Researchers working under Dr. Teresa Lever are studying how neurological conditions may affect people's ability to swallow and breathe as they lose control over their vocal folds. To do this they use a software called VFTrack to analyze hundreds of gigabytes of video footage recorded using an endoscope camera inserted through the nose or mouth into the top of the throat to visualize the vocal folds. This software is made on campus by Dr. Filiz Bunyak and her team, in collaboration with Dr. Lever’s group.

Using the RISE storage, researchers are able to quickly access the large amounts of data necessary for their project. The RISE storage also enables researchers to take advantage of Hellbender, a high-power computing environment. This allows researchers to work with the data without having to download it to their local machines. It also allows researchers to ensure they are working with an updated, unified version of the software at all times.

Our mitochondria image segmentation process aims to identify mitochondrial regions in electron microscope images. Due to imaging and sample complexities, microscope parameters need to be adjusted to improve image characteristics and outcomes during the data analysis process. Current workflow processes rely on in-situ researchers for instrument setting adjustments, data collection and transmission to simulation centers. Once data is analyzed, a request for new data is submitted indicating the adjustments needed to improve the characteristics of the images. Given the involvement of the human effort on the process, it requires a long time to complete the process with the expected outcomes. Implementing a mitochondria segmentation process in our RISE system allows adjustment of the microscope parameters, without human intervention in real-time based on in-situ imagery, which expedites the process and provides successful results faster.

Our experiments help develop an automated workflow using the RISE storage system that involves image data import, data collection and accurate analysis, and a real-time feedback guidance with the analysis output results. These steps will benefit the researchers in their experiments by automatically assessing image data quality and adjusting microscope settings as well as imaging parameters.

CNT growth automation process involves the following steps:
1) in-situ SEM synthesis techniques to acquire imagery of CNT forest growth and self-assembly,
2) computer vision to quantify and isolate the kinetics and assembly mechanisms of CNTs,
3) a finite element CNT forest synthesis and testing simulation,
4) a convolutional neural network (CNN) to predict CNTforest properties from experimental and simulated images, and
5) a distributed control algorithm that will transition experimental control from human researchers to autonomous decision algorithms. As the CNT growth kinetics and self-assembly process are understood for diverse synthesis conditions, the CNN will predict the process-structure-property relationships for CNT forests.

To understand and map the process-structure-property relationships for CNT forests, we conduct storage-intensive experiments involving remote control and an iterative experimentation approach with integrated analysis, simulation, and feedback mechanisms to gradually remove humans, with their inherent error and bias, from the CNT property discovery loop.

The University of Missouri’s (MU) Electron Microscopy Core (EMC), as part of its involvement in the NextGen Precision Health Initiative, recently acquired a number of state-of-the-art microscopes. These devices that capture small molecule structures to cellular organization in tissue are capable of generating 5,000 to 420,000 images per day, leading to a significant need for data storage, processing and associated on-campus services.

Using the RISE storage, researchers will establish a mechanism for transferring data from VEC to a long-term solution with a capacity of 500TB within the RISE Storage infrastructure. This will ensure that space on VEC is regularly freed up, enabling continuous data collection. Additionally, storing data within the RISE storage platform will enable the provision of various data management features