A generalist medical language model for disease diagnosis assistance.

The delivery of accurate diagnoses is crucial in healthcare and represents the gateway to appropriate and timely treatment. Although recent large language models (LLMs) have demonstrated impressive capabilities in few-shot or zero-shot learning, their effectiveness in clinical diagnosis remains unproven. Here we present MedFound, a generalist medical language model with 176 billion parameters, pre-trained on a large-scale corpus derived from diverse medical text and real-world clinical records.

Self-improving generative foundation model for synthetic medical image generation and clinical applications.

In many clinical and research settings, the scarcity of high-quality medical imaging datasets has hampered the potential of artificial intelligence (AI) clinical applications. This issue is particularly pronounced in less common conditions, underrepresented populations and emerging imaging modalities, where the availability of diverse and comprehensive datasets is often inadequate. To address this challenge, we introduce a unified medical image-text generative model called MINIM that is capable of synthesizing medical images of various organs across various imaging modalities based on textual instructions.

Ranked Importance of Visual Function Outcome Measures in Choroideremia Clinical Trials.

Purpose: Clinical trials of novel therapies for choroideremia require robust and clinically meaningful visual function outcome measures. Best-corrected visual acuity (BCVA) is mostly insensitive to changes in disease state, until late stages, and hence also to potential therapeutic gains after gene therapies. While the insensitivity of BCVA as an effective outcome measure is common wisdom, its low importance has not been rigorously demonstrated in the literature.

Real-world six-month outcomes in patients switched to faricimab following partial response to anti-VEGF therapy for neovascular age-related macular degeneration and diabetic macular oedema.

Background: Landmark studies reported on faricimab efficacy and safety predominantly in treatment naïve patients, but outcomes following switch from other anti-VEGF therapies are lacking. We evaluated patients switched to faricimab who had previously shown a partial response to other anti-VEGF injections for neovascular age-related macular degeneration (nAMD) and diabetic macular oedema (DMO).

Methods: Retrospective study at the Oxford Eye Hospital.

Multimodal Evaluation and Management of Wagner Syndrome-Three Patients from an Affected Family.

Wagner syndrome is a rare autosomal dominant vitreoretinopathy caused by mutations in chondroitin sulphate proteoglycan 2 (CSPG2)/Versican (VCAN). Here, we present a retrospective case series of a family pedigree with genetically confirmed Wagner syndrome (heterozygous VCAN exon 8 deletion), as follows: a 34-year-old mother (P1), 12-year-old daughter (P2), and a 2-year-old son (P3). The phenotype included early-onset cataract (P1), optically empty vitreous with avascular membranes (P1, 2), nasal dragging of optic nerve heads associated with foveal hypoplasia (P1, 2), tractional retinoschisis on optical coherence tomography (P2), and peripheral circumferential vitreo-retinal interface abnormality resembling white-without-pressure (P3) progressing to pigmented chorio-retinal atrophy (P1, 2).

Expanding the genotypic and phenotypic spectra with a novel variant in the ciliopathy gene, , associated with selective cone degeneration.

Background: (Cilia and Flagella Associated Protein 410) encodes a protein that has an important role in the development and function of cilia. In ophthalmology, pathogenic variants in have been described in association with cone rod dystrophy, retinitis pigmentosa, with or without macular staphyloma, or with systemic abnormalities such as skeletal dysplasia and amyotrophic lateral sclerosis. Herein, we report a consanguineous family with a novel homozygous c.

Single-cell transcriptomic analysis of retinal immune regulation and blood-retinal barrier function during experimental autoimmune uveitis.

Uveitis is characterised by breakdown of the blood-retinal barrier (BRB), allowing infiltration of immune cells that mediate intraocular inflammation, which can lead to irreversible damage of the neuroretina and the loss of sight. Treatment of uveitis relies heavily on corticosteroids and systemic immunosuppression due to limited understanding of disease pathogenesis. We performed single-cell RNA-sequencing of retinas, as well as bulk RNA-sequencing of retinal pigment epithelial (RPE) cells from mice with experimental autoimmune uveitis (EAU) versus healthy control.

Accuracy and safety of an autonomous artificial intelligence clinical assistant conducting telemedicine follow-up assessment for cataract surgery.

Background: Artificial intelligence deployed to triage patients post-cataract surgery could help to identify and prioritise individuals who need clinical input and to expand clinical capacity. This study investigated the accuracy and safety of an autonomous telemedicine call (Dora, version R1) in detecting cataract surgery patients who need further management and compared its performance against ophthalmic specialists.

Methods: 225 participants were recruited from two UK public teaching hospitals after routine cataract surgery between 17 September 2021 and 31 January 2022.

Robot-Assisted Eye Surgery: A Systematic Review of Effectiveness, Safety, and Practicality in Clinical Settings.

Purpose: Surgical innovation in ophthalmology is impeded by the physiological limits of human motion, and robotic assistance may facilitate an expansion of the surgical repertoire. We conducted a systematic review to identify ophthalmic procedures in which robotic systems have been trialled, evaluate their performance, and explore future directions for research and development of robotic techniques.

Methods: The Cochrane Library, Embase, MEDLINE, Scopus, and Web of Science were searched.

Optimized glycemic control of type 2 diabetes with reinforcement learning: a proof-of-concept trial.

The personalized titration and optimization of insulin regimens for treatment of type 2 diabetes (T2D) are resource-demanding healthcare tasks. Here we propose a model-based reinforcement learning (RL) framework (called RL-DITR), which learns the optimal insulin regimen by analyzing glycemic state rewards through patient model interactions. When evaluated during the development phase for managing hospitalized patients with T2D, RL-DITR achieved superior insulin titration optimization (mean absolute error (MAE) of 1.

