Home > About > Faculty Profiles > Fatima Boukari, Ph.D.

Fatima Boukari, Ph.D.

Boukari F

Associate Professor in Computer Science
Engineering Division of Physics, Engineering, Mathematics and Computer Sciences

Email: fboukari [at] desu.edu

Education

  • B.Sc., Computer Science Engineering, University of Annaba, Algeria
  • M.Sc., Computer Systems Architectures & M.Sc., Parallel Computing, Algeria-Glasgow
  • Ph.D., Mathematics & Physics, Delaware State University

Bio

Dr. Fatima Boukari’s research interests have evolved over her career path. She started working on highly focused computer systems such as operating systems, graph theory, simulation, parallel machines, and the implementation of computing models on parallel distributed architectures. Then she transitioned into interdisciplinary research, including the development of mathematical prediction-based algorithms, AI, classification, statistical approaches, transfer learning and physics-based models to address complex problems.

During her Ph.D., she developed new algorithms and computational tools for medical diagnosis and biomedical applications. Her research focuses on foundational Deep Learning architectures, mathematical modeling, and AI techniques, applied across various fields to build trustworthy and robust AI-driven solutions. She has worked on key architectures and fundamentals such as optimization techniques and machine learning algorithms applied in multidisciplinary areas such as computer vision, gaming, healthcare, autonomous systems, sensors, and robotics.

With newly acquired research grants from NSF CISE, she is working on building Deep Learning models for spectral and signal data analysis and prediction, applied to medical diagnosis, analysis of treatment responses.

In another DoD Air Force research project on tactical autonomy, she is developing robust tactical autonomy solutions to enhance multi-domain situational awareness and improve decision-making in complex environments using reinforcement learning, generative AI, as well as building machine learning models using a multi-modal distributed data training framework that maintains privacy and efficiency without data sharing. Additionally, she is focusing on transfer learning, reinforcement learning, and distributed learning to enhance decision-making in dynamic environments.

Her projects also include mathematical modeling, supported by a recently acquired grant from UARC RITA Air Force (PI) for metacognition modeling and understanding of awareness and human perception. She is working on modeling and simulating human cognitive mechanisms such as memory, learning, and decision-making. Her team has recently developed brain EEG signal modeling using PDEs, neural networks (CNNs), recurrent neural networks (RNNs), Residual Neural Networks, Generative Adversarial Networks (GANs), and Variational Autoencoders (VAEs). Collaborative projects with DSU involve building trustworthy machine learning algorithms, particularly in environmental and agricultural applications, including smart farming projects with professors from the Center of Excellence in Emerging Internet of Things (E-IoT) Technologies. She serves as a Data Science and AI ambassador at DSU for the HBCU Data Science Consortium. She is the core representative of DSU at the National Center for Women & Information Technology (NCWIT), an advocate for women in science and engineering, a member of SPIE women in Optics, and IEEE Women in Engineering. She serves as a reviewer for multiple journals and is an active force in strengthening educational and research infrastructure at Delaware State University. 

