Facilities and Resources:

 Laboratory:

Sundaresh Ram, PhD, Laboratory:

 Dr. Ram’s laboratory consists of two offices and shared space for postdoctoral fellows, technicians and students. Each devoted space includes a high-end Windows PC Workstation for signal & image processing and analysis.

 Computers:

There are three high-end Windows PC Workstations, and a mobile high-end laptop in the laboratory and Dr. Ram’s office, which are capable of prototyping and running data intensive computing jobs (for e.g., deep learning/artificial intelligence techniques and methodologies). Two of the workstations are Dell Precision 5860 Tower Workstations with Intel Xeon W3-2435 processor with 22MB cache, 8 cores, 16 threads. 4.5GHz turbo CPU, with 64GB RAM, 1TB hard disk space with an NVIDIA GPU (RTX A4000, 16GB GDDR6 RAM, 4DP), running windows OS. The other workstation is an artificial intelligence (AI) Dell Precision 7865 Tower Workstation with AMD Ryzen Threadripper PRO 5945WX processor with 64MB cache, 12 cores, 24 threads, 4.1GHz CPU, with 64GB RAM, 1 TB hard disk space with an NVIDIA GPU (RTX A4500, 20GB GDDR6 RAM, 4DP), running Windows OS. The laptop is a Dell Latitude 2-in-1 laptop that has an Intel Core i7-1370 CPU 3.9 GHz processor with 4 cores, 64 GB of RAM, 1 TB GB SSD hard disk space, with an NVIDIA GPU (GeForce GTX 1080 8 GB RAM), running Windows OS.

 Computing Resources:

The Department of Biomedical Informatics has HIPAA-compliant storage and compute environment involving a complex distributed PHI protected infrastructure consisting of an HPC cluster with a total of 19 nodes with a total of 764 CPU cores, 124544 CUDA (tm) GPU cores, 4.2 TBs of aggregated system RAM, 732 GB GPU RAM, and a parallel file system with 580TB of storage powered by ZFS & BeeGFS, running Centos Linux 7 OS, with relevant software packages to support all biomedical research and machine learning work.

Signal/Image/Video Processing and Analysis Toolkit: Dr. Ram’s laboratory has been active in developing a ‘toolkit’ of MATLAB and PYTHON signal/image/video processing and analysis methodologies. This toolkit speeds up and assists in analyzing both preclinical and clinical image data acquired internally at Emory and externally from other institutions. This toolkit includes open-sourced (e.g., 3DSlicer and Elastix) and developed signal/image/video processing and analysis, and machine learning and AI software packages. These packages include image restoration, image enhancement, image segmentation, image registration and motion estimation, image classification and phenotyping, and prediction. The toolkit also includes analytical techniques for calculating summary statistics, voxel-to-voxel techniques, machine learning and AI algorithms. Additionally, in-house algorithms have been developed to queue large datasets allowing processes to continue without user input.

 Office

Dr. Ram has an office space assigned to him in the Woodruff Memorial Research Building. The office has access to standard office machines such as copiPage in Progress