Muhammad Junaid Ali

Data Scientist - Healthcare AI Developer
image

I am a third-year PhD research scholar at the IRIMAS Lab, Universite de Haute Alsace, located in Mulhouse City, France's Haut-Rhin department. Previously I have worked in Gomal University and Virtual University of Pakistan as a Tutor/Instructor teaching multiple computer science courses. Prior to this, I have worked as a research assistant at Medical Imaging and Diagnostics (MID) lab where I have proposed multiple approaches for brain tumor segmentation, survival prediction, and breast pectoral muscle segmentation.

CV Download

Work Experiences

Ph.D. Research Scholar

UHA | December 2021- December 2024
  • During my PhD I have worked on the problem of designing AutoML approaches for medical image analysis tasks. I have proposed multiple AutoML Neural Architecture Search appraoches for designing deep learning architectures for 2D/3D medical image classification and segmentation problems.
  • Furthermore, i have designed robust NAS approaches which are resiliant against different adversarial attacks on medical images and formulated a multi-objective NAS approaches for searching lightweight architectures for medical images.

TECHNICAL SKILLS

    Data Science: A/B testing, optimization, big data pipeline (cleansing, wrangling, visualization, modeling, interpretation), AutoML, Image Processing
    Medical Imaging: MONAI, nnUNET, nibabel, pydicom
    Web Development: Flask
    Programming Languages: Python (Pandas, scikit-learn, pytest, Tensorflow, PyTorch, SciPy, NLTK, Gensim), SQL, R, C++, Java
    Cloud Machine Learning: AWS (SageMaker, ECR, EMR, S3, RedShift), Spark, DataBricks, Airflow

Projects

Evolutionary Neural Architecture Search

Evolutionary Neural Architecture Search for 2D and 3D Medical Image Classification

Proposed an evolutionary Neural Architecture Search approach for 2D and 3D medical image classification and performed a comparative study of multiple metaheuristics for the given problem.

More details

Robust Neural Architecture Search

Robust Neural Architecture Search Using Differential Evolution for Medical Image

This study proposes an evolutionary Neural Architecture Search (NAS) approach for searching robust architectures for medical image classification. The Differential Evolution (DE) algorithm is used as a search algorithm. Furthermore, an attention-based search space consisting of five different attention layers and sixteen convolution and pooling operations is used.

More details

Tashkhees AI diagnostic app

Tashkhees

Contributed to the Tashkhees AI diagnostic app for breast cancer detection at MID Lab, COMSATS by incorporating breast cancer and pectoral segmentation AI module and building flask module and frontend using bootstrap.

More details

Surrogate assisted Genetic Algorithm

Designing Convolutional Neural Networks using Surrogate assisted Genetic Algorithm for Medical Image Classification

Proposed a surrogate assisted genetic algorithm for medical image classification.

More details

Brain Tumor Segmentation

Brain Tumor Segmentation and Survival Prediction

Proposed a deep learning approach for brain tumor segmentation and survival prediction.

More details

Challenge Ranking

Brain Tumor Segmentation

Hardware Aware Neural Architecture Search for Medical Image Classification

Proposed a multi-objective NAS approach based on the NSGA- II algorithm for searching lightweight architecture for medical image classification in this study.

Under Review

Brain Tumor Segmentation

Training Free U-Net for retinal Vessel Segmentation

Proposed a Neural Architecture Search approach for searching U-Net shaped architecture for retinal vessel segmentation.