Research interests & Projects

  • Cell delineation and segmentation: In 2016, we proposed and developed an automated cell quantification and segmentation system in video sequences using non-linear motion diffusion partial differential equations-based formulation in the joint spatio-temporal domain for cell motion detection in each three consecutive frames. In addition, we introduce an intensity standardization technique using histogram transformation to address intensity variability complicating frame-to-frame analysis in differential techniques followed by watershed segmentation to delineate cells, using different kernel size parameters for best delineation. To further refine cell delineation accuracy produced by motion diffusion-based segmentation, we propose to use energy minimizing geometric active contours that assume a piece-wise constant image region model as a special case of the Mumford-Shah region based levelset optimization framework. We then applied mathematical and morphological operations to separate cells in cell clusters to reduce under segmentation. Furthermore, we introduce temporal linking of the region-based level sets to allow for faster convergence and to resolve non-convexity that affects energy-based minimization that is typical in image analysis inverse problems. An ongoing project consists on building Machine learning and Deep Learning algorithms for accurate segmentation, lineage, classification and Tracking and tracking of cellular structures over time, thereby enhancing our understanding of cell dynamics in biological processes.
  • Cell tracking: We developed a fully automated tracking system. The goal of cell tracking is to identify all cells throughout the time-lapse sequence to follow their motion and detect the main events such as migration, mitosis, apoptosis, entering and leaving the field of view. We propose a method for automated tracking of biological cells in time-lapse microscopy by motion prediction and minimization of a global probabilistic function for each set of cell tracks. We identify cell events by backtracking the cell track stack and forming new tracks to determine a partition of the complete track set. A pre-tracking stage is required to separate cell clusters to reduce under-segmentation, and to compute cell characteristics to be used for probabilistic cell matching and cell quantification. Then we estimate the cell motion by using a variational multi-scale optical flow technique. Next, we apply the motion field to calculate warped cells, and we apply a maximum likelihood decision approach on a probabilistic function to find cell correspondences. We then construct the cell linked lists to represent cell tracks and we backtrace the lists to detect overlapping tracks and identify and handle the cell events. After finding all cell events we construct the cell lineage tree that stores and visualizes the cell events. In addition, we calculate cell characteristics and their evolution with time to perform quantitative analysis and visualization. Our system automatically constructs the lineages of proliferating migrating cells, which is a critical and required step for understanding cell behaviors. It produces measurements of static and dynamic cell attributes. The proposed system is applicable to varying cell shapes, types, densities and image sequences of reduced image quality. Overall, it enables accurate quantitative analysis of cell events and provides a valuable tool for high-throughput biological studies.
  • Representational Assessment Tools for Neuroscience, Computational Modeling, and Artificial Intelligence Engineering: Investigating novel meta-learning techniques that apply advanced computational analyzes to understand intricate patterns in brain activity.
  • Building resiliency and robustness in Multi-Modal Distributed Learning systems: The integration of AI is revolutionizing surveillance, and reconnaissance capabilities, with technological advancements shaping new approaches and solutions for augmenting military decision-making. Enhancing ISR capabilities through the assimilation of diverse multimodal architectures and restructuring as a decentralized network of surveillance systems has the potential to elevate surveillance and decision-making capabilities. To solve mixed data input, we allow federated training over multimodal distributed data without assuming similar active sensors, but this model lacks the ability to impute missing data values or maintain their effectiveness in the event of the sudden loss of one or more sensors. The objective of this project is to propose a robust, resilient, and adaptive distributed learning system that addresses the distributed learning and distributed ISR architectures challenges. The scope of this project is to address the challenges of training ML models across diverse heterogeneous systems and propose an efficient Learning of Deep Networks from Multi-Modal Decentralized Data.
  • AI in Healthcare and medical diagnostic: This project includes Deep Learning and Computer Vision tools for medical diagnosis in cancer and brain cognition deterioration using. Exploring AI-driven approaches to enhance diagnostic accuracy and treatment customization with patient response to treatment analysis for complex health conditions.

In addition to my primary research initiatives, a variety of innovative projects have been developed and coded by my undergraduate research assistant students. Through guidance, support and mentorship, students build their own machine learning models, Generative AI, Reinforcement Learning, Deep Learning and Computer Vision tools for medical diagnosis in cancer and brain cognition deterioration, as well as a variety of projects in virtual reality, agriculture, business, signal analysis, autonomous systems and autonomous agents.

Student research projects for training and research

  • Deep Learning models for Cell segmentation and Tracking
  • Multimodal Deep Learning for early leukemia diagnosis and rapid therapy response analysis 
  • Machine Learning mine detection using SONAR data
  • Autonomous driving using LIDAR
  • Healthcare Sampling techniques to overcome disparities arising from biomedical data inequality
  • Automated Image for Histopathological assessment (skin Cancer detection and classification) Automated image analysis combined with machine learning algorithms to identify and classify various types of cancerous lesions
  • Real-Time Monitoring of EKG heart signals
  • Knowledge Graphs from text using LLMs; Researching the creation and utilization of knowledge graphs using large language models to enhance data interpretation and accessibility and brain cognition
  • Meta-Learning Mathematical and Computational Analysis of Brain activities
  • Building Web and phone Apps for development of practical applications that have the potential to make meaningful contributions to the fields
  • Multi-Modal Distributed Learning
  • Fine Tuning LLMs for mental health
  • AI in Agriculture, Cybersecurity, Business
  • Human Machine Teaming and Tactical Autonomy

