Search

In recent years, there has been considerable interest in the prospect of machine learning models demonstrating expert-level diagnosis in multiple disease contexts. However, there is concern that the excitement around this field may be associated with inadequate scrutiny of methodology and insufficient adoption of scientific good practice in the studies involving artificial intelligence in health care. This article aims to empower clinicians and researchers to critically appraise studies of clinical applications of machine learning, through: (1) introducing basic machine learning concepts and nomenclature; (2) outlining key applicable principles of evidence-based medicine; and (3) highlighting some of the potential pitfalls in the design and reporting of these studies.

Systematic screening for diabetic retinopathy (DR) has been widely recommended for early detection in patients with diabetes to address preventable vision loss. However, substantial manpower and financial resources are required to deploy opportunistic screening and transition to systematic DR screening programs. The advent of artificial intelligence (AI) technologies may improve access and reduce the financial burden for DR screening while maintaining comparable or enhanced clinical effectiveness. To deploy an AI-based DR screening program in a real-world setting, it is imperative that health economic assessment (HEA) and patient safety analyses are conducted to guide appropriate allocation of resources and design safe, reliable systems. Few studies published to date include these considerations when integrating AI-based solutions into DR screening programs. In this article, we provide an overview of the current state-of-the-art of AI technology (focusing on deep learning systems), followed by an appraisal of existing literature on the applications of AI in ophthalmology. We also discuss practical considerations that drive the development of a successful DR screening program, such as the implications of false-positive or false-negative results and image gradeability. Finally, we examine different plausible methods for HEA and safety analyses that can be used to assess concerns regarding AI-based screening.

Among the most noteworthy developments in ophthalmology over the last decade has been the emergence of quantifiable high-resolution imaging modalities, which are typically non-invasive, rapid and widely available. Such imaging is of unquestionable utility in the assessment of ocular disease however evidence is also mounting for its role in identifying ocular biomarkers of systemic disease, which we term oculomics. In this review, we highlight our current understanding of how retinal morphology evolves in two leading causes of global morbidity and mortality, cardiovascular disease and dementia. Population-based analyses have demonstrated the predictive value of retinal microvascular indices, as measured through fundus photography, in screening for heart attack and stroke. Similarly, the association between the structure of the neurosensory retina and prevalent neurodegenerative disease, in particular Alzheimer’s disease, is now well-established. Given the growing size and complexity of emerging multimodal datasets, modern artificial intelligence techniques, such as deep learning, may provide the optimal opportunity to further characterize these associations, enhance our understanding of eye-body relationships and secure novel scalable approaches to the risk stratification of chronic complex disorders of ageing.

PURPOSE: To develop an automated method for quantifying vitreous signal intensity on optical coherence tomography (OCT), with particular application for use in the assessment of vitreous inflammation. METHODS: This retrospective, observational case-control series comprised 30 patients (30 eyes), with vitreous haze secondary to intermediate, posterior, or panuveitis; 12 patients (12 eyes) with uveitis without evidence of vitreous haze; and 18 patients (18 eyes) without intraocular inflammation or vitreoretinal disease. The presence and severity of vitreous haze was classified according to the National Eye Institute system; other inflammatory indices and clinical parameters were also documented. Spectral-domain OCT images were analyzed using custom VITreous ANalysis software (termed 'VITAN'), which is fully automated and avoids the need for manual segmentation. RESULTS: VITAN performed accurate segmentation in all scans. Automated measurements of the vitreous:retinal pigment epithelium (RPE) signal ratio showed a moderate correlation with clinical vitreous haze scores (r = 0.585, P < 0.001), comparable to that reported using manual segmentation in our previous study (r = 0.566, P = 0.0001). The novel parameter of vitreous:RPE textural ratio showed a marginally stronger correlation (r = 0.604, P < 0.001) with clinical vitreous haze scores than the Vitreous:RPE signal ratio. CONCLUSIONS: The custom OCT image analysis software (VITAN) allows rapid and automated measurement of vitreous parameters, that is comparable to our previously reported vitreous:RPE index, and correlates with clinically measured disease activity. Such OCT-based indices may provide the much needed objective markers of vitreous activity, which may be used in both clinical assessment, and as outcome measures in clinical trials for intermediate, posterior, and panuveitis. TRANSLATIONAL RELEVANCE: We describe a rapid automated method for quantifying vitreous signal intensity on optical coherence tomography (OCT) and show that this correlates with clinical assessment of vitreous inflammation. Such OCT-based indices may provide the much needed objective markers of vitreous activity, which may be used both in routine clinical assessment, and as outcome measures in clinical trials for intermediate, posterior, and panuveitis.

Purpose: To investigate the structure���function relationship in eyes with drusen with mesopic and scotopic microperimetry. Methods: We analyzed structural and functional data from 43 eyes with drusen. Functional data were acquired with mesopic and scotopic two-color (red and cyan) microperimetry. Normative values were calculated using data from 56 healthy eyes. Structural measurements were green autofluorescence and dense macular optical coherence tomography scans. The latter were used to calculate the retinal pigment epithelium elevation (RPE-E) and the photoreceptor reflectivity ratio (PRR). The pointwise structure���function relationship was measured with linear mixed models having the log-transformed structural parameters as predictors and the sensitivity loss (SL, deviation from normal) as the response variable. Results: In the univariable analysis, the structural predictors were all significantly correlated (P < 0.05) with the SL in the mesopic and scotopic tests. In a multivariable model, mesopic microperimetry yielded the best structure���function relationship. All predictors were significant (P < 0.05), but the predictive power was weak (best R2 = 0.09). The relationship was improved when analyzing locations with abnormal RPE-E (best R2 = 0.18). Conclusions: Mesopic microperimetry shows better structure���function relationship compared to scotopic microperimetry; the relationship is weak, likely due to the early functional damage and the small number of tested locations affected by drusen. The relationship is stronger when locations with drusen are isolated for the mesopic and scotopic cyan test. Translational Relevance: These results could be useful to devise integrated structure���function methods to detect disease progression in intermediate age-related macular degeneration., DPC and GM received funding from the Innovative Medicines Initiative 2 Joint Undertaking under grant 116076 (Macustar). This joint undertaking receives support from the European Union's Horizon 2020 research and innovation program and European Federation of Pharmaceutical Industries and Associations (EFPIA). The funding for the NICOLA and NISA studies are reported in the acknowledgments as part of a mandatory statement for these studies. The article was first published in TVST. The final publication is available at https://tvst.arvojournals.org/article.aspx?articleid=2772110