RESUMO
Importance: Although augmenting large language models (LLMs) with knowledge bases may improve medical domain-specific performance, practical methods are needed for local implementation of LLMs that address privacy concerns and enhance accessibility for health care professionals. Objective: To develop an accurate, cost-effective local implementation of an LLM to mitigate privacy concerns and support their practical deployment in health care settings. Design, Setting, and Participants: ChatZOC (Sun Yat-Sen University Zhongshan Ophthalmology Center), a retrieval-augmented LLM framework, was developed by enhancing a baseline LLM with a comprehensive ophthalmic dataset and evaluation framework (CODE), which includes over 30â¯000 pieces of ophthalmic knowledge. This LLM was benchmarked against 10 representative LLMs, including GPT-4 and GPT-3.5 Turbo (OpenAI), across 300 clinical questions in ophthalmology. The evaluation, involving a panel of medical experts and biomedical researchers, focused on accuracy, utility, and safety. A double-masked approach was used to try to minimize bias assessment across all models. The study used a comprehensive knowledge base derived from ophthalmic clinical practice, without directly involving clinical patients. Exposures: LLM response to clinical questions. Main Outcomes and Measures: Accuracy, utility, and safety of LLMs in responding to clinical questions. Results: The baseline model achieved a human ranking score of 0.48. The retrieval-augmented LLM had a score of 0.60, a difference of 0.12 (95% CI, 0.02-0.22; P = .02) from baseline and not different from GPT-4 with a score of 0.61 (difference = 0.01; 95% CI, -0.11 to 0.13; P = .89). For scientific consensus, the retrieval-augmented LLM was 84.0% compared with the baseline model of 46.5% (difference = 37.5%; 95% CI, 29.0%-46.0%; P < .001) and not different from GPT-4 with a value of 79.2% (difference = 4.8%; 95% CI, -0.3% to 10.0%; P = .06). Conclusions and Relevance: Results of this quality improvement study suggest that the integration of high-quality knowledge bases improved the LLM's performance in medical domains. This study highlights the transformative potential of augmented LLMs in clinical practice by providing reliable, safe, and practical clinical information. Further research is needed to explore the broader application of such frameworks in the real world.
RESUMO
AIM: To investigate the association of floor area ratio (FAR), an indicator of built environments, and myopia onset. METHODS: This prospective cohort study recruited 136 753 children aged 6-10 years from 108 schools in Shenzhen, China at baseline (2016-2017). Refractive power was measured with non-cycloplegic autorefraction over a 2-year follow-up period. FAR was objectively evaluated using geographical information system technology. Mixed-effects logistic regression models were constructed to examine the association of FAR with a 2-year cumulative incidence of myopia among individuals without baseline myopia; multiple linear regression model, with a 2-year cumulative incidence rate of myopia at each school. RESULTS: Of 101 624 non-myopic children (56.3% boys; mean (SE) age, 7.657±1.182 years) included in the study, 26 391 (26.0%) of them developed myopia after 2 years. In the individual-level analysis adjusting for demographic, socioeconomic and greenness factors, an IQR in FAR was associated with a decreased risk of 2-year myopia incidence (OR 0.898, 95% CI 0.866 to 0.932, p<0.001). Similar findings were observed in the analysis additionally adjusted for genetic and behavioural factors (OR 0.821, 95% CI 0.766 to 0.880, p<0.001). In the school-level, an IQR increase in FAR was found to be associated with a 2.0% reduction in the 2-year incidence rate of myopia (95% CI 1.3% to 2.6%, p<0.001). CONCLUSIONS: Exposure to higher FAR was associated with a decreased myopia incidence, providing insights into myopia prevention through school built environments in China.
RESUMO
Fundus fluorescein angiography (FFA) examinations are widely used in the evaluation of fundus disease conditions to facilitate further treatment suggestions. Here, we present a protocol for performing deep learning-based FFA image analytics with classification and segmentation tasks. We describe steps for data preparation, model implementation, statistical analysis, and heatmap visualization. The protocol is applicable in Python using customized data and can achieve the whole process from diagnosis to treatment suggestion of ischemic retinal diseases. For complete details on the use and execution of this protocol, please refer to Zhao et al.1.
