Digital Commons @ Texas A&M University-San Antonio
Not a member yet
3365 research outputs found
Sort by
Efficient Training on Alzheimer’s Disease Diagnosis with Learnable Weighted Pooling for 3D PET Brain Image Classification
Three-dimensional convolutional neural networks (3D CNNs) have been widely applied to analyze Alzheimer’s disease (AD) brain images for a better understanding of the disease progress or predicting the conversion from cognitively impaired (CU) or mild cognitive impairment status. It is well-known that training 3D-CNN is computationally expensive and with the potential of overfitting due to the small sample size available in the medical imaging field. Here we proposed a novel 3D-2D approach by converting a 3D brain image to a 2D fused image using a Learnable Weighted Pooling (LWP) method to improve efficient training and maintain comparable model performance. By the 3D-to-2D conversion, the proposed model can easily forward the fused 2D image through a pre-trained 2D model while achieving better performance over different 3D and 2D baselines. In the implementation, we chose to use ResNet34 for feature extraction as it outperformed other 2D CNN backbones. We further showed that the weights of the slices are location-dependent and the model performance relies on the 3D-to-2D fusion view, with the best outcomes from the coronal view. With the new approach, we were able to reduce 75% of the training time and increase the accuracy to 0.88, compared with conventional 3D CNNs, for classifying amyloid-beta PET imaging from the AD patients from the CU participants using the publicly available Alzheimer’s Disease Neuroimaging Initiative dataset. The novel 3D-2D model may have profound implications for timely AD diagnosis in clinical settings in the future
The Change was as Big as Night and Day : Experiences of Professors Teaching Students with Intellectual Disabilities
Since the inception of the Higher Education Opportunity Act in 2008, there has been an increase in the number of post-secondary education institutions in the United States that have established inclusive postsecondary programs for individuals with intellectual disabilities to attend college and achieve higher levels of employment. Previous studies have investigated the development and outcomes of these programs, however, less has been explored related to professors\u27 experiences and perceptions regarding this unique student population, particularly within Hispanic Serving Institutions (HSI). The current study focused on professors teaching inclusive courses within a new Comprehensive Transition and Postsecondary Program at a HSI and aimed to identify their perceptions and experiences related to instructing students with intellectual disabilities. Six professors participated in pre- and post-semester in-depth interviews. Findings from applied thematic analysis included: (a) barriers to success; (b) academic supports and strategies; (c) successful outcomes and (d) considerations for future, related programming
A Case for Increased Rigor in AAC Research: A Methodological Quality Review
This comprehensive review reports on methodological quality of 162 single-case studies on augmentative and alternative communication interventions for communication and challenging behavior in individuals diagnosed with autism or intellectual disabilities and with complex communication needs. Following review for inclusion criteria, documents were excluded if they failed to meet basic methodological standards. Each remaining study was evaluated for 10 detailed quality criteria. No studies met all standards without reservations. Only three of the included studies met all of the standards with reservations and the remainder met some but not all standards, with or without reservations. The included studies reported adequate detail for half of the quality indicators, but insuffi- cient details for participant, setting, maintenance, generalization, and social validity descriptions. An increased quantity and quality of research were found in over four decades. More recent studies have adequately reported half of the criteria investigated, including describing the materials, defining the outcome variables, describing baseline and intervention procedures, and evaluating procedural integrity. After identifying quality features, the authors report in more detail on low-rated quality indicators particularly relevant to studies addressing social-communication interventions. The literature infrequently reported race, ethnicity, or home language. Future research should report characteristics of participants to ensure that research becomes representative of the population
Mathematical Modeling of the Marital Interaction Dynamics
We may never think the marriage interaction can be mathematically modeled. It is indeed possible applying only fundamental college math. By the Rapid Couples Interaction Scoring System, a videotaped interactive discussion between a couple and detailed aspects of their emotions are coded to give scores for each one of the couple. Humor and smiling are important for good relations generating positive scores, while anger or criticism are bad for relations generating negative scores. These scores are employed to construct score curves for both the husband and wife reacting to each other. We establish discrete mathematical models for these score curves and use dynamics to study under what conditions the couple has a stable or unstable marriage. Then our model is used to analyze how several personalities (including volatile, validating, and conflict-avoiding) of each one of the couple impact the stability of marriage
Examining the Association between Awareness and Acceptance of Impermanence and Humility
