With the growth of digital media and social networks, sharing visual content has become common in people’s daily lives. In the food industry, visually appealing food images can attract attention, drive engagement, and influence consumer behavior. Therefore, it is crucial for businesses to understand what constitutes attractive food images. Assessing the attractiveness of food images poses significant challenges due to the lack of large labeled datasets that align with diverse public preferences. Additionally, it is challenging for computer assessments to approach human judgment in evaluating aesthetic quality. This paper presents a novel framework that circumvents the need for explicit human annotation by leveraging user engagement data that are readily available on social media platforms. We propose procedures to collect, filter, and automatically label the attractiveness classes of food images based on their user engagement levels. The data gathered from social media are used to create predictive models for category-specific attractiveness assessments. Our experiments across five food categories demonstrate the efficiency of our approach. The experimental results show that our proposed user-engagement-based attractiveness class labeling achieves a high consistency of 97.2% compared to human judgments obtained through A/B testing. Separate attractiveness assessment models were created for each food category using convolutional neural networks (CNNs). When analyzing unseen food images, our models achieve a consistency of 76.0% compared to human judgments. The experimental results suggest that the food image dataset collected from social networks, using the proposed framework, can be successfully utilized for learning food attractiveness assessment models.

Corneal ulcer is a prevalent ocular condition that requires ophthalmologists to diagnose, assess, and monitor symptoms. During examination, ophthalmologists must identify the corneal ulcer area and evaluate its severity by manually comparing ocular staining images with severity indices. However, manual assessment is time-consuming and may provide inconsistent results. Variations can occur with repeated evaluations of the same images or with grading among different evaluators. To address this problem, we propose an automated corneal ulcer grading system for ocular staining images based on deep learning techniques and the Hough Circle Transform. The algorithm is structured into two components for cornea segmentation and corneal ulcer segmentation. Initially, we apply a deep learning method combined with the Hough Circle Transform to segment cornea areas. Subsequently, we develop the corneal ulcer segmentation model using deep learning methods. In this phase, the predicted cornea areas are utilized as masks for training the corneal ulcer segmentation models during the learning phase. Finally, this algorithm uses the results from these two components to determine two outputs: (1) the percentage of the ulcerated area on the cornea, and (2) the severity degree of the corneal ulcer based on the Type–Grade (TG) grading standard. These methodologies aim to enhance diagnostic efficiency across two key aspects: (1) ensuring consistency by delivering uniform and dependable results, and (2) enhancing robustness by effectively handling variations in eye size. In this research, our proposed method is evaluated using the SUSTech-SYSU public dataset, achieving an Intersection over Union of 89.23% for cornea segmentation and 82.94% for corneal ulcer segmentation, along with a Mean Absolute Error of 2.51% for determining the percentage of the ulcerated area on the cornea and an Accuracy of 86.15% for severity grading.

The rapid growth of dockless electric scooter (e-scooter) sharing services has transformed short-distance urban mobility, offering convenience and sustainability benefits while amplifying challenges related to demand imbalance, fleet rebalancing, and spatial inequity. Accurate spatiotemporal demand prediction is therefore essential for optimizing resource allocation and supporting data-driven policy interventions. This study proposes a hybrid deep learning framework that integrates a Graph Convolutional Network (GCN) with a Gated Recurrent Unit (GRU) and community detection to enhance short-term prediction of e-scooter pick-up and drop-off demands. The Louvain algorithm is employed to partition urban areas into mobility-based communities, enabling the model to capture functional connectivity rather than relying solely on geographic proximity. Using real-world e-scooter trip data from Calgary, Canada, the model’s performance is evaluated against established baselines, including a Masked Fully Convolutional Network (MFCN) and conventional GRU architectures. Results show that the proposed approach achieves up to 11.8% improvement in mean absolute error (MAE) compared with the MFCN baseline and more robust generalization across temporal horizons. The findings demonstrate that integrating community structures into graph-based learning effectively captures complex urban dynamics, providing practical insights for sustainable micromobility operation and service deployment.

Topical glaucoma therapy can affect the ocular surface, but objective longitudinal data on conjunctival hyperemia are limited. We prospectively followed 31 patients naive to topical antiglaucoma therapy for 24 weeks after treatment initiation, analyzing one baseline severity-selected eye per participant. All 31 participants contributed complete analyzable conjunctival image series at baseline and 6, 12, and 24 weeks. Automated image processing quantified strong red area as the primary hyperemia metric and blood vessel area as a secondary metric. Tear breakup time (TBUT), Schirmer I testing, National Eye Institute (NEI) corneal staining, Ocular Surface Disease Index (OSDI), and NEI Visual Function Questionnaire-25 were also assessed. Strong red area increased numerically by 2.56 percentage points at 24 weeks (95% CI − 1.06 to 6.18; P = 0.165), while blood vessel area remained stable. NEI staining decreased, TBUT and Schirmer I showed no significant overall change, OSDI scores decreased (overall P = 0.002), and NEI-VFQ-25 scores improved. Over 24 weeks, image-derived conjunctival hyperemia showed no statistically significant persistent increase, whereas clinical ocular-surface signs and patient-reported outcomes followed differing trajectories. These preliminary findings characterize ocular-surface changes after initiation of topical glaucoma therapy, although the small uncontrolled design precludes causal or equivalence conclusions.

Background: Patient-reported quality of life during routine use of mobile medication-support applications in glaucoma is poorly characterized. Objective: To characterize longitudinal health-related quality of life among self-selected OcuDrop users and explore associated factors. Methods: This single-center prospective observational cohort included adults with glaucoma or ocular hypertension recruited through a clinic announcement that specified Android-only availability. Participants voluntarily used the medication-reminder application during routine care. EQ-5D-5L was assessed at baseline and months 1, 2, and 3; NEI-VFQ-25 was exploratory. Random-intercept mixed models evaluated change over time. App events were linked using pseudonymous identifiers. Results: Of 81 voluntary app users, 64 enrolled and 54 had EQ-5D-5L data at all visits. Mean utility was 0.681 at baseline, 0.730 at month 1, 0.777 at month 2, and 0.775 at month 3 (overall P = 0.001). The month-3 within-person change was 0.094 (95% CI: 0.030 to 0.159; Cohen dz = 0.40); clinical importance could not be determined because no applicable glaucoma-specific Thai minimal clinically important difference was identified. Older age was associated with lower baseline utility, and baseline utility with month-3 utility, in exploratory models. Among 54 participants, 50 generated at least one app event and 39 generated an event in month 3. Ten enrolled participants had no available EQ-5D-5L records data and used more topical medications than included participants. Conclusion: EQ-5D-5L utility increased over three months among self-selected users, but the uncontrolled design, Android-only eligibility, missing outcomes, and variable engagement preclude attribution to OcuDrop. Findings are hypothesis-generating.