grants

  • PI DE-CTR ACCEL project: Deep Learning Techniques for the automatic Diagnosis of COVID-19 Respiratory Diseases using Big Data and Convolutional Neural Networks
  • PI Data Science South Hub: Title: Introducing and Promoting Data Science in Education and Research at Delaware State University
  • PI: Air Force RITA/UARC project TO2#14:  Representational Assessment Tools for Neuroscience, Computational Modeling, and Artificial Intelligence Engineering
  • PI Air Force project Building Robustness and resiliency in Heterogeneous Multi-Modal Distributed Learning Systems
  • Project Lead at the Delaware State University CAST E-IoT Center, a hub for research and education in agriculture emerging technologies. This was established and propelled under four (4) interdisciplinary research thrusts (IRT): 1)IRT1: “Integrating Emerging Technologies in Research and Education for Workforce Development; 2) IRT2: “Climate Smart Agriculture;” 3) IRT3: “Precision Agriculture and Livestock for Smart Farming;” 3) IRT4: “Precision Food Processing and Sustainable Agricultural Systems
  • Team Lead at the 1890 Working Group on Artificial Intelligence (AI) to develop responsible and innovative AI systems solutions in the field of adapting and mitigating the effects of climate change, agricultural resilience and sustainability, energy security and global food security. The 1890 working group was established in January 2024 to explore the current state, potential applications, and ethical implications of AI technologies, enhance knowledge of AI and its potential to support research, teaching, outreach/engagement, Promote AI Literacy and Invest in Workforce Development and explore AI’s Long-Term Impacts on Food, Agriculture and Natural Resources.
  • Co-PI NSF CISE grant “CISE-MSI: DP: III: Training and Partnership in Data Science for Advancing Research in Biomolecular Detection.” This is a collaborative project with the University of Delaware and the University of Virgin Islands. The project focuses on the development and application of multi-modal, physics-informed machine learning models for the detection and characterization of biomolecules with the use of spectral data. Further, the project provides support for training students in data science and machine learning.
  • Lead Scientist, NSF grant “Collaborative Research: HDR DSC: Delaware and Mid-Atlantic Data Science Corps”. This is a collaborative project with the University of Delaware and Lincoln University. The project focuses on building capacity in data science to train students in data science.
  • Research scientist member of the national AI Institute for Climate-Land Interactions, Mitigation, Adaptation, Tradeoffs and Economy, (AI-CLIMATE) which is one of seven National AI Research Institutes (NAIRIs) announced on May 4, 2023. AI-CLIMATE is a joint effort of the University of Minnesota Twin Cities (lead), Colorado State University, Cornell University, Delaware State University, North Carolina State University, and Purdue University. Supported by the United States Department of Agriculture - National Institute of Food and Agriculture and award #2023-67021-39829 in collaboration with the National Science Foundation.