RESUMO
BACKGROUND/AIMS: The aim of this study was to develop and evaluate digital ray, based on preoperative and postoperative image pairs using style transfer generative adversarial networks (GANs), to enhance cataractous fundus images for improved retinopathy detection. METHODS: For eligible cataract patients, preoperative and postoperative colour fundus photographs (CFP) and ultra-wide field (UWF) images were captured. Then, both the original CycleGAN and a modified CycleGAN (C2ycleGAN) framework were adopted for image generation and quantitatively compared using Frechet Inception Distance (FID) and Kernel Inception Distance (KID). Additionally, CFP and UWF images from another cataract cohort were used to test model performances. Different panels of ophthalmologists evaluated the quality, authenticity and diagnostic efficacy of the generated images. RESULTS: A total of 959 CFP and 1009 UWF image pairs were included in model development. FID and KID indicated that images generated by C2ycleGAN presented significantly improved quality. Based on ophthalmologists' average ratings, the percentages of inadequate-quality images decreased from 32% to 18.8% for CFP, and from 18.7% to 14.7% for UWF. Only 24.8% and 13.8% of generated CFP and UWF images could be recognised as synthetic. The accuracy of retinopathy detection significantly increased from 78% to 91% for CFP and from 91% to 93% for UWF. For retinopathy subtype diagnosis, the accuracies also increased from 87%-94% to 91%-100% for CFP and from 87%-95% to 93%-97% for UWF. CONCLUSION: Digital ray could generate realistic postoperative CFP and UWF images with enhanced quality and accuracy for overall detection and subtype diagnosis of retinopathies, especially for CFP.\ TRIAL REGISTRATION NUMBER: This study was registered with ClinicalTrials.gov (NCT05491798).
RESUMO
AIM: To establish a classification for congenital cataracts that can facilitate individualized treatment and help identify individuals with a high likelihood of different visual outcomes. METHODS: Consecutive patients diagnosed with congenital cataracts and undergoing surgery between January 2005 and November 2021 were recruited. Data on visual outcomes and the phenotypic characteristics of ocular biometry and the anterior and posterior segments were extracted from the patients' medical records. A hierarchical cluster analysis was performed. The main outcome measure was the identification of distinct clusters of eyes with congenital cataracts. RESULTS: A total of 164 children (299 eyes) were divided into two clusters based on their ocular features. Cluster 1 (96 eyes) had a shorter axial length (mean±SD, 19.44±1.68 mm), a low prevalence of macular abnormalities (1.04%), and no retinal abnormalities or posterior cataracts. Cluster 2 (203 eyes) had a greater axial length (mean±SD, 20.42±2.10 mm) and a higher prevalence of macular abnormalities (8.37%), retinal abnormalities (98.52%), and posterior cataracts (4.93%). Compared with the eyes in Cluster 2 (57.14%), those in Cluster 1 (71.88%) had a 2.2 times higher chance of good best-corrected visual acuity [<0.7 logMAR; OR (95%CI), 2.20 (1.25-3.81); P=0.006]. CONCLUSION: This retrospective study categorizes congenital cataracts into two distinct clusters, each associated with a different likelihood of visual outcomes. This innovative classification may enable the personalization and prioritization of early interventions for patients who may gain the greatest benefit, thereby making strides toward precision medicine in the field of congenital cataracts.
RESUMO
Background: This study aimed to explore the postoperative myopic shift and its relationship to visual acuity rehabilitation in patients with bilateral congenital cataracts (CCs). Methods: Bilateral CC patients who underwent cataract extraction and primary intraocular lens implantations before 6 years old were included and divided into five groups according to surgical ages (<2, 2-3, 3-4, 4-5, and 5-6 years). The postoperative myopic shift rates, spherical equivalents (SEs), and the best corrected visual acuity (BCVA) were measured and analyzed. Results: A total of 1,137 refractive measurements from 234 patients were included, with a mean follow-up period of 34 months. The postoperative mean SEs at each follow-up in the five groups were linearly fitted with a mean R2 = 0.93 ± 0.03, which showed a downtrend of SE with age (linear regression). Among patients with a follow-up of 4 years, the mean postoperative myopic shift rate was 0.84, 0.81, 0.68, 0.24, and 0.28 diopters per year (D/y) in the five age groups (from young to old), respectively. The BCVA of those with a surgical age of <2 years at the 4-year visit was 0.26 (LogMAR), and the mean postoperative myopic shift rate was 0.84 D/y. For patients with a surgical age of 2-6 years, a poorer BCVA at the 4-year visit was found in those with higher postoperative myopic shift rates (r = 0.974, p = 0.026, Pearson's correlation test). Conclusion: Performing cataract surgery for patients before 2 years old and decreasing the postoperative myopic shift rates for those with a surgical age of 2-6 years may benefit visual acuity rehabilitation.