More details

Publications

Journal Publications

  • Muhammad Junaid Ali, Mokhtar Essaid, Laurent Moalic, Lhassane Idoumghar. "A Review of AutoML Optimization Techniques for Medical Image Applications". Computrized Medical Imaging and Graphics
  • Nazir, Maria, Muhammad Junaid Ali, Hafiz Zahid Tufail, Ahmad Raza Shahid, Basit Raza, Sadia Shakil, and Khurram Khurshid. "Multi-task learning architecture for brain tumor detection and segmentation in MRI images." Journal of Electronic Imaging 31, no. 5 (2022): 051606-051606.
  • Muhammad Junaid Ali, Basit Raza, Ahmad Raza Shahid. Multi-level Kronecker Convolutional Neural Network (ML-KCNN) for Gliomas Segmentation from Multi-Modal MRI Volumetric Data. Journal of Digital Imaging.
  • Muhammad Junaid Ali, Basit Raza, Ahmad Raza Shahid, et al. Enhancing breast pectoral muscle segmentation performance by using skip connections in fully convolutional network. Int J Imaging Syst Technol. 2020;1–11.
  • Asra Rafi, Tahir Mustafa Madni, Uzair Khan, Muhammad Junaid Ali et al. Multi-Level Dilated Convolutional Neural Network for Brain Tumor Segmentation and Multi-View based Radiomics for Overall Survival Prediction. Int J Imaging Syst Technol. 2021;1–11.

Conference Publications

  • Muhammad Junaid Ali, Laurent Moalic, Mokhtar Essaid, and Lhsanne Idoumghar. 2024. Evolutionary Neural Architecture Search for 2D and 3D Medical Image Classification. 24th International Conference on Computational Science.
  • Muhammad Junaid Ali, Laurent Moalic, Mokhtar Essaid, and Lhsanne Idoumghar. 2024. Robust Neural Architecture Search using Differential Evolution for Medical Images. In 27th International Conference on Applications of Evolutionary Computing (EVOAPPS ’24)
  • Muhammad Junaid Ali, Laurent Moalic, Mokhtar Essaid, and Lhsanne Idoumghar. 2023. Designing Convolutional Neural Networks using Surrogate assisted Genetic Algorithm for Medical Image Classification. In Genetic and Evolutionary Computation Conference , July 15–19, 2023, Lisbon, Portugal. ACM, New York, NY, USA https://doi.org/10.1145/3583133.3590678
  • Haris Ali Khan, Muhammad Junaid Ali, Umm e Hanni. Poster: A novel approach for POS tagging of Pashto. 1st Int conference on smart technologies.
  • Muhammad Junaid Ali, Laurent Moalic, Mokhtar Essaid, and Lhsanne Idoumghar. Training Free U-Net for Retinal Vessel Segmentation [Under Review]
  • Muhammad Junaid Ali, Laurent Moalic, Mokhtar Essaid, and Lhsanne Idoumghar. Hardware-Aware Neural Architecture Search (HW-NAS) for medical image analysis tasks. [Under Review]
  • Rafi, Asra, Ali MJ et al. "U-Net Based Glioblastoma Segmentation with Patient’s Overall Survival Prediction." International Symposium on Intelligent Computing Systems. Springer, Cham, 2020.
  • Muhammad KaleemUllah Khan, Nadeem Javaid, Shakeeb Murtaza, Maheen Zahid, Wajahat Ali Gilani, and Muhammad Junaid Ali. "Efficient Energy Management Using Fog Computing", in the 21st International Conference on Network-Based Information System (NBiS-2018).
  • Muhammad Junaid Ali, Muhammad Tahir Akram, Hira Saleem, Basit Raza, and Ahmad Raza Shahid. "Glioma Segmentation Using Ensemble of 2D/3D U-Nets and Survival Prediction Using Multiple Features Fusion." International MICCAI Brainlesion Workshop. Springer, Cham, 2020. Status: Accepted
  • Muhammad Junaid Ali, Nadeem Javaid, Mubariz Rehman, Muhammad Usman Sharif, Muhammad KaleemUllah Khan, and Haris Ali Khan. "State Based Load Balancing Algorithm for Smart Grid Energy Management in Fog Computing", in 10-th International Conference on Intelligent Networking and Collaborative Systems (INCoS-2018).
  • Saif, Talha, et al. "Round Robin Inspired History Based Load Balancing Using Cloud Computing." International Conference on P2P, Parallel, Grid, Cloud and Internet Computing. Springer, Cham, 2018.
  • Rehman, Mubariz, et al. "Threshold based load balancer for efficient resource utilization of smart grid using cloud computing." International Conference on P2P, Parallel, Grid, Cloud and Internet Computing. Springer, Cham, 2018.