Providing quality computer science education (CSEd) in rural areas is essential for reducing computing disparities. Compared to teachers from non-rural areas, teachers in rural school face some unique challenges in teaching CS, such as limited access to professional learning resources. However, with close access to agriculture and farming land, rural teachers also have unique opportunities to integrate CS instruction into science or other subject areas in a way that is relevant to students' daily life. This study will share lessons learned from the implementation of an initiative to prepare rural teachers to integrate CS into their lessons through the context of agriculture. Teachers who participated in the initiative increased their preparedness to integrate CS into their classrooms and indicated that it would benefit their students' CS and agricultural learning. Some teachers noted that the initiative would ultimately contribute to the local community by increasing student awareness of CS-related career pathways and by expanding technology use in farming. Findings from the project's initial implementation highlight the value of a CS curriculum that connects to local agricultural industries and the importance of providing teachers with the materials and resources needed to implement it within their context. Initial findings also offered critical insights to inform expansion of the project and to strengthen its support of rural teachers and schools.

In recent years, there has been an increased recognition of the importance of professional development (PD) programs that prepare grade K-5 teachers to teach computer science (CS) and computational thinking (CT), primarily through integrating these topics into core curriculum. Prior studies using quantitative and qualitative methods have examined outcomes of such PD programs, revealing important aspects to measure when evaluating teacher readiness and concerns about integrating CS and CT into their instruction, including individual capacity, network and resources, and barrier and concerns. The current study aims to fill in the gap of validated instruments to measure teacher readiness and concerns to teach or integrate CS and CT in the K-5 educational context. An instrument was developed based on existing validated measures and adapted for the K-5 teacher population. Pre-survey data from 641 K-5 teachers trained to integrate CS and CT into their instruction was used to validate the instrument. Internal reliability and exploratory factor analysis (EFA) were conducted on two subscales: readiness and concerns to integrate CS and CT. Results suggested satisfactory psychometric properties for both subscales overall, but suggested the removal of one item on each subscale to further improve internal reliability and factor structure. Pearson correlation results suggested the readiness and concerns subscales were moderately correlated (r=-0.38), indicating readiness and concerns to be two correlated but separate aspects in teachers' preparedness to integrate CS and CT in K-5 classrooms. The development and validation of this instrument provide a reliable and valuable tool to assess future PD program outcomes in K-5 CS education.

Providing high-quality computing education to all students in primary school is essential for reducing computing disparities, by preparing students for success in secondary education. To impact the quality of computing education experiences for large numbers of diverse primary schools and students, a teacher professional development model must be designed for scalability from the outset. As many countries work to expand national access to computing education at the primary level and reach hundreds or thousands of teachers in the process, there is a need to understand effective systems-based strategies for scaling teacher professional learning across diverse schools and communities. This experience report will share lessons learned from the implementation of an initiative to prepare thousands of primary school (K-5) teachers to integrate computational thinking into their instruction.

In recent years, grade K-5 teachers have increasingly integrated computer science instruction into different subject areas. To identify strategies to support K-5 teachers in providing computer science instruction, it is crucial to understand the primary challenges and facilitators that they experience. This study surveyed 1,054 teachers from diverse Texas schools and identified key supports and barriers to integrating new state technology education standards in K-5 classrooms. Teachers completed a survey before participating in a professional development program. The teachers identified facilitators to integrating the standards into their instruction, such as access to technology, professional development, and administrator support. They also shared barriers such as time constraints, lack of curriculum and training, and insufficient technology infrastructure. The findings underscore the need for stronger teacher support systems, including improved resources, more targeted professional development, and enhanced administrator buy-in, to ensure all K-5 students benefit from comprehensive technology education.

There is greater demand for Computer Science (CS) to be taught in elementary education as more states pass policies requiring it. Integrating CS into elementary education provides a viable avenue to teach CS to all students and can result in more equitable outcomes. However, our research demonstrates many elementary educators have concerns about not having enough time, training, knowledge, curricula, resources, and/or support to teach or integrate CS in their general education classrooms. This session will bring together educators, curriculum developers, professional learning providers, researchers, and practitioners interested in fostering CS and computational thinking skills among young learners. Topics will include how CS is currently being taught or integrated into elementary education, reasons why teachers are or are not integrating CS, resources and supports to address barriers to teaching CS, and what is needed to increase the integration of CS in elementary education on larger scales.