RESUMO
Utilization of digital technologies for cataract screening in primary care is a potential solution for addressing the dilemma between the growing aging population and unequally distributed resources. Here, we propose a digital technology-driven hierarchical screening (DH screening) pattern implemented in China to promote the equity and accessibility of healthcare. It consists of home-based mobile artificial intelligence (AI) screening, community-based AI diagnosis, and referral to hospitals. We utilize decision-analytic Markov models to evaluate the cost-effectiveness and cost-utility of different cataract screening strategies (no screening, telescreening, AI screening and DH screening). A simulated cohort of 100,000 individuals from age 50 is built through a total of 30 1-year Markov cycles. The primary outcomes are incremental cost-effectiveness ratio and incremental cost-utility ratio. The results show that DH screening dominates no screening, telescreening and AI screening in urban and rural China. Annual DH screening emerges as the most economically effective strategy with 341 (338 to 344) and 1326 (1312 to 1340) years of blindness avoided compared with telescreening, and 37 (35 to 39) and 140 (131 to 148) years compared with AI screening in urban and rural settings, respectively. The findings remain robust across all sensitivity analyses conducted. Here, we report that DH screening is cost-effective in urban and rural China, and the annual screening proves to be the most cost-effective option, providing an economic rationale for policymakers promoting public eye health in low- and middle-income countries.
Assuntos
Catarata , Análise Custo-Benefício , Programas de Rastreamento , Humanos , China/epidemiologia , Catarata/economia , Catarata/diagnóstico , Catarata/epidemiologia , Pessoa de Meia-Idade , Programas de Rastreamento/economia , Programas de Rastreamento/métodos , Masculino , Tecnologia Digital/economia , Feminino , Cadeias de Markov , Idoso , Inteligência Artificial , Telemedicina/economia , Telemedicina/métodosRESUMO
PURPOSE: To characterize the morphology of persistent pupillary membranes (PPMs) in pediatric patients and explore the corresponding surgical approaches. SETTING: Zhongshan Ophthalmic Center, Sun Yat-sen University, Guangzhou, China. DESIGN: Prospective observational study. METHODS: Consecutive pediatric patients with PPMs who underwent surgery from April 2020 to July 2022 were included. PPM morphology was assessed and categorized according to its anatomic relationship with crystalline lens and distribution of iris strands. The surgical approaches for different morphologies of PPMs were described in detail. The visual outcome and operation-related complications were recorded. RESULTS: 31 eyes from 19 patients were included with the mean age of 7.2 years. 3 morphological variants of PPMs were observed: type I (51.6%, 16/31), a spider-like appearance and no adhesion to the anterior lens capsule (ALC); type II (38.7%, 12/31), a loose central adherence to the ALC and partially thick iris strands attached to the iris collarette; type III (9.7%, 3/31), a tight central adherence to the ALC and only silk-like iris strands. Surgeries were performed with a natural pupil size in type I, while dilated pupil in the other types. The adhesions between PPM and the ALC were separated by viscoelastic injection in type II and by discission needles in type III. The corrected distance visual acuity was significantly improved from 0.34 ± 0.18 logMAR preoperatively to 0.17 ± 0.09 logMAR postoperatively ( P < .001). No operation-related complications were observed during 9.5-month follow-up. CONCLUSIONS: PPMs were categorized into 3 types according to their different morphologies, which helped to determine the best surgical strategy.
