About me

Short Bio

Sinan Kockara received his BS degree in Computer Engineering in 2001 at Dokuz Eylul University, Izmir, Turkey. In 2008, he received his Ph.D. degree in Applied Computing at University of Arkansas at Little Rock. Since 2008 he is with Department of Computer Science at University of Central Arkansas. He is currently an Associate Professor in the same department. Research interests include biomedical informatics, medical image processing & analysis especially for dermoscopy, surgical simulation development for virtual reality environments, parallel computing, and GPGPU. He is an IEEE member since 2006. He is also a member of Sigma Xi scientific research society. He is the director of Biomedical Image Processing and High Performance Computing (BioMed-HPC) lab. Some of the recent and funded active research projects are:

Distinguishing malignant pigmented lesions from benign lesions

 Skin cancer is the most common form of cancer in the US and over 3.5 million cases are diagnosed annually. Skin cancer detection is the most important indication of dermoscopy. We develop algorithms to distinguish malignant pigmented lesions from benign lesions using dermoscopy images. We extract dermoscopic features for detection of neoplastic behavior.
   This research has been partially supported by Arkansas Science and Technology Association Award# 15-B-25.

Virtual Arthroscopic Tear Diagnosis and Evaluation Platform (VATDEP)

Shoulder arthroscopy is a minimally invasive surgery for diagnosis and repairing of the tissues/joints in  the shoulder area. Virtual Reality (VR) based surgical simulators offer a realistic, low cost, realistic risk-free training and assessment platform where the trainees can repeatedly perform tasks and receive quantitative feedback on their performances.  In our ongoing study we are developing a VR arthroscopic rotator cuff repair surgery simulator with Dr. Tansel Halic. 
​   This research has been partially supported by NIH.

Dynamic Voxelization for Virtual Rotator Cuff Surgery

In the arthroscopic rotator cuff surgery, drilling a suture anchor into the humeral head is one of the critical tasks. We use voxel based method for realistic real-time virtual simulation of drilling. Voxelization allows ability to create convex and concave surfaces and enables the robust haptic interactions during drilling. However, the voxelization cannot represent exact geometry of the humeral head. This worsens especially with large voxel sizes. On the contrary, the use of finer and fixed resolution voxel sizes could be computationally not attainable. Therefore, a dynamic voxelization is highly desirable for realistic and computationally efficient interaction. In this project, we introduce a novel voxelization method based on dynamic proximity hierarchy (DPH). DPH is a graph spanner based hierarchical representation of approximate shortest paths of the points in the geometry.
    This research has been partially supported by NIH.