Understanding teachers' motivation to teach computer science (CS) plays a significant role in recruiting, supporting, and retaining CS teachers. Prior literature has identified the existence of differences among female and male teachers in terms of their motivation to teach. The goal of the current study was to examine the psychometric properties and measurement invariance of the Revised Motivation to Teach Computer Science (MTCS_R) scale between female and male groups. The MTCS_R scale is a shortened, more concise version of the original Motivation to Teach Computer Science (MTCS) scale, which measures teachers' motivation to teach CS on a continuum from external to internal motivation. We used the MTCS_R scale to collect survey data during 2022 and 2023 from 310 educators enrolled in a professional learning course designed to prepare teachers for a CS certification exam. We then conducted a confirmatory factor analysis with all survey respondents (N=310) and further examined measurement invariance among those who disclosed their gender (N=298). Results from the confirmatory factor analysis suggested satisfactory psychometric properties of the MTCS_R scale. In addition, we identified strong evidence to support the configural, metric, and scalar invariance across the gender groups, confirming that the MTCS_R scale is a valid measure of motivation to teach CS for both females and males. This study represents a significant advancement in the measurement of motivation to teach CS. Implications of using this instrument to assess teachers' motivation in CS teaching and further refinement of the instrument are discussed.

Photoacoustic microscopy (PAM) leverages the optical absorption contrast of blood hemoglobin for high-resolution, multi-parametric imaging of the microvasculature in vivo. However, to quantify the blood flow speed, dense spatial sampling is required to assess blood flow-induced loss of correlation of sequentially acquired A-line signals, resulting in increased laser pulse repetition rate and consequently optical fluence. To address this issue, we have developed a sparse modeling approach for blood flow quantification based on downsampled PAM data. Evaluation of its performance both in vitro and in vivo shows that this sparse modeling method can accurately recover the substantially downsampled data (up to 8 times) for correlation-based blood flow analysis, with a relative error of 12.7 ± 6.1 % across 10 datasets in vitro and 12.7 ± 12.1 % in vivo for data downsampled 8 times. Reconstruction with the proposed method is on par with recovery using compressive sensing, which exhibits an error of 12.0 ± 7.9 % in vitro and 33.86 ± 26.18 % in vivo for data downsampled 8 times. Both methods outperform bicubic interpolation, which shows an error of 15.95 ± 9.85 % in vitro and 110.7 ± 87.1 % in vivo for data downsampled 8 times.

Uniquely capable of simultaneous imaging of the hemoglobin concentration, blood oxygenation, and flow speed at the microvascular level in vivo, multi-parametric photoacoustic microscopy (PAM) has shown considerable impact in biomedicine. However, the multi-parametric PAM acquisition requires dense sampling and thus a high laser pulse repetition rate (up to MHz), which sets a strict limit on the applicable pulse energy due to safety considerations. A similar limitation is shared by high-speed PAM, which also uses lasers with high pulse repetition rates. To achieve high quantitative accuracy besides good structural visualization at low levels of laser fluence in PAM, we have developed a new, sparse coding-based two-step denoising technique. In the setting of intravital brain imaging, we demonstrated that this unsupervised learning approach enabled the reduction of the laser fluence in PAM by 5 times without compromise of the image quality (structural similarity index measure or SSIM: >0.92) and the quantitative accuracy (errors: <4.9%). Offering a significant relaxation in the requirement of PAM on laser fluence while maintaining the quality of structural imaging and accuracy of quantitative measurements, this sparse coding-based approach is expected to facilitate the application and clinical translation of multi-parametric PAM and high-speed PAM, which have a tight photon budget due to either safety considerations or laser source limitations.

Existing CNN-based methods for semantic segmentation heavily depend on multi-scale features to meet the requirements of both semantic comprehension and detail preservation. State-of-the-art segmentation networks widely exploit conventional scale-transfer operations, i.e., up-sampling and down-sampling to learn multi-scale features. In this work, we find that these operations lead to scale-confused features and suboptimal performance because they are spatial-invariant and directly transit all feature information cross scales without spatial selection. To address this issue, we propose the Gated Scale-Transfer Operation (GSTO) to properly transit spatial-filtered features to another scale. Specifically, GSTO can work either with or without extra supervision. Unsupervised GSTO is learned from the feature itself while the supervised one is guided by the supervised probability matrix. Both forms of GSTO are lightweight and plug-and-play, which can be flexibly integrated into networks or modules for learning better multi-scale features. In particular, by plugging GSTO into HRNet, we get a more powerful backbone (namely GSTO-HRNet) for pixel labeling, and it achieves new state-of-the-art results on multiple benchmarks for semantic segmentation including Cityscapes, LIP, and Pascal Context, with a negligible extra computational cost. Moreover, experiment results demonstrate that GSTO can also significantly boost the performance of multi-scale feature aggregation modules like PPM and ASPP.

As a step towards scaling personalized instruction, we seek to automatically identify the key features of the interactive learning process teachers use to select the next activity when teaching a single student. Such features could both inform computational student models designed to facilitate instructional decisions, and help enable automated self-improving teaching systems that leverage this identified feature set. We present preliminary results that a very small set of features is almost as good as a much larger set of features at predicting human tutor decisions when teaching students about histograms.