Assuntos
Acuidade Visual , Humanos , Estudos Prospectivos , Criança , Acuidade Visual/fisiologia , Feminino , Masculino , Pré-Escolar , Iris/cirurgia , Iris/anatomia & histologia , Anormalidades do Olho/cirurgia , Adolescente , Distúrbios Pupilares/cirurgia , Distúrbios Pupilares/fisiopatologia , Pupila/fisiologiaRESUMO
AIMS: To report the incidence and associated risk factors for developing suspected and definitive glaucoma after bilateral congenital cataract (CC) removal with a 5-year follow-up. METHODS: Secondary analysis of a prospective longitudinal cohort study. Bilateral CC patients who had undergone cataract surgery between January 2011 and December 2014 at Zhongshan Ophthalmic Centre were recruited. Suspected glaucoma was defined as persistent ocular hypertension requiring medical treatment. Definitive glaucoma was defined as accompanied by the progression of glaucomatous clinical features. According to postoperative lens status in 5 years follow-up: 130 eyes in the aphakia group; 219 in the primary intraocular lens (IOL) implantation group and 337 in the secondary IOL implantation group. The Kaplan-Meier survival and Cox regression analyses were used to explore the cumulative incidence and risk factors for suspected and definitive glaucoma. RESULTS: Three hundred fifty-one children (686 eyes) with bilateral CCs were enrolled in the study. The mean age at surgery was 1.82±2.08 years, and the mean follow-up duration was 6.26±0.97 years. Suspected and definitive glaucoma developed at a mean time of 2.84±1.75 years (range 0.02-7.33 years) postoperatively. The cumulative incidence of suspected and definitive glaucoma was 9.97% (35 of 351 patients), including 6.12% (42 eyes) for definitive glaucoma and 2.48% (17 eyes) for suspected glaucoma. Microcornea (HR 4.103, p<0.0001), CC family history (HR 3.285, p=0.001) and initial anterior vitrectomy (HR 2.365 p=0.036) were risk factors for suspected and definitive glaucoma. Gender, age at surgery, intraocular surgery frequency, length of follow-up and frequency of neodymium-doped yttrium aluminumaluminium garnet laser were non-statistically significant. Primary IOL implantation was a protective factor (HR 0.378, p=0.007). CONCLUSIONS: Identifying suspected and definitive glaucoma after bilateral CC surgery can lower the risk of secondary blindness in children. Patients with related risk factors need to pay more attention and thus reach early intervention and treatment during clinical practice. Primary IOL implantation may be a potential protective factor, need more clinical trials to be verified. TRIAL REGISTRATION NUMBER: NCT04342052.
Assuntos
Extração de Catarata , Catarata , Glaucoma , Hipertensão Ocular , Criança , Humanos , Lactente , Incidência , Seguimentos , Estudos Longitudinais , Estudos Prospectivos , Acuidade Visual , Complicações Pós-Operatórias , Catarata/complicações , Catarata/epidemiologia , Catarata/congênito , Extração de Catarata/efeitos adversos , Glaucoma/diagnóstico , Glaucoma/epidemiologia , Glaucoma/etiologia , Hipertensão Ocular/cirurgia , Fatores de RiscoRESUMO
BACKGROUND: Emerging three-dimensional digital visualization technology (DVT) provides more advantages than traditional microscopy in microsurgery; however, its impact on microsurgeons' visual and nervous systems and delicate microsurgery is still unclear, which hinders the wider implementation of DVT in digital visualization for microsurgery. METHODS AND MATERIAL: Forty-two microsurgeons from the Zhongshan Ophthalmic Center were enrolled in this prospective self-controlled study. Each microsurgeon consecutively performed 30 min conjunctival sutures using a three-dimensional digital display and a microscope, respectively. Visual function, autonomic nerve activity, and subjective symptoms were evaluated before and immediately after the operation. Visual functions, including accommodative lag, accommodative amplitude, near point of convergence and contrast sensitivity function (CSF), were measured by an expert optometrist. Heart rate variability was recorded by a wearable device for monitoring autonomic nervous activity. Subjective symptoms were evaluated by questionnaires. Microsurgical performance was assessed by the video-based Objective Structured Assessment of Technical Skill (OSATS) tool. RESULTS: Accommodative lag decreased from 0.63 (0.18) diopters (D) to 0.55 (0.16) D ( P =0.014), area under the log contrast sensitivity function increased from 1.49 (0.15) to 1.52 (0.14) ( P =0.037), and heart rate variability decreased from 36.00 (13.54) milliseconds (ms) to 32.26 (12.35) ms ( P =0.004) after using the DVT, but the changes showed no differences compared to traditional microscopy ( P >0.05). No statistical significance was observed for global OSATS scores between the two rounds of operations [mean difference, 0.05 (95% CI: -1.17 to 1.08) points; P =0.95]. Subjective symptoms were quite mild after using both techniques. CONCLUSIONS: The impact of DVT-based procedures on microsurgeons includes enhanced accommodation and sympathetic activity, but the changes and surgical performance are not significantly different from those of microscopy-based microsurgery. Our findings indicate that short-term use of DVT is reliable for microsurgery and the long-term effect of using DVT deserve more consideration.
