Signal and Image Processing (SIP) Research Group
The SIP research group is focused on the applications of multidimensional signal processing to solve medical challenges and problems related to the content distribution. The group is dedicated towards the development of innovative imaging techniques and computational algorithms for efficient content distribution and advanced processing of physiological signals for better understanding, diagnosis, and treatment of human diseases.
Through coordination with multiple faculties, university departments and institutes, the SIP research group strives to become the leading and well-renowned research group in the area of signal and image processing. The research group strongly welcomes the involvement of young graduate students in the research activities of the group and has full potential to provide the essential training programs to enable them to learn state-of-the-art techniques and methods. We believe that their involvement will enable them to become outstanding independent researchers.
Key Research Themes
- Multidimensional Signal Processing
- Biomedical Imaging and Analysis
- Multimedia Compression and Communication
- Machine Learning
Research Leads
Prof. Dr. Gulistan Raja
PhD (UET Taxila) MSc (Osaka University, Japan)
Professor, Department of Electrical Engineering,
University of Engineering and Technology, Taxila, Pakistan.
Email: [email protected]
Phone: +92 51 9047549
Dr. Furqan Shoukat
PhD (UET Taxila) MSc (UET Taxila)
Associate Professor, Department of Electronics Engineering,
University of Engineering and Technology, Taxila, Pakistan.
Email: [email protected]
Phone: +92 543-551278
Dr. Laiq Ur Rahman Shahid
PhD (Jacobs University, Germany) MSc (UET Taxila)
Assistant Professor, Department of Electronics Engineering,
University of Engineering and Technology, Taxila, Pakistan.
Email: [email protected]
Phone: +92 51 9047XXX
Dr. Junaid Mir
PhD (University of Surrey, UK) MSc (UET Taxila)
Assistant Professor, Department of Electrical Engineering,
University of Engineering and Technology, Taxila, Pakistan.
Email: [email protected]
Phone: +92 51 9047548
Selected Publications
2020
- M. Kashif, Gulistan Raja, and Furqan Shaukat, "An Efficient Content based Image Retrieval System for Diagnosis of Lung Diseases", Journal of Digital Imaging, Vol. 33. Issue 4, , pp. 971-987, Sep. 2020
- I. Bibi, Junaid Mir, and Gulistan Raja, "Automated Detection of Diabetic Retinopathy in Fundus Images using Fused Features", Physical and Engineering Sciences in Medicine, Published Online, Sep. 2020
2019
- Furqan Shaukat, Gulistan Raja, R. Ashraf, S. Khalid, M. Ahmad and A. Ali, “Artificial Neural Network based Classification of Lung Nodules in CT images using Intensity, Shape and Texture Features,” Journal of Ambient Intelligence and Humanized Computing, vol. 10, no. 10, pp. 4135-4149, Oct. 2019.
- Furqan Shaukat, K. Javed, Gulistan Raja, Junaid Mir, and Muhammad Laiq-Ur-Rahman Shahid, “Automatic Lung Nodule Detection in CT Images using Convolutional Neural Networks,” IEICE Transactions on Fundamentals on Electronics, Communications and Computer Sciences, vol. E102-A, no.10, pp.1364-1373, Oct. 2019.
- Furqan Shaukat, Gulistan Raja and A. Frangi, “Computer-Aided Detection of Lung Nodules: A Review,” SPIE Journal of Medical Imaging, vol. 6, no. 2, Apr-Jun. 2019.
- M.S. Ahmad, Junaid Mir, M.O. Ullah, Muhammad Laiq-Ur-Rahman Shahid and S.M. Adnan, “An Efficient Heart Murmur Recognition and Cardiovascular Disorders Classification System,” Australasian Physical and Engineering Sciences in Medicine, vol. 42, no. 3, pp. 733-743, 2019.
- R. Chughtai, Gulistan Raja, Junaid Mir and Furqan Shaukat, “An Efficient Scheme for Automatic Pill Recognition Using Neural Networks,” The Nucleus Journal, vol. 56, no. 1, Jun. 2019, pp. 42-48.
- Junaid Mir, D.S. Talagala, A. Fernando and S.S. Hussain, “Improved HEVC λ -domain rate control algorithm for HDR video,” Signal, Image, and Video Processing, vol. 13, no. 3, pp. 439-445, 2019.
- Junaid Mir, D.S. Talagala, A. Fernando and H.K. Arachchi, “A Comprehensive Study and Evaluation of HDR Video Coding,” Arabian Journal of Science and Engineering, vol. 44, no. 3, pp. 2427-2444, 2019.
