Artificial intelligence for retinal diseases.

PURPOSE: To discuss the worldwide applications and potential impact of artificial intelligence (AI) for the diagnosis, management and analysis of treatment outcomes of common retinal diseases. METHODS: We performed an online literature review, using PubMed Central (PMC), of AI applications to evaluate and manage retinal diseases. Search terms included AI for screening, diagnosis, monitoring, management, and treatment outcomes for age-related macular degeneration (AMD), diabetic retinopathy (DR), retinal surgery, retinal vascular disease, retinopathy of prematurity (ROP) and sickle cell retinopathy (SCR). Additional search terms included AI and color fundus photographs, optical coherence tomography (OCT), and OCT angiography (OCTA). We included original research articles and review articles. RESULTS: Research studies have investigated and shown the utility of AI for screening for diseases such as DR, AMD, ROP, and SCR. Research studies using validated and labeled datasets confirmed AI algorithms could predict disease progression and response to treatment. Studies showed AI facilitated rapid and quantitative interpretation of retinal biomarkers seen on OCT and OCTA imaging. Research articles suggest AI may be useful for planning and performing robotic surgery. Studies suggest AI holds the potential to help lessen the impact of socioeconomic disparities on the outcomes of retinal diseases. CONCLUSIONS: AI applications for retinal diseases can assist the clinician, not only by disease screening and monitoring for disease recurrence but also in quantitative analysis of treatment outcomes and prediction of treatment response. The public health impact on the prevention of blindness from DR, AMD, and other retinal vascular diseases remains to be determined.

EYE-Llama, an in-domain large language model for ophthalmology.

Background: Training Large Language Models (LLMs) with in-domain data can significantly enhance their performance, leading to more accurate and reliable question-answering (QA) systems essential for supporting clinical decision-making and educating patients. Methods: This study introduces LLMs trained on in-domain, well-curated ophthalmic datasets. We also present an open-source substantial ophthalmic language dataset for model training. Our LLMs (EYE-Llama), first pre-trained on an ophthalmology-specific dataset, including paper abstracts, textbooks, EyeWiki, and Wikipedia articles. Subsequently, the models underwent fine-tuning using a diverse range of QA datasets. The LLMs at each stage were then compared to baseline Llama 2, ChatDoctor, and ChatGPT (GPT3.5) models, using four distinct test sets, and evaluated quantitatively (Accuracy, F1 score, and BERTScore) and qualitatively by two ophthalmologists. Results: Upon evaluating the models using the American Academy of Ophthalmology (AAO) test set and BERTScore as the metric, our models surpassed both Llama 2 and ChatDoctor in terms of F1 score and performed equally to ChatGPT, which was trained with 175 billion parameters (EYE-Llama: 0.57, Llama 2: 0.56, ChatDoctor: 0.56, and ChatGPT: 0.57). When evaluated on the MedMCQA test set, the fine-tuned models demonstrated a higher accuracy compared to the Llama 2 and ChatDoctor models (EYE-Llama: 0.39, Llama 2: 0.33, ChatDoctor: 0.29). However, ChatGPT outperformed EYE-Llama with an accuracy of 0.55. When tested with the PubmedQA set, the fine-tuned model showed improvement in accuracy over both the Llama 2, ChatGPT, and ChatDoctor models (EYE-Llama: 0.96, Llama 2: 0.90, ChatGPT: 0.93, ChatDoctor: 0.92). Conclusion: The study shows that pre-training and fine-tuning LLMs like EYE-Llama enhances their performance in specific medical domains. Our EYE-Llama models surpass baseline Llama 2 in all evaluations, highlighting the effectiveness of specialized LLMs in medical QA systems. (Funded by NEI R15EY035804 (MNA) and UNC Charlotte Faculty Research Grant (MNA).).

Federated learning for diagnosis of age-related macular degeneration.