Selected Publications

  • Jinjie Liu, Qi Lu, Hacene Boukari and Fatima Boukari, “Iterated Crank–Nicolson Runge–Kutta Methods and Their Application to Wilson–Cowan Equations and Electroencephalography Simulations”, Journal paper ID: foundations-3108690,  Type of manuscript: Artic, Foundations 2024, 4(4), 673-689; https://doi.org/10.3390/foundations4040042 - 13 Nov 2024
  • Hobbs A, Stephenson K, Ferretti T, Ferretti M, Boukari F, Pokrajac D. Creation of database for detection of unusual behaviour in surveillance videos. 2013 11th International Conference on Telecommunications in Modern Satellite, Cable and Broadcasting Services (TELSIKS). TELSIKS 2013 - 2013 11th International Conference on Telecommunication in Modern Satellite, Cable and Broadcasting Services; ; Nis, Serbia. IEEE; c2013. Available from: http://ieeexplore.ieee.org/document/6704899/ DOI: 10.1109/TELSKS.2013.6704899, SCV Biographical Sketch v.2024-1 Page 1 of 3
  • Boukari F, Makrogiannis S. “Spatio-temporal Level-Set Based Cell Segmentation in Time-Lapse Image Sequences”. Lecture Notes in Computer Science [Internet] Cham: Springer International Publishing; 2014. Chapter 541-50p. Available from: http://link.springer.com/10.1007/978-3-319-14364-4_5 DOI: 10.1007/978-3-319-14364-4_5
  • Boukari F, Makrogiannis S. Spatio-temporal diffusion-based dynamic cell segmentation. 2015, IEEE International Conference on Bioinformatics and Biomedicine (BIBM). 2015 IEEE International Conference on Bioinformatics and Biomedicine (BIBM); ; Washington, DC, USA. IEEE; c2015. Available from: http://ieeexplore.ieee.org/document/7359701/ DOI: 10.1109/BIBM.2015.7359701
  • Boukari F, Makrogiannis S. Automated Cell Tracking Using Motion Prediction-Based Matching and Event Handling. IEEE/ACM Trans Comput Biol Bioinform. 2020 May-Jun;17(3):959-971. PubMed Central PMCID: PMC6832744.
  • Boukari A, Boukari F. Transfer and deep learning techniques for the diagnosis of COVID-19 respiratory diseases. In: Azar F, Intes X, Fang Q, editors. Multimodal Biomedical Imaging XVI. Multimodal Biomedical Imaging XVI; ; Online Only, United States. SPIE; c2021. Available from: https://www.spiedigitallibrary.org/conference-proceedings-ofspie/11634/2… DOI: 10.1117/12.2579006
  • Boukari F, et all, Assessing Raman spectra of concentrated carbohydrate solutions. SPIE Photonics Wets; c2023. Available from: https://spie.org/photonics-west/presentation/Assessing-Raman-spectraof-c… other-id: SPIE
  • Rivera-Lopez Y, Salih M, Boukari F, Barnett C, Boukari H. Intensity correlation analysis of Raman spectra of concentrated Ficoll solutions. In: Maitland K, Roblyer D, Campagnola P, editors. Multiscale Imaging and Spectroscopy III. Multiscale Imaging and Spectroscopy III; San Francisco, United States. SPIE; c2022. https://www.spiedigitallibrary.org/conference-proceedings-of-spie/11944/…, 119440B (2022), SPIE BiOS, 2022, San Francisco, CA.
  • Makrogiannis S, Boukari F, Ferrucci L. Automated skeletal tissue quantification in the lower leg using peripheral quantitative computed tomography. Physiology Meas. 2018 Apr 3;39(3):035011. PubMed Central PMCID: PMC5933065.
  • Boukari F, Makrogiannis S, Nossal R, Boukari H. Imaging and tracking HIV viruses in human cervical mucus. J Biomed Opt. 2016 Sep 1;21(9):96001. PubMed Central PMCID: PMC5010625.
  • Boukari F, Makrogiannis S. Joint level-set and spatio-temporal motion detection for cell segmentation. BMC Med Genomics. 2016 Aug 10;9 Suppl 2(Suppl 2):49. PubMed Central PMCID: PMC4980781.

Selected Presentations

  • Ian Mitchel, Qi Lu, Fatima Boukari, Title: “The interaction of music with the brain state can cause change in the electroencephalogram (EEG)” , presented our EEG data that was collected at DSU using BIOPAC, 2024 Annual Meeting of the APS Mid-Atlantic Section, at Temple University, November 15 - 17, 2024
  • Daniel Boyce, Jinjie Liu, Qi Lu, Fatima Boukari, Presentation, (submitted paper), SPIE Biophotonics Conference: Neural Imaging and Sensing 2025 Conference, January 2025, Program: https://spie.org/BO201,   Paper Number: 13303-23:, Machine learning analysis of electroencephalography (EEG) signals for sentiment categorization, Application tracks: Brain Function , AI/ML
  • Fatima Boukari, “Innovative Ideas: Computing pathways and Computing in Undergraduate education”,  Level Up Workshop, November 27th 2023, Boston Level UP
  • NIH Presentation, November 16th 2023, Fatima Boukari: “AI and Machine Learning for Brain activity and behavioral research”, Bethesda, MD, NIH
  • A. Boukari and Fatima Boukari, Transfer and deep learning techniques for the automatic diagnosis of respiratory diseases, Date January 2021, Conference presentation: SPIE BiOS 2021
  • Y. Lopez, M. Salih, F. Boukari, et al., Intensity correlation analysis of Raman spectra of concentrated Ficoll solutions, January 2022, Multiscale Imaging and Spectroscopy III. Multiscale Imaging and Spectroscopy III, Conference: SPIE BiOS 2022
  • Presentation for the Community Research Exchange with Nemours hospital, UD, Christiana Care, DSU and Musc Health. Title: “Transfer and Deep Learning Techniques for the automatic Diagnosis of COVID-19 Respiratory Diseases using Big Data and Convolutional Neural Networks”.
  • More than 15 research projects presentations by Students participating at Delaware State research day: Delaware State Research Day April 2021; Delaware State Research Day April 2022; Darwin Day 2020, 2021, 2023, 2024, 2025; Delaware State Research Day April 2023; Delaware State Research Day Presentation, April 2024.
  • Ian Hayes and Fatima Boukari, “Resiliency and Robust Solutions for Heterogeneous Federated Learning”, Annual Delaware State University Research Day, Wednesday, April 10, 2024.
  • Lillie Hunter and Fatima Boukari, “The Ethics of AI Art and web application development”, Annual Delaware State University Research Day, Wednesday, April 10, 2024.
  • Myles Davie and Fatima Boukari, “Multi-Modality and Data Fusion in Machine Learning”, Annual Delaware State University Research Day, Wednesday, April 10, 2024.
  • FLL Lego League students Oral Presentations 2023
  • Three Poster Presentations at Provost day at UD, Spring 2024
  • Presentation, Friday, February 4, 2022, at 12:30 pm 12:50 pm, at the HBCU Data Science Consortium Celebration
  • F. Boukari et all, 13th IEEE Integrated STEM Education Conference - Saturday, March 11, Laurel, Maryland and Delaware Bio (April 2023)
  • SPIE Presentation by students 2024: Yahira Rivera, S-A Elelu, M. Robinson, M. Carattini Colon, Fatima Boukari and H. Boukari, “ Optical Studies of concentrated ficoll and Glycol solutions”, SPIE. Photonics West, January 28th 2024; Fatima Boukari, “ResNet-Based Deep Learning Modeling of Bacteria and Antibiotic”, SPIE. Photonics West, January 28th, 2024. 
  • DARWIN high performance computing Symposium, February 12th, 2024, University of Delaware
  • DARWIN and HPC computing day Poster Presentations by students, 2021, 2022, 2023, 2024 and 2025
  • Student Presentation at the Provost Symposium at UD, March 14, 2024
  • Abstract accepted for competition at the HBCU Data Science Consortium 3rd Annual Celebration
  • Spring 2024 PEMaCS Day Presentations and Capstone Projects, Friday, April 26, 2024
  • Presentation at DSU, July 2024 (INBRE) projects