Deep-learning-enabled protein-protein interaction analysis for prediction of SARS-CoV-2 infectivity and variant evolution.

Host-pathogen interactions and pathogen evolution are underpinned by protein-protein interactions between viral and host proteins. An understanding of how viral variants affect protein-protein binding is important for predicting viral-host interactions, such as the emergence of new pathogenic SARS-CoV-2 variants. Here we propose an artificial intelligence-based framework called UniBind, in which proteins are represented as a graph at the residue and atom levels.

PHOTORECEPTOR DAMAGE IN TERSON SYNDROME.

Purpose: To describe photoreceptor damage in patients with Terson syndrome as a potential cause for inconsistent clinical outcomes.

Methods: Clinical evaluation and retinal imaging in six patients.

Results: Four patients were women and two men, with an average age of 46.

Gene-agnostic therapeutic approaches for inherited retinal degenerations.

Inherited retinal diseases (IRDs) are associated with mutations in over 250 genes and represent a major cause of irreversible blindness worldwide. While gene augmentation or gene editing therapies could address the underlying genetic mutations in a small subset of patients, their utility remains limited by the great genetic heterogeneity of IRDs and the costs of developing individualised therapies. Gene-agnostic therapeutic approaches target common pathogenic pathways that drive retinal degeneration or provide functional rescue of vision independent of the genetic cause, thus offering potential clinical benefits to all IRD patients.

A Network Meta-Analysis of Retreatment Rates following Bevacizumab, Ranibizumab, Aflibercept, and Laser for Retinopathy of Prematurity.

Topic: To compare bevacizumab, ranibizumab, aflibercept, and laser treatment as primary therapies for retinopathy of prematurity (ROP) in terms of retreatment rate.

Clinical Relevance: Anti-VEGF agents are increasingly used as primary treatment for ROP and may provide superior outcomes compared with laser in posterior disease. Head-to-head comparisons between different anti-VEGFs are lacking.

A deep-learning system predicts glaucoma incidence and progression using retinal photographs.

BackgroundDeep learning has been widely used for glaucoma diagnosis. However, there is no clinically validated algorithm for glaucoma incidence and progression prediction. This study aims to develop a clinically feasible deep-learning system for predicting and stratifying the risk of glaucoma onset and progression based on color fundus photographs (CFPs), with clinical validation of performance in external population cohorts.

First-in-Human Robot-Assisted Subretinal Drug Delivery Under Local Anesthesia.

Purpose: To report the results of a first-in-human study using a robotic device to assist subretinal drug delivery in patients undergoing vitreoretinal surgery for macular hemorrhage.

Design: Double-armed, randomized controlled surgical trial (ClinicalTrials.gov identifier: NCT03052881).

Characterizing the cellular immune response to subretinal AAV gene therapy in the murine retina.

Although adeno-associated viral (AAV) vector-mediated retinal gene therapies have demonstrated efficacy, the mechanisms underlying dose-dependent retinal inflammation remain poorly understood. Here, we present a quantitative analysis of cellular immune response to subretinal AAV gene therapy in mice using multicolor flow cytometry with a panel of key immune cell markers. A significant increase in CD45 retinal leukocytes was detected from day 14 post-subretinal injection of an AAV8 vector (1 × 10 genome copies) encoding green fluorescent protein (GFP) driven by a ubiquitous promoter.

Safety and Acceptability of a Natural Language Artificial Intelligence Assistant to Deliver Clinical Follow-up to Cataract Surgery Patients: Proposal.

Background: Due to an aging population, the demand for many services is exceeding the capacity of the clinical workforce. As a result, staff are facing a crisis of burnout from being pressured to deliver high-volume workloads, driving increasing costs for providers. Artificial intelligence (AI), in the form of conversational agents, presents a possible opportunity to enable efficiency in the delivery of care.

Expression of Rab Prenylation Pathway Genes and Relation to Disease Progression in Choroideremia.

Purpose: Choroideremia results from the deficiency of Rab Escort Protein 1 (REP1), encoded by CHM, involved in the prenylation of Rab GTPases. Here, we investigate whether the transcription and expression of other genes involved in the prenylation of Rab proteins correlates with disease progression in a cohort of patients with choroideremia.

Methods: Rates of retinal pigment epithelial area loss in 41 patients with choroideremia were measured using fundus autofluorescence imaging for up to 4 years.

Interactions between Apolipoprotein E Metabolism and Retinal Inflammation in Age-Related Macular Degeneration.

Age-related macular degeneration (AMD) is a multifactorial retinal disorder that is a major global cause of severe visual impairment. The development of an effective therapy to treat geographic atrophy, the predominant form of AMD, remains elusive due to the incomplete understanding of its pathogenesis. Central to AMD diagnosis and pathology are the hallmark lipid and proteinaceous deposits, drusen and reticular pseudodrusen, that accumulate in the subretinal pigment epithelium and subretinal spaces, respectively.

CRISPR genome engineering for retinal diseases.

Novel gene therapy treatments for inherited retinal diseases have been at the forefront of translational medicine over the past couple of decades. Since the discovery of CRISPR mechanisms and their potential application for the treatment of inherited human conditions, it seemed inevitable that advances would soon be made using retinal models of disease. The development of CRISPR technology for gene therapy and its increasing potential to selectively target disease-causing nucleotide changes has been rapid.