Integrando la Inteligencia Artificial y los nuevos modelos de lenguaje en Oftalmología: Complicaciones del uso del aceite de silicona en cirugía vítreo-retiniana / Artificial Intelligence and new language models in Ophthalmology: Complications of the use of silicone oil in vitreoretinal surgery

La inteligencia artificial (IA) es una tecnología emergente que facilita tareas cotidianas y automatiza tareas en diversos campos como medicina. Sin embargo, la irrupción de un modelo de lenguaje en el mundo académico ha generado mucho interés. En esta comunicación se evalúa el potencial de Generative Pre-trained Transformer (ChatGPT), un modelo de lenguaje desarrollado por OpenAI; y de DALL-E 2, un generador de imágenes, en la escritura de artículos científicos en Oftalmología. El tema seleccionado es el de las complicaciones del uso de aceite de silicona (AS) en cirugía vítreo-retiniana. Se utilizó ChatGPT para generar un resumen y un artículo estructurado, sugerencias para un título y referencias bibliográficas. Se puede concluir que, a pesar del conocimiento demostrado, la precisión y fiabilidad científica en temas específicos es insuficiente para la generación automática de artículos con rigor científico. Por otro lado, todo científico debe conocer las posibles implicaciones éticas y legales de estas herramientas (AU)

Artificial intelligence (AI) is an emerging technology that facilitates everyday tasks and automates tasks in various fields such as medicine. However, the emergence of a language model in academia has generated a lot of interest. This paper evaluates the potential of ChatGPT, a language model developed by OpenAI, and DALL-E 2, an image generator, in the writing of scientific articles in ophthalmology. The selected topic is the complications of the use of silicone oil in vitreoretinal surgery. ChatGPT was used to generate an abstract and a structured article, suggestions for a title and bibliographical references. In conclusion, despite the knowledge demonstrated by this tool, the scientific accuracy and reliability on specific topics is insufficient for the automatic generation of scientifically rigorous articles. In addition, scientists should be aware of the possible ethical and legal implications of these tools (AU)

Artificial Intelligence and new language models in Ophthalmology: Complications of the use of silicone oil in vitreoretinal surgery.

Artificial intelligence (AI) is an emerging technology that facilitates everyday tasks and automates tasks in various fields such as medicine. However, the emergence of a language model in academia has generated a lot of interest. This paper evaluates the potential of ChatGPT, a language model developed by OpenAI, and DALL-E 2, an image generator, in the writing of scientific articles in ophthalmology. The selected topic is the complications of the use of silicone oil in vitreoretinal surgery. ChatGPT was used to generate an abstract and a structured article, suggestions for a title and bibliographical references. In conclusion, despite the knowledge demonstrated by this tool, the scientific accuracy and reliability on specific topics is insufficient for the automatic generation of scientifically rigorous articles. In addition, scientists should be aware of the possible ethical and legal implications of these tools.

Hair cortisol level as a molecular biomarker in retinitis pigmentosa patients.

PURPOSE: Retinitis pigmentosa (RP) patients commonly experience negative psychological states due to their progressive and unpredictable loss of vision and visual variations related to stress. The aim of this study was to examine hair cortisol concentrations (HCCs), which is usually associated with chronic stress, pretending to unveil possible associations between underlying psychological factors and disease severity in RP patients. METHODS: Seventy-eight RP patients and 148 healthy controls were included in this study. A complete ophthalmological exam was performed in all patients to grade into severity disease groups. Perceived stress and trait-anxiety were measured by the State-Trait Anxiety Inventory (STAI) questionnaire. RESULTS: Fifty-two (67%) patients had severe RP and 26 (33%) mild-moderate RP. Fifty-eight (58,9%) patients reported severely levels of stress and 18 (23.,1%) highly levels assessed by STAI questionnaire. RP patients exhibited higher HCCs (500.04 ± 120.99 pg/mg) than in controls (136.17 ± 60.51 pg/mg; p < 0.001). Severe RP patients had significant higher HCCs than mild-moderate patients differing in 274.27 pg/mg (p < 0.001). RP severity grade and perceived anxiety levels in the questionaries were not associated. Group differences were not affected by relevant covariates (age, grade of severity, stress status, and gender). CONCLUSIONS: HCC seems an effective biomarker associated with chronic stress in RP patients. This study shows that HCC in patients with RP are elevated compared to population-based controls, and association between HCC and RP severity was found. Future research is needed to characterize the effect of untreated negative psychological states on progression of the disease if any.

