Parkinson's disease (PD) is a neurodegenerative disease that affects motor abilities with increasing severity as the disease progresses. Traditional methods for diagnosing PD include a section where a trained specialist scores qualitative symptoms using the motor subscale of the Unified Parkinson's Disease Rating Scale (UPDRS-III). The aim of this feasibility study was twofold. First, to evaluate quiet standing as an additional, out-of-clinic, objective feature to predict UPDRS-III subscores related to motor symptom severity; and second, to use quiet standing to detect the presence of motor symptoms. Force plate data were collected from 42 PD patients and 43 healthy controls during quiet standing (a task involving standing still with eyes open and closed) as a feasible task in clinics. Predicting each subscore of the UPDRS-III could aid in identifying progression of PD and provide specialists additional tools to make an informed diagnosis. Random Forest feature importance indicated that features correlated with range of center of pressure (i.e., the medial-lateral and anterior-posterior sway) were most useful in the prediction of the top PD prediction subscores of postural stability (r = 0.599; p = 0.014), hand tremor of the left hand (r = 0.650; p = 0.015), and tremor at rest of the left upper extremity (r = 0.703; p = 0.016). Quiet standing can detect body bradykinesia (AUC-ROC = 0.924) and postural stability (AUC-ROC = 0.967) with high predictability. Although there are limited data, these results should be used as a feasibility study that evaluates the predictability of individual UPDRS-III subscores using quiet standing data.

We discuss the design, effectiveness, and curricular impacts of an introductory data science course that uniquely meets a number of competing constraints. Student needs range from an initial exploration for future data scientists to the only computing course for some science students. Initially, associated research labs, departments, and industries provided differing perspectives on skills and preferred programming languages. The focus was on practical programming skills and learning through demos to develop useful skills quickly. The task-driven nature also enabled the mutual development of resources for each of the top three requested languages, Python, R, and MATLAB, allowing students to concurrently complete the course in any of the three languages. Tutorials in all three languages were produced to prepare students for the language-independent task-driven assignments given each day. For consistency all tutorials and assignments use the Jupyter notebook cell-based workflow. Feedback and performance measures of three student cohorts (Spring 2018, Fall 2018, and Spring 2019) are analyzed. Feedback included an average rating that met or exceeded 4 ("substantial progress") out of 5 for each separate cohort on all four general educational objectives (Basic Understanding, Developing Skills, Application, and Numeric Methods). 75% of responding students indicated a preference for the multi-language course design. The course also attracted a majority of women every semester (63% female, average). Because of these benefits, this course became the recommended first programming course for a newly developed and approved undergraduate data science major. All material is online and available at https://sites.google.com/view/comp-180

Identifying the trending topics in journals and conferences is valuable for understanding the role of authors, institutions, and funding agencies in the progression of knowledge produced in the field. However, many available clustering methods do not accommodate a desire for temporally clustered results that are typical of trends, in part because time of publication is often neglected as a feature. As a demonstration of how time can be emphasized in trend detection, we use a novel approach of introducing a weighted temporal feature to bias a topic clustering toward articles in a similar time frame; this is performed over a set of finance journal abstracts from 1974 to 2020. Latent Dirichlet Allocation (LDA) is used to parameterize each abstract, followed by dimensionality reduction using Singular Value Decomposition (SVD). We detect trending finance topics that are not identifiable when we use a standard clustering approach with no temporal bias. To identify trending topics, we utilize a metric of the silhouette score divided by the standard deviation of clusters over time. We then isolate topics identified by this metric and validate them using expert judgment. Our clustering strategy using temporal bias can be readily utilized in other fields for discovering the rise and fall of trends.

Radiation therapy (RT) of lung tumors requires an accurate and real-time localization of the tumor while the patient is being treated. Tumor motion caused by breathing can impact dose delivery in patients with lung cancer, leading to poor disease management and damage to surrounding normal tissues. A major challenge in tumor tracking using fluoroscopic imaging is to track the tumor while it is occluded by the overlapping bones (ribs and spine). In this work, we propose a series of three modeling strategies combining template matching for initial probability estimation and Kalman filter for sequential measurement updates. The images were acquired using a fast-kV switching real-time fluoroscope utilizing a dynamic thorax motion phantom. We utilized three distinct physics-based state representations to predict the motion of the tumor: Velocity, Acceleration, and Spring Motion. The Kalman filter with Spring Motion physics performed best with the average RMSE rates of 0.44 mm improving template matching alone with RMSE 1.81 cm. Additionally, occlusion and noise were simulated with general improvements from the Spring Motion Kalman filter. The proposed method will enable more accurate and precise lung radiotherapy using existing hardware and workflow. Our future work is focused on the clinical implementation of this method.

The Modified Ashworth Scale (MAS) is one of the most commonly used classification metrics for spasticity. While MAS is easy to implement, it relies on the subjective judgement of the clinician which leads to the poor inter-rater reliability between clinicians. We developed a device worn over the arm which measures angular velocity, angular acceleration, resistance, mechanical power, and elbow extension angle of the elbow joint. This device was used on 32 subjects who have spasticity in the upper forearm by 29 clinicians who evaluated the subjects, yielding a data set of 634 evaluations. Predictive models were developed including linear, decision tree, random forest, and support vector machines using variants of the models for regression and classification. Model hyperparameter tuning and testing was performed using clinician-wise nested cross-validation. The support vector classifier model was found to have the best root mean square error of 0.4449. Using a trained models can help eliminate much of the known variability in the MAS. By creating a more accurate way of assessing spasticity, better treatment plans can be more readily compared to aid in the treatment of spasticity and improve their quality of life.

Patterns of spontaneous activity in the developing retina, LGN, and cortex are necessary for the proper development of visual cortex. With these patterns intact, the primary visual cortices of many newborn animals develop properties similar to those of the adult cortex but without the training benefit of visual experience. Previous models have demonstrated how V1 responses can be initialized through mechanisms specific to development and prior to visual experience, such as using axonal guidance cues or relying on simple, pairwise correlations on spontaneous activity with additional developmental constraints. We argue that these spontaneous patterns may be better understood as part of an "innate learning" strategy, which learns similarly on activity both before and during visual experience. With an abstraction of spontaneous activity models, we show how the visual system may be able to bootstrap an efficient code for its natural environment prior to external visual experience, and we continue the same refinement strategy upon natural experience. The patterns are generated through simple, local interactions and contain the same relevant statistical properties of retinal waves and hypothesized waves in the LGN and V1. An efficient encoding of these patterns resembles a sparse coding of natural images by producing neurons with localized, oriented, bandpass structure-the same code found in early visual cortical cells. We address the relevance of higher-order statistical properties of spontaneous activity, how this relates to a system that may adapt similarly on activity prior to and during natural experience, and how these concepts ultimately relate to an efficient coding of our natural world.