The application of artificial intelligence (AI) in the diagnosis and management of RCES is a promising yet underexplored area of research

The gap

Evidence profile

Sourced from the future work of the source papers, classified as general, drawn from work published between 2025 and 2026, spanning 3 journals. Those papers have been cited 8 times in total.

Supporting evidence — 3 representative gaps

• Artificial intelligence in refractive surgery: progress, challenges, and future directions (2026) · Frontiers in Cell and Developmental Biology · doi

  Overall, the application of AI technology in the field of refractive surgery has made significant progress, with far-reaching impacts on improving surgical accuracy, formulating personalized surgical plans, and predicting postoperative complications. With the rapid development of big data and AI algorithms, the application of AI in refractive surgery will continue to deepen, advancing toward greater precision and intelligence. In the future, the application of AI will be in the following aspects: (1) AI-Driven primarily reflected Frontiers in Cell and Developmental Biology 08 frontiersin.org Ji et al. 10.3389/fcell.2026.1824307 surgical planning approaches will Preoperative Screening and Candidate Selection. In the future, with the accumulation of more high-quality clinical data, AI screening models will be continuously optimized. By integrating diverse examination data, such as corneal topography and corneal thickness, AI will form a more intelligent screening system. AI can not only improve screening accuracy but also significantly enhance diagnostic efficiency, enabling clinicians to identify high- risk patients more quickly and provide them with personalized treatment recommendations. As AI continues to advance in the field of medical image recognition, fully automated preoperative screening is expected to become a reality. This will allow clinicians to obtain comprehensive patient evaluation reports more efficiently, thereby optimizing clinical decision-making and improving the convenience and accuracy of screening. (2) AI- Assisted Surgical Planning and Individualized Treatment. In the future, AI will not only assist clinicians in surgical planning but also enhance the accuracy of surgical recommendations through continuous training and optimization. In complex cases, AI may even provide multiple surgical options. By learning the relationship between extensive preoperative examination data and postoperative refractive outcomes, AI models can accurately predict optimal surgical parameters. Furthermore, AI is expected to refine the calculation methods for surgical parameters and integrate real- time image analysis technology, enabling dynamic adjustments to surgical strategies during the procedure. This will further improve the precision and personalization of refractive surgery. Such intelligent enhance postoperative visual quality and reduce incidence of complications. (3) Application of AI in Predicting Postoperative Complications. The application of AI technology in postoperative management of refractive surgery is progressively deepening, offering new methods to improve postoperative safety and reduce complication rates. With the continued advancement of AI in predicting postop

  generalfuture work
  Keywords: surgical postoperative screening application refractive surgery accuracy technology predicting complications future planning preoperative improve enhance
• Seeing Through the Layers: Imaging Innovations in Recurrent Corneal Erosion Syndrome (2026) · Current Ophthalmology Reports · doi

  The application of artificial intelligence (AI) in the diagnosis and management of RCES is a promising yet underexplored area of research. While AI has demon- strated significant utility in the detection and monitor- ing of other corneal diseases such as keratoconus and infectious keratitis, its use in RCES-related imaging remains in its infancy. Future research should prioritize the development and validation of artificial intelligence (AI)-driven algorithms capable of synthesizing complex data from multimodal imaging platforms. By leveraging deep learning to analyze high-resolution imaging, these systems can identify discrete microstructural corneal changes that serve as biomarkers of disease recurrence, thus aiding in lowering disease morbidity. in corneal

  generalfuture work
  Keywords: corneal imaging artificial intelligence rces disease application diagnosis management promising underexplored area demon strated significant
• Navigating the Challenges of Acanthamoeba Keratitis: Current Trends and Future Directions (2025) · Life · cited 8× · doi

  4.1. Use of Artificial Intelligence in AK Diagnosis The diagnosis of AK is based on clinical findings (with a high index of suspicion), corneal tissue culture and biopsy, confocal microscopy, and PCR. These diagnostic tech- niques rely on experience, time, and cost and result in tissue loss. In low-resource or re- mote settings, diagnosis can be particularly challenging. Artificial intelligence (AI) models have been utilized in the diagnostic workup of AK, especially in the realms of image anal- ysis, early detection with IVCM image analysis, and within decision support systems to generate a risk score. This enables less dependence on human expert interpretation, al- lowing better accessibility. AI models can analyze IVCM images to detect patterns indicative of AK, potentially leading to earlier and more accurate diagnoses. Essalat et al. utilized a deep learning model that showed good ability to distinguish between Acanthamoeba (sensitivity of 91% and specificity of 98%) and fungal keratitis (sensitivity of 97% and specificity of 96%) [94]. A recent retrospective cohort study comprising 3312 IVCM images from 17 culture-posi- tive AK patients was used to train a deep learning model that resulted in sensitivity and specificity of 76%, respectively [95]. Another study by Koyama et al. applied techniques of facial recognition on slit-lamp images (different angles, resolution, and levels of illumi- nation) to diagnose Acanthamoeba with an accuracy rate of 98% [96]. Combining several different AI models can improve diagnostic performance. Zhang et al. utilized combined AI models and had an accuracy of 83.8% for AK diagnosis, which was better than the Life 2025, 15, 933 10 of 15 accuracy of their study’s invited corneal experts [97]. Given the diversity of AK presenta- tion globally, more validation is required across diverse datasets to improve AI models for diagnosis and treatment. Future research will be aimed toward developing AI tools that can be utilized in the process of early diagnosis and treatment, and determining how to integrate these into the clinical workflows to facilitate daily practice. 4.2. Standardized Treatment Protocols Protocolized treatment is established in medicine for severe infections such as sepsis, with clear improvement in clinical outcomes for patients [98]. Currently, it is rare outside of randomized controlled trials for protocolized treatments to be given to keratitis pa- tients. The prolonged treatment course for AK, with its individualized treatment that is practitioner-dependent and with no standardized termination protocols, makes outcome evaluation challenging. Sharma et al. published a protocol for fungal keratitis (Topical, Systemic, and Targeted Therapy; TST) that has demonstrated good outcomes compared to individualized treatment [99]. Recently, Dart et al. showed a 1.59-fold improvement in clinical outcomes for AK patients using a protocol-based treatment plan compared with individualized treatment [100]. With the advent of standardized preparations of anti- amoebic drugs becoming available, using standardized treatment protocols will allow bet- ter evaluation of treatment outcomes. 4.3. Cross-Linking Corneal collagen cross-linking has been applied to treat microbial keratitis, including AK (PACK-CXL; photoactivated chromophore for keratitis corneal cross-linking). The proposed mechanism of action entails the impairment of the Acanthamoeba cell mem- brane and nucleic acids by reactive oxygen species [101]. Atalay et al. established that PACK-CXL with riboflavin 0.1% and 0.25% exhibited no amoebicidal impact, whereas rose-bengal-mediated PACK-CXL did have in vitro anti-amoebic action against Acan- thamoeba castellanii [102]. In a subsequent animal model, the same authors showed that rose bengal PACK-CXL effectively decreased the AK clinical severity and AK load [103]. A recent meta-analysis comprising 46 studies totaling 435 patients concluded that PACK- CXL helped to expedite healing as an adjuvant treatment. Still, there was insufficient evi- dence for its primary use in AK [104]. Histologically, it has been shown that, despite PACK-CXL treatment, AK cysts can persist within the corneal stroma [105]. Hence, PACK-CXL is not routinely used by AK.

  generalfuture work
  Keywords: treatment pack diagnosis clinical corneal models keratitis utilized patients standardized outcomes diagnostic ivcm images model

Questions about this gap