Predictores del riesgo de fractura en una población de mujeres postmenopáusicas mediante el procedimiento estadístico binario CART / Fracture risk predictors of a postmenopausal female population by binary statistical procedure

Objetivo: La principal consecuencia de la osteoporosis es la fractura por fragilidad asociada a elevada morbimortalidad. La predicción de la misma puede ayudar a identificar la población de mayor riesgo y establecer medidas de prevención. El objetivo de este estudio fue valorar la utilidad de diversos factores en su prevención comparando la densidad mineral ósea (DMO), el cálculo del riesgo absoluto de fractura con la herramienta FRAX® con y sin DMO, y los datos clínicos. Material y métodos: Se realizó un estudio longitudinal de 8 años de duración en una población de mujeres postmenopáusicas, osteoporóticas y no osteoporóticas. A todas ellas se les realizó una historia clínica protocolizada, DMO de columna y cadera, y el FRAX con y sin DMO. A los 8 años se identificaron las fracturas existentes. Además de realizar una estadística paramétrica y no paramétrica con SPSS 21.1, se realizó un método del árbol de clasificación y regresión (CART) para evaluar las posibles interacciones entre los factores de riesgo de fractura. Resultados: Se incluyeron 276 pacientes postmenopáusicas cuya edad media al inicio del estudio fue de 61,08±8,43 años y un índice de masa corporal (IMC) de 25,67±4,04. El 56,5% de las pacientes (n=156) fueron diagnosticadas de osteoporosis antes del inicio de nuestro estudio, y todas ellas fueron tratadas. Pasados los 8 años de seguimiento, 72 pacientes (24,6%) sufrieron fractura y 17 (6,2%) también sufrieron una segunda fractura. Los resultados del análisis CART nos mostraron que el principal factor de riesgo para sufrir una fractura osteoporótica tras 8 años de seguimiento fue el haber sufrido fracturas previas. Entre las pacientes que habían sufrido una fractura previa, el tener una DMO del cuello femoral menor de 0,67 fue el principal factor de riesgo. (AU)

Objetivo: The main consequences of osteoporosis are fragility fractures, associated with high morbimortality. The prediction of these fractures can help identify the most-at-risk population and implement preventive measures. The aim of this study was to assess the usefulness of multiple factors in their prevention, comparing the bone mineral density (BMD), the calculation of absolute risk of fracture using the tool FRAX® in the presence and absence of BMD, and the clinical data. Material and methods: An eight-year-duration longitudinal study was conducted on a postmenopausal female population, with and without osteoporosis. All of them were taken a standardised clinical history, spinal and hip BMD, and FRAX with and without BMD. Eight years later we identified the existent fractures. In addition to a parametric and non-parametric statistic in SPSS 21.1, we used the classification and regression tree (CART) method to assess possible interactions among fracture risk factors. Results: We studied 276 postmenopausal patients whose average age at the beginning of the study was 61.08±8.43 years-old and had a body mass index (BMI) of 25.67±4.04. 56.5% of the patients (n=156) were diagnosed with osteoporosis before the beginning of our study, and all of them were treated. After eight years of follow-up, 72 patients (24.6%) suffered a fracture and 17 patients (6.2%) also suffered a second one. The results of the CART analysis showed that the main risk factor to suffer an osteoporotic fracture after 8 years of following up is having preceding fractures. Having a femoral neck BMD lower than 0.67 was the main risk factor among those with a previous fracture. Conclusions: The use of a binary statistical procedure (CART) on a cohort of patients allow us identify those most at risk of fractures, according to clinical parameters and simple additional tests, in order to establish more effective therapeutic measures. (AU)

Association of IL1Β (-511 A/C) and IL6 (-174 G > C) polymorphisms with higher disease activity and clinical pattern of psoriatic arthritis.

The objective of this study is to analyze whether IL1ß (-511G > A) and IL6 (-174 G > C) polymorphisms are associated with inflammatory activity, radiographic damage or clinical pattern of psoriatic arthritis (PsA). One hundred twenty-five patients classified as PsA according to the Classification of Psoriatic Arthritis (CASPAR) criteria were included. Patients were stratified according to their clinical pattern at inclusion as peripheral, axial, or mixed involvement. Disease activity in peripheral or mixed forms was measured using the number of swollen and tender joints, pain analog visual scale, erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), and disease activity score 28 (DAS28). Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) was used for axial and mixed forms, as were pain visual analog scale, ESR and CRP. Radiographic damage was evaluated using a modified Sharp score and modified stoke ankylosing spondylitis spinal score (SASSSm). The polymorphisms for the promoter region of IL1ß (-511 G/A) and IL-6 (-174 G/C) were analyzed. The G allele of IL1B (-511G/A) polymorphism was associated with higher peripheral joint disease activity (OR 3.13; p < 0.0004; CI 95 % 1.43-6.82, p (corrected) <0.008), while the G allele of the IL6 (174G > C) polymorphism presented a strong trend to be associated with peripheral forms (70.86 %) (OR 1.89; p < 0.03; CI 95 % 1.06-3.39, p-corrected 0.05). In addition, this allele showed a lower association with HLA-B27 (15.78 %) compared with C allele (28.57 %) (OR 0.469; p = 0.02; CI 95 % 0.238-0.923, p-corrected 0.03). This study suggests that the G allele polymorphism of IL1B (-511 A/C) is associated with higher peripheral joint disease activity. On the other hand, the IL6 (-174 G/C) polymorphism showed a strong trend to be associated with the peripheral pattern of PsA.