SoC Research Publications
Permanent URI for this collection
Browse
Recent Submissions
Item Incidence and risk factors of fellow-eyes wet conversion in unilateral neovascular age-related macular degeneration over 15-year follow-up(Springer Science and Business Media Deutschland GmbH, 2024-08-23) Sadeghi, Elham; Vupparaboina, Sharat Chandra; Bollepalli, Sandeep Chandra; Vupparaboina, Kiran; Agarwal, Komal; Sahel, Jose-Alain; Eller, Andrew W.; Chhablani, JayPurpose: Incidence and risk factors of fellow eye wet conversion in unilateral neovascular age-related macular degeneration (nAMD) over 15-years follow-up. Methods: This retrospective study reviewed 593 unilateral nAMD patients with a minimum of five years up to 15-years of follow-up. The demographic data, visual acuity, fellow eye nAMD conversion rate, and the number of anti-vascular endothelial growth factor (anti-VEGF) injections in the primary eye were evaluated. Also, the nAMD-converted fellow eyes were divided into two groups based on the time of conversion (less and more than two years from the first injection in the primary eye). Based on the data types, the T-test, Chi-square, and Mann-Whitney U test were used to analyze. Results: The total cases were 593 patients, and 248 eyes (41.82%) converted to nAMD in the mean interval of 34.92 ± 30.62months. The males exhibited a predisposition to wet conversion at 2.54years earlier than their female counterparts (P = 0.025). In all the converted fellow eyes, the mean age was 2.3years higher at presentation in the group who converted within two years of follow-up in compared to eyes that converted after two years (79.82 ± 8.64 vs 77.51 ± 8.5years, P = 0.035). Additionally, eyes converting within two years had a mean baseline LogMAR visual acuity of 0.44 ± 0.47, compared to 0.32 ± 0.41 for conversions after two years (P = 0.014). Conclusion: This study reported that males showed a predisposition to fellow eye nAMD conversion at an earlier age. Additionally, there was a trend of faster fellow eye nAMD conversion in individuals with higher age and lower baseline visual acuity. Key messages: What is known • Certain risk factors may make the fellow eye of neovascular age-related macular degeneration (nAMD) more likely to progress to wet conversion. • Identifying these risk factors for fellow eye wet conversion can help prevent it, potentially preserving the patient's vision quality for a longer duration. • The studies on the incidence of wet conversion in the fellow eye have yielded controversial results. What is new • During the 15-year follow-up period, nearly half (47.58%) of the fellow eyes that underwent wet conversion did so within the initial two years following the wet conversion of the first eye. • Males showed a predisposition to fellow eye nAMD conversion at an earlier age. • There was a trend of faster fellow eye nAMD conversion in individuals with higher age and lower baseline visual acuity. © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2024.Item On the use of a simulation framework for studying accessibility challenges faced by people with disabilities in indoor environments(Springer Science and Business Media Deutschland GmbH, 2024) Garfias, Francisco R.; Namboodiri, VinodNavigating indoor spaces is known to be significantly challenges for individuals with mobility and sensory impairments due to the presence of physical barriers and inadequate accessible signage. Current laws and efforts have not led to meeting diverse needs of these populations. In this work we provide a brief introduction to MABLESim (Mapping for Accessible BuiLt Environments Simulator), a simulation framework for studying indoor space accessibility. MABLESim recreates digital models of indoor environments, allowing for the simulation of diverse mobility scenarios for individuals with varying abilities. MABLESim enables the analysis of critical factors important for efficient mobility in indoor spaces such as route complexity and disability characteristics. Through careful configuration of simulation parameters, MABLESim facilitates the assessment of accessibility challenges in both simple and complex indoor spaces. This framework offers a tool for designers and planners to visualize and address accessibility barriers in built environments. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.Item Firefly algorithm-based LSTM model for Guzheng tunes switching with big data analysis(Elsevier Ltd, 2024) Han, Mingjin; Soradi-Zeid, Samaneh; Anwlnkom, Tomley; Yang, YuanyuanGuzheng tune progression involves intricately harmonizing melodic motif transitions. Effectively navigating this vast creative possibility space to expose musically consistent elaborations presents challenges. We develop a specialized large long short-term memory (LSTM) model for generating musically consistent Guzheng tune transitions. First, we propose novel firefly algorithm (FA) enhancements, e.g., adaptive diversity preservation and adaptive swim parameters, to boost exploration effectiveness for navigating the vast creative combinatorics when generating Guzheng tune transitions. Then, we develop a specialized stacked LSTM architecture incorporating residual connections and conditioned embedding vectors that can leverage long-range temporal dependencies in Guzheng music patterns, including unsupervised learning of concise Guzheng-specific melody embedding vectors via a variational autoencoder, encapsulating unique harmonic signatures from performance descriptors to provide style guidance. Finally, we use LSTM networks to develop adversarial generative large models that enable realistic synthesis and evaluation of Guzheng tunes switching. We gather an extensive 10+ hour corpus of solo Guzheng recordings spanning 230 musical pieces, 130 distinguished performing artists, and 600+ audio tracks. Simultaneously, we conduct thorough Guzheng data analysis. Comparative assessments against strong baselines over systematic musical metrics and professional listeners validate significant generation fidelity improvements. Our model achieves a 63 % reduction in reconstruction error compared to the standard FA optimization after 1000 iterations. It also outperforms baselines in capturing characteristic motifs, maintaining modality coherence with under 2 % dissonant pitch errors, and retaining desired rhythmic cadences. User studies further confirm the superior naturalness, novelty, and stylistic faithfulness of the generated tune transitions, with ratings close to real data. © 2024Item Learning motion primitives for the quantification and diagnosis of mobility deficits(IEEE Computer Society, 2024) Yan, Fujian; Gong, Jiaqi; Zhang, Qiyang; He, HongshengThe severity of mobility deficits is one of the most critical parameters in the diagnosis and rehabilitation of Parkinson's disease (PD). The current approach for severity evaluation is clinical scaling that relies on a clinician's subjective observations and experience, and the observation in laboratories or clinics may not suffice to reflect the severity of motion deficits as compared to daily living activities. The paper presents an approach to modeling and quantifying the severity of mobility deficits from motion data by using nonintrusive wearable physio-biological sensors. The approach provides a user-specific metric that measures mobility deficits in terms of the quantities of motion primitives that are learned from motion tracking data. The proposed method achieved 99.84% prediction accuracy on laboratory data and 93.95% prediction accuracy on clinical data. This approach presents the potential to supplant traditional observation-based clinical scaling, providing an avenue for real-time feedback to fortify positive progression throughout the course of rehabilitation. IEEEItem A Novel Salary Prediction System Using Machine Learning Techniques(Institute of Electrical and Electronics Engineers Inc., 2024) Asaduzzaman, Abu; Uddin, Md Raihan; Woldeyes, Yoel; Sibai, Fadi N.The purpose of this work is to build a salary prediction system using machine learning techniques. The experiments are done using the data from 1994 census database which has 32,561 records of employee data. The techniques used in determining whether an employee salary is less than or greater than $50,000 are: logistic regression, decision tree, Naive Bayes classifier, K-nearest neighbor, and support vector machine. We implement these algorithms using original train data and oversampled train data. The results of these models are analyzed and compared with respect to accuracy. According to the experimental results, decision tree model outperforms the other models with original train data. © 2024 IEEE.