Assuntos
Microscopia , Dispositivos Eletrônicos Vestíveis , Humanos , Microcirurgia/métodos , Estudos Prospectivos , TecnologiaRESUMO
Age is closely related to human health and disease risks. However, chronologically defined age often disagrees with biological age, primarily due to genetic and environmental variables. Identifying effective indicators for biological age in clinical practice and self-monitoring is important but currently lacking. The human lens accumulates age-related changes that are amenable to rapid and objective assessment. Here, using lens photographs from 20 to 96-year-olds, we develop LensAge to reflect lens aging via deep learning. LensAge is closely correlated with chronological age of relatively healthy individuals (R2 > 0.80, mean absolute errors of 4.25 to 4.82 years). Among the general population, we calculate the LensAge index by contrasting LensAge and chronological age to reflect the aging rate relative to peers. The LensAge index effectively reveals the risks of age-related eye and systemic disease occurrence, as well as all-cause mortality. It outperforms chronological age in reflecting age-related disease risks (p < 0.001). More importantly, our models can conveniently work based on smartphone photographs, suggesting suitability for routine self-examination of aging status. Overall, our study demonstrates that the LensAge index may serve as an ideal quantitative indicator for clinically assessing and self-monitoring biological age in humans.
Assuntos
Aprendizado Profundo , Cristalino , Humanos , Pré-Escolar , Envelhecimento/genéticaRESUMO
Importance: Retinal diseases are the leading cause of irreversible blindness worldwide, and timely detection contributes to prevention of permanent vision loss, especially for patients in rural areas with limited medical resources. Deep learning systems (DLSs) based on fundus images with a 45° field of view have been extensively applied in population screening, while the feasibility of using ultra-widefield (UWF) fundus image-based DLSs to detect retinal lesions in patients in rural areas warrants exploration. Objective: To explore the performance of a DLS for multiple retinal lesion screening using UWF fundus images from patients in rural areas. Design, Setting, and Participants: In this diagnostic study, a previously developed DLS based on UWF fundus images was used to screen for 5 retinal lesions (retinal exudates or drusen, glaucomatous optic neuropathy, retinal hemorrhage, lattice degeneration or retinal breaks, and retinal detachment) in 24 villages of Yangxi County, China, between November 17, 2020, and March 30, 2021. Interventions: The captured images were analyzed by the DLS and ophthalmologists. Main Outcomes and Measures: The performance of the DLS in rural screening was compared with that of the internal validation in the previous model development stage. The image quality, lesion proportion, and complexity of lesion composition were compared between the model development stage and the rural screening stage. Results: A total of 6222 eyes in 3149 participants (1685 women [53.5%]; mean [SD] age, 70.9 [9.1] years) were screened. The DLS achieved a mean (SD) area under the receiver operating characteristic curve (AUC) of 0.918 (0.021) (95% CI, 0.892-0.944) for detecting 5 retinal lesions in the entire data set when applied for patients in rural areas, which was lower than that reported at the model development stage (AUC, 0.998 [0.002] [95% CI, 0.995-1.000]; P < .001). Compared with the fundus images in the model development stage, the fundus images in this rural screening study had an increased frequency of poor quality (13.8% [860 of 6222] vs 0%), increased variation in lesion proportions (0.1% [6 of 6222]-36.5% [2271 of 6222] vs 14.0% [2793 of 19â¯891]-21.3% [3433 of 16â¯138]), and an increased complexity of lesion composition. Conclusions and Relevance: This diagnostic study suggests that the DLS exhibited excellent performance using UWF fundus images as a screening tool for 5 retinal lesions in patients in a rural setting. However, poor image quality, diverse lesion proportions, and a complex set of lesions may have reduced the performance of the DLS; these factors in targeted screening scenarios should be taken into consideration in the model development stage to ensure good performance.
