Multimodal Imaging in Retinal Disease: A Consensus Definition.

Advances in retinal imaging are fundamental to the recent paradigm shifts in the diagnosis and management of ocular disease and in the understanding of the pathogenesis of these disorders.1 Evaluation of a broad spectrum of various types of retinal images is essential to modern ophthalmic practice and has become the optimal standard of care for successful management of retinal disorders. There are a wide range of retinal technologies that can acquire two-dimensional or threedimensional en face and cross-sectional images, some of which are depth-resolved, that may provide information about anatomic structure, function, or dynamic vascular flow. Common imaging modalities include color fundus photography; near-infrared reflectance (NIR); fundus autofluorescence;2,3 fluorescein angiography; indocyanine green angiography;4 scanning-laser ophthalmoscopy with2,5 and without adaptive optics;6,7 optical coherence tomography (OCT);8-10 and, more recently, OCT angiography (OCTA).11,12 Each modality has unique advantages and limitations and requires a specific skill set for optimal application and interpretation. When analyzed together, different imaging modalities can increase diagnostic sensitivity and specificity. Recently, the term “multimodal imaging” has been used in an increasing number of clinical reports, emphasizing the importance of combining a variety of imaging modalities. However, no specific definition of this terminology has been established in ophthalmology. The term multimodal imaging has been used in a variety of medical specialties such as oncology,13 radiology14,15, and cardiology long before expansion of the term to describe ophthalmic imaging techniques. It has been defined as the combination of imaging modalities to provide improved preclinical assessment, diagnostics, and therapeutic monitoring.16 The Cardiology Imaging Council defined multimodal imaging as “the efficient integration of two or more methods of imaging to improve the ability to diagnose, guide therapy, or predict outcomes.”17 In ophthalmology, it is uncommon for clinicians to rely on a single imaging modality to determine diagnosis and therapy. Instead most choose more than one technique for a given clinical situation. To be successful, one must have knowledge of the potential benefits and shortcomings of each imaging modality to know which is ideal for any given clinical situation. Because there is considerable overlap in the information provided by certain imaging modalities, it is important for clinicians to choose the minimum numbers of studies needed to accomplish the clinical goal and to avoid unnecessary cost and burden to the patient. In ophthalmology, most reports classify multimodal imaging as the combination of en face fundus imaging, en face dye-base angiography, cross-sectional and/or en face OCT, and possibly autofluorescence when relevant to evaluate a specific disease.18-22 Analyzing images from multiple imaging modalities can be used as a preliminary step in understanding new technologies, such as when comparing dye-based angiography to OCTA.23 We propose a definition of multimodal imaging as it applies to the practice of ophthalmology: To comprise the use of more than one technological system to acquire images, concurrently or at a short period of time, that complement one another for the purpose of diagnosis, prognostication, management, and monitoring of disease. Thus, all modalities can analyze a specific structure unequivocally. This may include hybrid devices that can simultaneously perform more than one imaging modality. For example, a fundus camera may provide color images of the retina; however, if excitation and emission filters are added to the camera to obtain fundus autofluorescence imaging, this device could be considered a multimodal platform. Similarly, commercial systems that combine OCTA with a corresponding structural OCT B-scan of the retina utilize different technological platforms and provide contrasting anatomical (blood flow versus static tissue layers) information and could also be considered a multimodal approach. Therefore, when the retina is imaged by different technological systems (whether the machine is changed or not), providing a different image of the fundus, this would constitute, in the authors’ consensus, a multimodal-imaging platform. There are many clinical situations in which the application of a multimodal approach can enhance the evaluation and understanding of retinal disease.