Fatima Zerari Boukari, PhD, Visiting Assistant Professor in the Division of Physics, Mathematics and Computer Science at Delaware State University
Dr. Boukari is researching the application of machine learning methods to provide an automatic, quantitative, diagnostic tool for rapid and efficient detection and diagnosis of lung diseases based on chest X-ray images and other imaging modalities, and to provide insight into the evolving conditions of the lungs after infection.
- What is the importance of this research?
According to the National Heart, Lung, and Blood Institute (NHLBI) lung diseases affect tens of millions of people. In the U.S., more than 26 million people—including over 6 million children—have asthma, and nearly 16 million people have chronic obstructive pulmonary disease (COPD). One of the main ways to distinguish between different lung diseases is through x-ray imaging and CT scans. The significance of accurate imaging of lung disease became more important during the COVID-19 pandemic, which causes lung damage. Automated medical imaging could increase the speed of lung disease diagnosis and reduce pressure on public health systems during this pandemic.
2. Why did you want to conduct this research?
Machine learning and deep learning, in particular, has emerged as a powerful AI tool to detect, identify and characterize features in images. Initially, I was interested in extending the deep learning approach to build a fast learned-model to detect and classify signatures of respiratory diseases based on X-ray images from patients. The project took a more interesting turn when COVID-19 was found to be caused by SARS-CoV-2, which triggers a respiratory tract infection. Working to improve the diagnosis and monitor changes in the disease (COVID-19 or other respiratory diseases) has become my main focus.
3. How does this research relate to your other work?
Although I hold a PhD degree in Applied Mathematics and Physics, I am trained as a computer scientist and my passion is in developing and applying advanced computational approaches and analytical mathematical methods in various applications. Recently, I have directed my efforts towards the field of data science, especially deep learning, where we have seen an explosion of applications in diverse areas of applied research. In addition, I am a faculty member at Delaware State University where I teach diverse courses in computer science and I mentor students interested in data science.
4. What aspect of the DE-CTR was most helpful to you for this research?
As a junior investigator, the DE-CTR ACCEL Program and its support have been critical to jump start my research career. Delaware State University is distinctly a primarily undergraduate institution with a growing research portfolio, especially in STEM areas. Although short in time scale (3 months) the ACCEL support has not only been important to conduct the objectives of my project, to train students, to collect initial data and preliminary results, but more importantly, to connect me with the research community in Delaware and elsewhere. Furthermore, with the results, I have spent my summer working on putting together a proposal to broaden the scope of this project and to build collaborations, which I hope to submit to NSF or NIH.
5. What advice would you give to a junior researcher?
If you are working in Delaware and interested in research, there are many opportunities and funded programs in different areas, including biomedical research. There is a very motivated research community and, being in a small state (but the First State) it is easier to connect, to build collaborations, to get information, to join large projects, and to reach out to other communities, among others. Just Engage and Connect!