Assuntos
Aprendizado Profundo , Doenças Retinianas , Humanos , Feminino , Idoso , Sensibilidade e Especificidade , Fundo de Olho , Retina/diagnóstico por imagem , Retina/patologia , Doenças Retinianas/diagnóstico por imagem , Doenças Retinianas/patologiaRESUMO
Image quality variation is a prominent cause of performance degradation for intelligent disease diagnostic models in clinical applications. Image quality issues are particularly prominent in infantile fundus photography due to poor patient cooperation, which poses a high risk of misdiagnosis. Here, we developed a deep learning-based image quality assessment and enhancement system (DeepQuality) for infantile fundus images to improve infant retinopathy screening. DeepQuality can accurately detect various quality defects concerning integrity, illumination, and clarity with area under the curve (AUC) values ranging from 0.933 to 0.995. It can also comprehensively score the overall quality of each fundus photograph. By analyzing 2,015,758 infantile fundus photographs from real-world settings using DeepQuality, we found that 58.3% of them had varying degrees of quality defects, and large variations were observed among different regions and categories of hospitals. Additionally, DeepQuality provides quality enhancement based on the results of quality assessment. After quality enhancement, the performance of retinopathy of prematurity (ROP) diagnosis of clinicians was significantly improved. Moreover, the integration of DeepQuality and AI diagnostic models can effectively improve the model performance for detecting ROP. This study may be an important reference for the future development of other image-based intelligent disease screening systems.
RESUMO
Ischemic retinal diseases (IRDs) are a series of common blinding diseases that depend on accurate fundus fluorescein angiography (FFA) image interpretation for diagnosis and treatment. An artificial intelligence system (Ai-Doctor) was developed to interpret FFA images. Ai-Doctor performed well in image phase identification (area under the curve [AUC], 0.991-0.999, range), diabetic retinopathy (DR) and branch retinal vein occlusion (BRVO) diagnosis (AUC, 0.979-0.992), and non-perfusion area segmentation (Dice similarity coefficient [DSC], 89.7%-90.1%) and quantification. The segmentation model was expanded to unencountered IRDs (central RVO and retinal vasculitis), with DSCs of 89.2% and 83.6%, respectively. A clinically applicable ischemia index (CAII) was proposed to evaluate ischemic degree; patients with CAII values exceeding 0.17 in BRVO and 0.08 in DR may be associated with increased possibility for laser therapy. Ai-Doctor is expected to achieve accurate FFA image interpretation for IRDs, potentially reducing the reliance on retinal specialists.
Assuntos
Retinopatia Diabética , Oclusão da Veia Retiniana , Humanos , Inteligência Artificial , Angiofluoresceinografia/métodos , Oclusão da Veia Retiniana/diagnóstico , Oclusão da Veia Retiniana/terapia , Retinopatia Diabética/diagnóstico por imagem , Retinopatia Diabética/terapia , Isquemia/diagnóstico , Isquemia/terapiaRESUMO
Data quality issues have been acknowledged as one of the greatest obstacles in medical artificial intelligence research. Here, we present DeepFundus, which employs deep learning techniques to perform multidimensional classification of fundus image quality and provide real-time guidance for on-site image acquisition. We describe steps for data preparation, model training, model inference, model evaluation, and the visualization of results using heatmaps. This protocol can be implemented in Python using either the suggested dataset or a customized dataset. For complete details on the use and execution of this protocol, please refer to Liu et al.1.
Assuntos
Pesquisa Biomédica , Aprendizado Profundo , Inteligência ArtificialRESUMO
Deep learning (DL) is the new high-profile technology in medical artificial intelligence (AI) for building screening and diagnosing algorithms for various diseases. The eye provides a window for observing neurovascular pathophysiological changes. Previous studies have proposed that ocular manifestations indicate systemic conditions, revealing a new route in disease screening and management. There have been multiple DL models developed for identifying systemic diseases based on ocular data. However, the methods and results varied immensely across studies. This systematic review aims to summarize the existing studies and provide an overview of the present and future aspects of DL-based algorithms for screening systemic diseases based on ophthalmic examinations. We performed a thorough search in PubMed®, Embase, and Web of Science for English-language articles published until August 2022. Among the 2873 articles collected, 62 were included for analysis and quality assessment. The selected studies mainly utilized eye appearance, retinal data, and eye movements as model input and covered a wide range of systemic diseases such as cardiovascular diseases, neurodegenerative diseases, and systemic health features. Despite the decent performance reported, most models lack disease specificity and public generalizability for real-world application. This review concludes the pros and cons and discusses the prospect of implementing AI based on ocular data in real-world clinical scenarios.
