Evaluating the potential of ChatGPT-4 in ophthalmology: The good, the bad and the ugly.

There is growing interest nowadays for artificial intelligence (AI) in all medical fields. Beyond the direct medical application of AI to medical data, generative AI such as "pre-trained transformer" (GPT) could significantly change the ophthalmology landscape, opening up new avenues for enhancing precision, productivity, and patient outcomes. At present, ChatGPT-4 has been investigated in various ways in ophthalmology for research, medical education, and support for clinical decisions purposes. This article intends to demonstrate the application of ChatGPT-4 within the field of ophthalmology by employing a 'mise en abime' approach. While we explore its potential to enhance the future of ophthalmology care, we will also carefully outline its current limitations and potential risks.

Intravitreal triamcinolone injections in non-arteritic anterior ischemic optic neuropathy - A retrospective report.

BACKGROUND: Non-arteritic anterior ischemic optic neuropathy (NAION) is a common cause of vision loss but no treatment has demonstrated its efficiency. A preliminary study showed an improvement on the visual acuity (VA) in a group of patients who received intravitreal administration of triamcinolone acetonide (IVTA) versus a non-treated group. In the present series, the visual outcome of IVTA in NAION was evaluated on a larger group of patients. METHODS: This retrospective, unmasked and non-randomized study took place at Reims University Hospital between 2009 and 2017. The data of consecutive patients presenting with isolated optic disc edema characteristic of recent NAION (<1month of visual acuity loss) were included. After informed consent, a single intravitreal injection of filtrated 4mg/0.1mL triamcinolone acetonide were administered. Twenty-seven control patients chose not to be injected and therefore served as controls. LogMar visual acuity (VA), VA rating (VAR) (1 line=0.1LogMAR=5 VAR letters), retinal nerve fiber layer thickness assessed by OCT and static visual field were evaluated at presentation, after 7days, after 3months and after 6months. RESULTS: Sixty-eight patients with NAION were evaluated. Forty-one received IVTA, 29 were injected within 15days after the onset of symptoms and 12 after 15days. There was a higher proportion of patients improving VA of 2 lines or more (10 or more VAR letters) in the injected group (49%) compared with the non-injected group (11%, P=0.019). Among the patients injected before 15days, the proportion improving for 2 lines or more (55% vs. 11%, respectively, P=0.013) and for 3 lines or more (45% vs. 11%, respectively, P=0.035) were significantly higher than in the non-injected group. Also, comparing the VA at presentation with the VA after 6months in the injected eyes, it improved significantly (P=0.003) and also in the subgroup injected within 15days (P=0.0007) but not in the injected group after 15days (P=0.801). Visual field improvement was only observed in the subgroup of patients injected within 15days with a significant improvement of the mean deviation (dB) within 6months (P=0.015). CONCLUSIONS: This follow-up study confirms the results of the previous series displaying an apparent benefit of intravitreal steroids injected in the acute phase of NAION. Only patients receiving IVTA within 15days from onset of NAION have a significant improvement of VA and visual field during the follow-up period of 6months.

[Retrospective evaluation of residencies in private ophthalmology practice from 2015 to 2018 in Champagne-Ardenne]. / Évaluation rétrospective de stages d'interne d'ophtalmologie en cabinet libéral de 2015 à 2018 en Champagne-Ardenne.

INTRODUCTION: Between 2010 and 2018, the quota for admission to ophthalmology residencies increased by 50 % (106 in 2010 vs 150 in 2018). In order to accommodate this increasing number of residents, the University Hospital of Reims formulated an agreement with the Regional Health Agency in May 2015 enabling certain ophthalmologists in the private sector to train a resident within their private practice for a semester. We will present the results of three and one half years of this experience. METHODS: Two retrospective questionnaires were created and completed. One was addressed to the host supervisor, the other to the resident. Their objective was to evaluate and standardize the experience from both points of view. The following aspects were explored: the extent to which the environment was welcoming, the resident's clinical ability and progress, the resident's surgical ability and progress, time-management, resident-patient relations and respective assessments. We also requested information on the number of residents hosted by each private practice as well as the positive and negative aspects of the internship for both the supervisor and the intern. RESULTS: Between May 2015 and October 2018 (seven semesters), 12 residents from the University Hospital of Reims did an internship with one of the six proposed private sector ophthalmologists in the Champagne-Ardennes region. The residents were between their second and eighth semesters of professional training. Seven residents did their first or second semester of ophthalmology training in a private practice. The survey results led to the conclusion of a positive experience for both trainer and trainee regarding the progress made by the resident, both clinically and surgically. CONCLUSION: Given the growing number of ophthalmology residents and the limits of the training capacity of hospital-based residencies, private sector internships with practitioners actively involved in teaching increase the training capacity for residents and complement the conventional hospital training.

[Is Holladay 2 formula acurate enough for calculating intraocular lens power in non-standard eyes?] / Intérêt de la formule Holladay 2 dans le calcul de la puissance de l'implant intraoculaire pour les yeux de biométrie hors normes.

PURPOSE: To evaluate the benefit of the Holladay 2 formula versus the 3rd generation formulae in calculating the IOL power in eyes with non-standard axial length or keratometry before cataract surgery. PATIENTS: Retrospectiv study from January to December 2015. The inclusion critaeria were axial length (AL) <22mm or >26mm, or average keratometry <42 D or >46 D, dividing the patients in 4 groups respectively. The 7 parameters required to calculate the Holladay 2 formula were collected. The final refractive result was turned into spherical equivalent to calculate the optimal power retrospectively. Then, the results obtained using the other formulae were compared with the optimal IOL power. RESULTS: One hundred and twenty-six eyes operated by two surgeons were included. In the high AL group (n=32), the SRK/T was the most accurate formula; regarding the low AL group (n=36), the Hoffer Q and Holladay 2 formulae performed better; for the steep cornea group (n=27), the Hoffer Q, Haigis, Holladay 1 and 2 formulae were not different; last, the Holladay 1 and 2 were more accurate in the flat cornea group (n=33). CONCLUSION: In our study, the Holladay 2 formula does not seem to be better than the others for calculating IOL power in non-standard eyes. Preoperative eye features in such non-standard cases should be taken into account before the surgery to choose the more suitable formula.