Habte Tadesse LIKASSA

Habte Tadesse Likassa is a Research and Teaching Fellow at the College of Health Solutions, Arizona State University (ASU), USA. Previously, he served as an Assistant Research Scientist (Feb 2025 – Jun 2025) and Postdoctoral Research Scholar (Feb 2024 – Jan 2025) at ASU, focusing on development of methods to enhance the quality of high dimensional degraded retinal images for an early screening and medical diagnosis under the supervision of Prof. Ding-Geng Chen (Executive Director and Professor of Biostatistics) (https://search.asu.edu/profile/4022508 ), with co-supervision from Prof. Kewei Chen and Dr. Oana M. Dumitrascu of the Mayo Clinic, Arizona. His work on alternative retinal image enhancement methods has been published in reputable journals.

He earned his Ph.D. in Statistics with a focus on signal processing from the National Taiwan University of Science and Technology (2019). He also holds an M.Sc. in Statistics from Addis Ababa University (2011), where he studied the power of the spatial dependencies and Tobler’s Law of Geography in real-world spatial data, and a B.Sc. in Statistics from the University of Gondar (2008).

His doctoral research advanced Robust Principal Component Analysis (RPCA) methods to separate low-rank signals from high-dimensional imaging data corrupted by severe outliers and noise, supported by rigorous convergence proofs. This work, published in IEEE and other journals, extended to tensor RPCA for anomaly detection and robust regression models for head pose estimation and image recovery, with applications in radar imaging, facial anomaly detection, crime detection, and security in resource-limited settings. He is actively seeking collaborations and funding to translate these methods into practical technological solutions for crime detection and security. I am the author of the book entitled High-Dimensional Imaging Data Analysis: From Classical PCA to Tensor Robust PCA with AI/ML, Vol. 1, published by Springer written with Ding-Geng Chen. In addition, we developed a specialized medical imaging analyzer as part of this work, and a patent application has been filed through Arizona State University in the United States.

Building on his Ph.D. work, he now focuses on extending these methods to medical imaging including retinal, brain, and cancer-related images to improve diagnostic accuracy and early disease detection. He is passionate about developing tracer harmonization methods and exploring brain imaging related to Alzheimer’s disease. Along with his multidisciplinary team, including IT experts, he is translating RPCA methods combined with TWNN and AHE into practical tools, including an app to automatically enhance degraded retinal images, with deployment planned soon.

In 2024 and 2025, he delivered presentations on novel medical image enhancement methods at prestigious venues such as the University of Manchester (UK), Texas Statisticians Conference, ASU Blitz Talk, Addis Ababa University, and the Arizona Alzheimer’s Conference.

Before ASU, he was an Assistant Professor and briefly chaired the Department of Statistics at Addis Ababa University, and served as Teaching and Research Fellow and Vice Dean of the College of Natural and Computational Sciences at Ambo University, Ethiopia.

Personal Website: 

I delivered an engaging training on data analysis using R for PhD students and postdocs from public health, bioinformatics, engineering, and hospital settings, supported by the T32 project in the College of Health Solutions, Arizona State University, for 12 participants from August 11 to 15, 2025. The training provided an excellent opportunity to cover topics from basic to advanced statistical methods, including descriptive statistics, mean comparisons (such as t-tests, ANOVA, and ANCOVA), and various models, including linear regression model, multilevel and longitudinal models, as well as count models (Poisson, negative binomial and ....), and logistic regression, all applied to health data. 

Upon completion, the participants were certified and awarded certificates.