This paper presents a federated learning (FL) approach to train deep learning models for classifying age-related macular degeneration (AMD) using optical coherence tomography image data. We employ the use of residual network and vision transformer encoders for the normal vs. AMD binary classification, integrating four unique domain adaptation techniques to address domain shift issues caused by heterogeneous data distribution in different institutions. Experimental results indicate that FL strategies can achieve competitive performance similar to centralized models even though each local model has access to a portion of the training data. Notably, the Adaptive Personalization FL strategy stood out in our FL evaluations, consistently delivering high performance across all tests due to its additional local model. Furthermore, the study provides valuable insights into the efficacy of simpler architectures in image classification tasks, particularly in scenarios where data privacy and decentralization are critical using both encoders. It suggests future exploration into deeper models and other FL strategies for a more nuanced understanding of these models' performance. Data and code are available at https://github.com/QIAIUNCC/FL_UNCC_QIAI.

Chlorine resistance property improvement of polyamide reverse osmosis membranes through cross-linking degree increment.

Highly permeable polyamide reverse osmosis (RO) membranes are desirable for reducing the energy burden and ensuring future water resources in arid and semiarid regions. One notable drawback of thin film composite (TFC) polyamide RO/NF membranes is the polyamide's sensitivity to degradation by free chlorine, the most used biocide in water purification trains. This investigation demonstrated a significant increase in the crosslinking-degree parameter by the m-phenylenediamine (MPD) chemical structure extending in the thin film nanocomposite (TFN) membrane without adding extra MPD monomers to enhance the chlorine resistance and performance. Membrane modification was carried out according to monomer ratio changes and Nanoparticle embedding into the PA layer approaches. A new class of TFN-RO membranes incorporating novel aromatic amine functionalized (AAF)-MWCNTs embedded into the polyamide (PA) layer was introduced. A purposeful strategy was carried out to use cyanuric chloride (2,4,6-trichloro-1,3,5-triazine) as an intermediate functional group in the AAF-MWCNTs. Thus, amidic nitrogen, connected to benzene rings and carbonyl groups, assembles a structure similar to the standard PA, consisting of MPD and trimesoyl chloride. The resulting AAF-MWCNTs were mixed in the aqueous phase during the interfacial polymerization to increase the susceptible positions to chlorine attack and improve the crosslinking degree in the PA network. The characterization and performance results of the membrane demonstrated an increase in ion selectivity and water flux, impressive stability of salt rejection after chlorine exposure, and improved antifouling performance. This purposeful modification resulted in overthrowing two tradeoffs; i) high crosslink density-water flux and ii) salt rejection-permeability. The modified membrane demonstrated ameliorative chlorine resistance relative to the pristine one, with twice the increase in crosslinking degree, more than four times the enhancement of the oxidation resistance, negligible reduction in the salt rejection (0.83 %), and only 5 L/m2.h flux loss following a rigorous static chlorine exposure of 500 ppm.h under acidic conditions. The excellent performance of new chlorine resistant TNF RO membranes fabricated via AAF-MWCNTs together with the facile membrane manufacturing process offered the possibility of postulating them in the desalination field, which could eventually help the current freshwater supply challenge.

Impact of a new functionalization of multiwalled carbon nanotubes on antifouling and permeability of PVDF nanocomposite membranes for dye wastewater treatment.

Here, novel hydroxyl and carboxyl functionalized multiwalled carbon nanotubes (AHF-MWCNT and ACF-MWCNT) were successfully synthesized and introduced for modification and antifouling improvement of the PVDF membrane. The blending effect of AHF-MWCNT and ACF-MWCNT on the morphology and surface properties of the PVDF membrane was explored by SEM, AFM, water contact angle, and zeta potential analysis. The results indicated that the membrane surface has become more hydrophilic, smoother as well more negative. In addition, the overall porosity and mean pore radius are increased by MWCNTs embedding. The filtration performance, antifouling and dye separation of the nanocomposite membranes were improved by adding any amounts of AHF-MWCNT and ACF-MWCNT in the PVDF membrane matrix. The carboxylic modification presented better performance than the hydroxyl functionalization. The 0.1 wt% ACF-MWCNT blended membrane presented an optimum performance with 46 L m-2 h-1 bar-1 permeability, 93% FRR, and 97.3% dye rejection. Consequently, embedding functionalized MWCNT in the PVDF membrane matrix was led to improvement of membrane characteristics and enhancement of pure water flux, antifouling feature, and dye separation. So, the functionalized MWCNT could be a promising additive for the PVDF membrane modification.