Additional Activities

Student Training: Mentored and co-mentored graduate (2) and undergraduate (> 40) students from diverse backgrounds, mostly from underrepresented in STEM fields, in interdisciplinary fields, including Machine Learning, Artificial Intelligence, LLMs, Computer Vision, computer simulation, database systems, Diffusion Models, Transformers, signal processing, Deep Learning and Data Science in multi-disciplinary fields.

  • Engaging students in seminars, workshops, trainings, data science, research conference presentations, hackathons, competitions, summer internships, technical certificates, job interview preparation

Strengthening educational and research infrastructure: 

  • Lead efforts to promote data science at DSU, to build computer infrastructure for machine learning, and to train a diverse workforce
  • Co-led efforts for the establishment of the laboratory for Applied Interdisciplinary Data Science (AIDA Lab) at DSU

Course development and Workshops: 

  • Developed and applied innovative project-oriented approaches to improve teaching outcomes in computer science, especially the courses in data mining, algorithmics, compilers, and probability and statistics
  • Presented multiple workshops on Data Science, Data Visualization and Artificial Intelligence
  • Organized the first workshop on data science at DSU, which offered to DSU summer interns in 2021
  • Following that year, Prepared and organized multiple workshops on Data Science Fall 2022, Spring 2023 and Summer 2023 and Summer 2024 to train all Delaware State University interns on Data Science and AI

Events

  • Organized seminar on Introducing and promoting Data Science, AI and machine learning at Computer Science Department of Delaware State University Fall 2021
  • Participated in professional meeting: “Image analysis across domains”, Berkeley Institute of Data Science and ADSA at ImageXD 2023, March 16-17, 2023 at the National Academies Beckman Center in Irvine, CA
  • DARWIN Symposium at the University of Delaware, February 12th, 2023, “Bayesian non-parametric Statistical Learning” 
  • Presentation at BioBriefing June, 2023
  • F. Boukari et all, 13th IEEE Integrated STEM Education Conference - Saturday, March 11, Laurel, Maryland and Delaware Bio (April 2023)

Professional Memberships + More

  • Member of ACM Baltimore Chapter and IEEE Baltimore ComSoc Chapter
  • DSU Data Science Ambassador
  • Member of Women in Science, ACM
  • Data Science and AI ambassador at DSU for the HBCU Data Science Consortium
  • Core representative of DSU at the National Center for Women & Information Technology (NCWIT)
  • Advocate for women in science and engineering
  • SPIE women in Optics, IEEE Women in Engineering. IEEE women in Science
  • Scientific reviewer for multiple journals