Abstract
Purpose:
To evaluate the efficacy of in vivo confocal microscopy in the diagnosis of corneal disease.
Methods:
Thirty eyes of 30 patients with corneal diseases were included in the study. All patients underwent in vivo confocal microscopy and color picture of the anterior segment. Round hyper-reflective bodies seen with in vivo confocal microscopy were defined as UFOs (Unknown Full Objects). Frames containing UFOs were selected and analyzed by a masked observer (MP), to ascertain whether confocal images alone were sufficient to formulate a correct diagnosis. The masked observer was then provided with the color picture of the cornea and asked to re-assess his/her hypothesis of diagnosis if needed.
Results:
Fifteen out of the 30 patients presented Acanthamoeba keratitis; 4 conjunctival pigmented lesion; 3 Map-Dot-Fingerprint keratopathy; 3 post-LASIK Diffuse Lamellar Keratitis (DLK); 2 fungal keratitis; 2 epithelial in-growth and 1 corneal pigmented lesion. In vivo confocal microsopy allowed for correct diagnoses in 22 cases (73%) , whereas in 8 the diagnosis was incorrect. Patients with Acanthamoeba keratitis and fungal keratitis were correctly diagnosed. DLK patients were generically diagnosed as having “corneal scarring” and subsequently correctly diagnosed through examining the color picture. Out of the 8 misdiagnosed cases, 7 were correctly diagnosed once the color picture of the cornea was provided. One patient affected by Map-Dot-fingerprint was mis-diagnosed as suffering from Acanthamoeba keratitis even after examining the color picture.
Conclusions:
In vivo confocal microscopy is a non-invasive examination that provides relevant information on corneal anatomy. As it provides coronal images of the cornea, it is comparable to biopsies. Nevertheless, clinical pictures are instrumental to correct diagnosis. UFOs were mostly misinterpreted as Acanthamoeba cysts, probably due to the fact that they are easily identified by this tool thereby yielding a high rate of false positive findings. One single image of in vivo confocal microscopy is deemed insufficient if a correct diagnoses are to be made, as findings may well overlap in different diseases. Nevertheless, when integrated with bio-microscopical findings, this tool is essential if prompt, accurate and non-invasive diagnoses of corneal disease are to be formulated.
Keywords: 596 microscopy: confocal/tunneling •
479 cornea: clinical science •
550 imaging/image analysis: clinical