People who appreciate their smallness compared to the vastness of the world can detach themselves from their egos and become humble. We argue that humility is not only about the relative size of the ego, but also about the relativetime in which the ego exists. The current study examined the relationship between awareness and acceptance of the fleeting nature of time (impermanence) and humility. We collected data from 257 adults residing in the United Statesthrough an online study. Participants completed a measure of impermanence and measures of humility. Bivariate correlations suggested that there were significant positive correlations between acceptance and awareness of impermanence and humility. The results suggested people who were more aware and accepting of impermanence had higher reports of being humble, suggesting another strategy to increase humility
Water Resources Science and Technology Fall Seminar Series: Christopher Vaughn (San Antonio River Authority)
Chris Vaughn is the San Antonio River Authority\u27s Watershed Monitoring Supervisor. He introduced the concept and the protocol of biological index to monitor the health of the San Antonio River watershed
Playful Mouth-to-Mouth Interactions of Belugas (Delphinapterus leucas) in Managed Care
Belugas (Delphinapterus leucas) engage in many forms of play (e.g., object, water, locomotor), but no play is quite as curious as the unusual form of cooperative social play involving mouth-to-mouth interactions. These playful interactions are characterized by two belugas approaching each other head-to-head and interlocking their jaws, clasping one another, as if they were shaking hands. Observed in belugas both in the wild and in managed care, it is seemingly an important type of social play that offers a unique way of socializing with conspecifics. To describe this unusual behavior, a group of belugas in managed care was observed from 2007 to 2019. Although adults participated in mouth-to-mouth interactions, most were initiated and received by young belugas. Both males and females engaged in mouth-to-mouth interactions and did so at similar frequencies. Individual differences in how many mouth-to-mouth interactions were initiated among calves were also observed. Due to the unique, cooperative nature of mouth-to-mouth interactions, which require both social and motor skills, it is hypothesized that these interactions may be used to test social and motor competency
The Addition of Sprint Interval Training to Field Lacrosse Training Increases Rate of Torque Development and Contractile Impulse in Female High School Field Lacrosse Players
Field lacrosse requires sudden directional changes and rapid acceleration/deceleration. The capacity to perform these skills is dependent on explosive muscle force production. Limited research exists on the potential of sprint interval training (SIT) to impact explosive muscle force production in field lacrosse players. The purpose of this study is to examine SIT, concurrent to field-lacrosse-specific training, on the rate of torque development (RTD), contractile impulse, and muscle function in female high school field lacrosse players (n = 12; 16 ± 1 yrs.). SIT was performed three times per week, concurrent to field-lacrosse-specific training, for 12 weeks. Right lower-limb muscle performance was assessed pre-, mid-, and post-SIT training via isometric and isokinetic concentric knee extensor contractions. Outcomes included RTD (Nm·s−1), contractile impulse (Nm·s), and peak torque (Nm). RTD for the first 50 ms of contraction improved by 42% by midseason and remained elevated at postseason (p = 0.004, effect size (ES) = −577.3 to 66.5). Contractile impulse demonstrated a training effect across 0–50 ms (42%, p = 0.004, ES = −1.4 to 0.4), 0–100 ms (33%, p = 0.018, ES = 3.1 to 0.9), and 0–200 ms (22%, p = 0.031, ES = −7.8 to 1.6). Isometric (0 rad·s−1) and concentric (3.1 rad·s−1) strength increased by 20% (p = 0.002, ES = −60.8 to −20.8) and 9% (p = 0.038, ES = −18.2 to 0.0) from SIT and field-lacrosse-specific training, respectively (p \u3c 0.05). SIT, concurrent to field-lacrosse- specific training, enhanced lower-limb skeletal muscle performance, which may enable greater sport-specific gains
Self-Supervised Learning Application on COVID-19 Chest X- ray Image Classification Using Masked AutoEncoder
The COVID-19 pandemic has underscored the urgent need for rapid and accurate diagnosis facilitated by artificial intelligence (AI), particularly in computer-aided diagnosis using medical imag- ing. However, this context presents two notable challenges: high diagnostic accuracy demand and limited availability of medical data for training AI models. To address these issues, we proposed the implementation of a Masked AutoEncoder (MAE), an innovative self-supervised learning approach, for classifying 2D Chest X-ray images. Our approach involved performing imaging reconstruction using a Vision Transformer (ViT) model as the feature encoder, paired with a custom-defined decoder. Additionally, we fine-tuned the pretrained ViT encoder using a labeled medical dataset, serving as the backbone. To evaluate our approach, we conducted a comparative analysis of three distinct training methods: training from scratch, transfer learning, and MAE-based training, all employing COVID-19 chest X-ray images. The results demonstrate that MAE-based training produces superior performance, achieving an accuracy of 0.985 and an AUC of 0.9957. We explored the mask ratio influ- ence on MAE and found ratio = 0.4 shows the best performance. Furthermore, we illustrate that MAE exhibits remarkable efficiency when applied to labeled data, delivering comparable performance to utilizing only 30% of the original training dataset. Overall, our findings highlight the significant performance enhancement achieved by using MAE, particularly when working with limited datasets. This approach holds profound implications for future disease diagnosis, especially in scenarios where imaging information is scarce