Orbital and conjunctival nontuberculous mycobacteria infection.

A 64-year-old female developed refractory red-eye with itching and watery discharge 2 weeks after being injured by a comb in the left eye. It presented as diffuse pinkish thickening of the bulbar conjunctiva. Biopsy and histological examinations revealed granulomatous inflammation with microgranuloma. Acid-fast-positive bacilli were found within the tissue, which was identified by culture 5 weeks later as Mycobacterium Abscessus. The orbital computed tomography with contrast medium showed irregular enhancement with an ill-defined margin along the inferior sclera. Due to symptomatic and recurrent bulbar conjunctival thickening and abscess-like lesion formations, wide excision of the conjunctival and orbital granuloma with amniotic membrane transplantation was performed twice. Conjunctiva inflammation subsided after the surgical treatment was combined with 4 months of topical and parenteral antimycobacterial treatment. The presentation, diagnosis, and treatment of ocular nontuberculous mycobacterial (NTM) infection will be discussed in this article. NTM can cause infections of all adnexal and ocular tissues in patients with ocular trauma or surgical history. The pathological findings were granulomatous inflammation without true caseating. Periocular cutaneous, adnexal, and orbital NTM infections remain rare and require surgical debridement and long-term parenteral antibiotic therapy.

Necessity of Local Modification for Deep Learning Algorithms to Predict Diabetic Retinopathy.

Deep learning (DL) algorithms are used to diagnose diabetic retinopathy (DR). However, most of these algorithms have been trained using global data or data from patients of a single region. Using different model architectures (e.g., Inception-v3, ResNet101, and DenseNet121), we assessed the necessity of modifying the algorithms for universal society screening. We used the open-source dataset from the Kaggle Diabetic Retinopathy Detection competition to develop a model for the detection of DR severity. We used a local dataset from Taipei City Hospital to verify the necessity of model localization and validated the three aforementioned models with local datasets. The experimental results revealed that Inception-v3 outperformed ResNet101 and DenseNet121 with a foreign global dataset, whereas DenseNet121 outperformed Inception-v3 and ResNet101 with the local dataset. The quadratic weighted kappa score (κ) was used to evaluate the model performance. All models had 5-8% higher κ for the local dataset than for the foreign dataset. Confusion matrix analysis revealed that, compared with the local ophthalmologists' diagnoses, the severity predicted by the three models was overestimated. Thus, DL algorithms using artificial intelligence based on global data must be locally modified to ensure the applicability of a well-trained model to make diagnoses in clinical environments.

Diagnosis of microsporidial keratitis with in vivo confocal microscopy.

As a rare cause of microbial keratitis, microsporidial keratitis (MK) is first described in a patient with acquired immunodeficiency syndrome. As increased use of topical steroid creates a localized immunosuppressive environment of the eyes, MK occurs more commonly than expected in immunocompetent patients nowadays. Owing to initial insidious growth of pathogens and nonspecific ocular symptoms of infected patients, its frequent misdiagnosis has posed a major clinical challenge in recent decades. Without appropriate treatments, MK can progress deeply into corneal stroma, anterior and posterior segments, subsequently deteriorating vision severely and ultimately requiring corneal transplant. Related risk factors for the occurrence of MK in immunocompetent individuals include contact lens wear, topical steroid use, previous corneal trauma, and a history of laser refractive surgery. The conventional standard of MK diagnosis is based on a tissue biopsy by superficial corneal scrapping. In vivo confocal laser scanning microscopy can obtain images through the cornea in a plane paralleling to the vertical axis. This approach provides an effective method of identifying tissue layers that correspond to corneal histologic structures. This current study investigates the efficacy of \textit{in vivo} confocal laser scanning microscopy in diagnosing MK in immunocompetent patients. The clinical presentations of enrolled patients, including features of slit lamp biomicroscopy and the histopathological results of corneal scrapping, were described. In these patients, the confocal microscopy identified multiple small intracellular hyper-reflective dots in the cytoplasm of corneal epithelial cells and stromal keratocytes. Additionally, the confocal microscopic images clearly revealed the enhanced cytoplasm of cell with intracellular round hyper-reflective dots. The size and morphology of hyper-reflective dots were compatible with the spores of microsporidia found in corneal tissue. Moreover, vision recovered after topical use of antimicrobial medicine. This observation suggests that in vivo confocal laser scanning microscopy provides a rapid, non-invasive, and high resolution scheme for diagnosing MK. In addition to diminishing the risk of secondary infection from epithelial defect created by superficial debridement, this approach facilitates early diagnosis and appropriate treatments. Furthermore, from a series of images taken during the clinical courses, this method is highly promising for use in monitoring treatment effects and identifying the recurrence of MK.