Color vision deficiency (CVD) is an often-overlooked issue within the medical community, and its consequences remain insufficiently explored. We aim to evaluate how CVD affects diagnostic accuracy and distinguish between malignant choroidal melanoma and benign choroidal nevus among ophthalmologists. Methods: In this cross-sectional study, we engaged ophthalmologists through a web-based survey distributed via the professional ophthalmology society’s social media channels. The survey encompassed a series of three fundus images representing normal fundus, choroidal nevus, and choroidal melanoma. Each image underwent simulation for the three primary types of CVD—protanopia, deuteranopia, and tritanopia—alongside a non-simulated version. Results: The study included 41 participants, averaging 40 years of age (±9.2), comprising 28 (68%) men and 13 (32%) women. Significantly lower rates of identifying orange pigments were observed in simulated protanopia images compared to non-simulated ones (p = 0.038). In simulated deutranopia images, the recognition of melanotic lesions was notably reduced compared to non-simulated images (p = 0.048). No such limitation was observed for tritanopia. However, participants retained their ability to identify subretinal fluid and estimate tumor thickness in simulated and non-simulated images. Concerning simulated images of choroidal nevi, participants misdiagnosed nevi as choroidal melanoma in 37% of cases in simulated protanopia nevi images and 41% in simulated deutranopia nevi images. This resulted in unnecessary referrals of benign lesions as malignant, emphasizing the potential for mistaken diagnoses. Nevertheless, almost all simulated images of malignant melanoma were correctly referred for specialized oncological treatment. Conclusions: The simulated CVD conditions of protanopia and deuteranopia affected the accuracy of identifying the melanotic nature of the choroidal tumor and the presence of orange pigments. This limitation led to challenges in correctly diagnosing choroidal melanoma and choroidal nevus, resulting in extra referrals for nevus cases. However, participants were safe and could still determine the possible risk of eyes with choroidal melanoma, so most referred melanoma cases to specialized oncologists as needed.

show abstract

“…This might not be a major issue for other ophthalmic diseases like diabetic retinopathy, where a single fundus image can be used reliably to diagnose the presence of bleeding or vascular abnormalities. 8 However, glaucoma is a disease with no consensus “gold-standard” definition to rely on making the development of a robust AI-based diagnostic model very problematic.…”

Section: The Need For a Gold Standardmentioning

confidence: 99%

“…of bleeding or vascular abnormalities. 8 However, glaucoma is a disease with no consensus "gold-standard" definition to rely on making the development of a robust AI-based diagnostic model very problematic.…”

mentioning

confidence: 99%

Artificial Intelligence and Glaucoma: Going Back to Basics

Singh¹,

Kalpathy–Cramer²,

AlRyalat³

2023

OPTH

Self Cite

View full text Add to dashboard Cite

There has been a recent surge in the number of publications centered on the use of artificial intelligence (AI) to diagnose various systemic diseases. The Food and Drug Administration has approved several algorithms for use in clinical practice. In ophthalmology, most advances in AI relate to diabetic retinopathy, which is a disease process with agreed upon diagnostic and classification criteria. However, this is not the case for glaucoma, which is a relatively complex disease without agreed-upon diagnostic criteria. Moreover, currently available public datasets that focus on glaucoma have inconstant label quality, further complicating attempts at training AI algorithms efficiently. In this perspective paper, we discuss specific details related to developing AI models for glaucoma and suggest potential steps to overcome current limitations.

show abstract

Simulating a colour-blind ophthalmologist for diagnosing and staging diabetic retinopathy

Cited by 4 publications

References 9 publications

The Impact of Color Vision Deficiency on the Capability of Ophthalmologists to Diagnose Benign and Malignant Choroidal Tumors

The Impact of Color Vision Deficiency on the Capability of Ophthalmologists to Diagnose Benign and Malignant Choroidal Tumors

Assessing the Impact of Simulated Color Vision Deficiency on Ophthalmologists’ Ability to Differentiate between Choroidal Melanoma and Choroidal Nevus

Artificial Intelligence and Glaucoma: Going Back to Basics

Contact Info

Product

Resources

About