RESUMO
Medical artificial intelligence (AI) has been moving from the research phase to clinical implementation. However, most AI-based models are mainly built using high-quality images preprocessed in the laboratory, which is not representative of real-world settings. This dataset bias proves a major driver of AI system dysfunction. Inspired by the design of flow cytometry, DeepFundus, a deep-learning-based fundus image classifier, is developed to provide automated and multidimensional image sorting to address this data quality gap. DeepFundus achieves areas under the receiver operating characteristic curves (AUCs) over 0.9 in image classification concerning overall quality, clinical quality factors, and structural quality analysis on both the internal test and national validation datasets. Additionally, DeepFundus can be integrated into both model development and clinical application of AI diagnostics to significantly enhance model performance for detecting multiple retinopathies. DeepFundus can be used to construct a data-driven paradigm for improving the entire life cycle of medical AI practice.
Assuntos
Inteligência Artificial , Citometria de Fluxo , Curva ROC , Área Sob a CurvaRESUMO
Early detection of visual impairment is crucial but is frequently missed in young children, who are capable of only limited cooperation with standard vision tests. Although certain features of visually impaired children, such as facial appearance and ocular movements, can assist ophthalmic practice, applying these features to real-world screening remains challenging. Here, we present a mobile health (mHealth) system, the smartphone-based Apollo Infant Sight (AIS), which identifies visually impaired children with any of 16 ophthalmic disorders by recording and analyzing their gazing behaviors and facial features under visual stimuli. Videos from 3,652 children (≤48 months in age; 54.5% boys) were prospectively collected to develop and validate this system. For detecting visual impairment, AIS achieved an area under the receiver operating curve (AUC) of 0.940 in an internal validation set and an AUC of 0.843 in an external validation set collected in multiple ophthalmology clinics across China. In a further test of AIS for at-home implementation by untrained parents or caregivers using their smartphones, the system was able to adapt to different testing conditions and achieved an AUC of 0.859. This mHealth system has the potential to be used by healthcare professionals, parents and caregivers for identifying young children with visual impairment across a wide range of ophthalmic disorders.
Assuntos
Aprendizado Profundo , Smartphone , Masculino , Lactente , Humanos , Criança , Pré-Escolar , Feminino , Olho , Pessoal de Saúde , Transtornos da Visão/diagnósticoRESUMO
Background: We aimed to evaluate the morphological characteristics of the subfoveal choroid and explore the possible association of these characteristics with best-corrected visual acuity (BCVA) in postoperative patients with unilateral congenital cataracts (CCs). Methods: This was a cross-sectional study. Subfoveal choroidal structures were measured by spectral-domain optical coherence tomography with enhanced depth imaging (EDI-OCT). Several choroidal parameters, including subfoveal choroidal thickness (SFCT), total choroidal area (TCA), luminal area (LA), stromal area (SA) and choroidal vascularity index (CVI), were compared between pseudophakic and contralateral healthy eyes. Then, the choroidal parameters were compared between pseudophakic eyes with a poor BCVA (>0.3 logMAR) and those with a good BCVA (≤0.3 logMAR). The performance of the choroidal parameters in detecting a poor BCVA in pseudophakic eyes was evaluated by using the area under the receiver operating characteristic curve (AUC). A logistic regression model was used to assess the association between choroidal parameters and BCVA in postoperative patients with unilateral CCs. Results: A total of 55 postoperative patients with unilateral CCs were included. The age was 6.67±2.64 years. Thinner SFCT and smaller TCA, LA, SA and CVI were observed in pseudophakic eyes than in contralateral healthy eyes. In addition, in pseudophakic eyes, those with a poor BCVA had a thinner SFCT and a smaller TCA, LA and SA than those with a good BCVA. TCA [AUC, 0.75; 95% confidence interval (CI), 0.62, 0.88], LA (AUC, 0.74; 0.61, 0.87) and SA (AUC, 0.74; 0.60, 0.87) showed acceptable discriminatory abilities on BCVA. Pseudophakic eyes with TCA ≤0.594 mm² [odds ratio (OR), 8.90; 95% CI: 1.99, 39.94; P=0.004], LA ≤0.402 mm² (OR 8.90; 95% CI: 1.99, 39.94; P=0.004) or SA ≤0.218 mm (OR, 6.53; 95% CI: 1.69, 25.27; P=0.007) were more likely to have a poor visual acuity. Conclusions: The pseudophakic eyes in patients with unilateral CCs had thinner SFCT and smaller TCA, LA, SA and CVI than the contralateral healthy eyes. In pseudophakic eyes, smaller TCA, LA and SA values were associated with a poor visual acuity.