2018
- N. U. Ain, Furqan Shaukat, A.S. Nagra and Gulistan Raja, “An Efficient Algorithm for Fingerprint Recognition using Minutiae Extraction,” Pakistan Journal of Science, vol. 70, no.2, Jun. 2018, pp. 169-176.
- Furqan Shaukat and Gulistan Raja, “An Efficient Algorithmic Solution for Automatic Segmentation of Lungs from CT Images,” Pakistan Journal of Science, vol. 70, no. 1, Mar. 2018, pp. 71-78.
2017
- Furqan Shaukat, Gulistan Raja, A. Gooya and A. Frangi, “Fully Automatic and Accurate Detection of Lung Nodules in CT images using a Hybrid Feature Set,” Medical Physics Journal, vol. 44, no. 7, July 2017, pp. 3615-3629.
- I. Amjad, Furqan Shaukat, Gulistan Raja and A.K. Khan, “An Algorithm to Segment the Mid-Brain Structures Using Multiresolution Non-Rigid Registration,” Pakistan Journal of Science, vol. 69, no. 2, June 2017, pp. 221-227.
- Muhammad Laiq-Ur-Rahman Shahid, Teodora Chitiboi, Tatyana Ivanovska, Vladimir Molchanov, Henry Völzke and Lars Linsen, “Automatic MRI Segmentation of Para-Pharyngeal Fat Pads using Interactive Visual Feature Space Analysis for Classification,” BMC Medical Imaging, vol. 17, no. 1, 2017.
2016
- Gulistan Raja, M. J. Mirza, and T. Song, “H.264/AVC Deblocking Filter based on Motion Activity in Video Sequences,” Journal of IEICE Electronics Express, Vol. 5, No. 19, 2008, pp. 809-814.
- M. Asghar, I. Arshad, I. A. Taj, Gulistan Raja and A. K. Khan, “Palm and Finger Segmentation of High Resolution Images using Hand Shape and Texture,” Proceedings of Pakistan Academy of Sciences: A, vol. 53, no. 4, Dec. 2016, pp. 401-416.
- Z. Shabbir, I. Arshad, Gulistan Raja and A. K. Khan, “Content Based Image Retrieval using Improved Local Tetra Pattern and Neural Network,” The Nucleus Journal, vol. 53, no. 4, Dec. 2016, pp. 225-232.
- Z. Bashir, Gulistan Raja and M. Obaid Ullah, “A Video Enhancement Algorithm for Low-Lighting Environment using FPGA Architecture,” NED University Journal of Research, Vol. XIII, No. 4, Sep. 2016, pp. 81-89.
- B. Hassan, Gulistan Raja, T. Hassan and M. U. Akram, “Structure Tensor Based Automated Detection of Macular Edema and Central Serous Retinopathy using Optical Coherence Tomography Images,” Journal of Optical Society of America A, vol. 33, no. 4, Apr. 2016, pp. 455-463.
- B. Hassan and Gulistan Raja, “Fully Automated Assessment of Macular Edema using Optical Coherence Tomography (OCT) Images,” International Conference on Intelligent Systems Engineering, ICISE 2016, Islamabad, 15-18 Jan. 2016.
2015
- Tatyana Ivanovska, R. Laqua, Muhammad Laiq-Ur-Rahman Shahid, Lars Linsen, K. Hegenscheid and Henry Völzke, “Automatic Pharynx Segmentation from MRI Data for Analysis of Sleep Related Disorders,” International Journal on Artificial Intelligence Tools, vol. 24, no. 4, 2015.
- Muhammad Laiq-Ur-Rahman Shahid, Teodora Chitiboi, Tatyana Ivanovska, Vladimir Molchanov, Henry Völzke, Horst K. Hahn and Lars Linsen; "Automatic Pharynx Segmentation from MRI Data for Obstructive Sleep Apnea Analysis", Proceedings of 10th International Conference on Computer Vision Theory and Applications, pp. 599-608, 2015, Berlin, Germany.
- A. Ahmad, I. Arshad and Gulistan Raja, “Partial Fingerprint Image Enhancement using Region Division Technique and Morphological Transform,” The Nucleus Journal, vol. 52, no. 2, June 2015, pp. 63-70.
Earlier Publications
- I. Arshad, Gulistan Raja and A. K. Khan, “Latent Fingerprints Segmentation: Feasibility of using Clustering based Automation Approach,” Arabian Journal of Science and Engineering, vol. 39, no. 11, Nov. 2014, pp. 7933-7944.
- S. U. Rehman and Gulistan Raja, “Performance Evaluation of HEVC High Efficiency Video Coding over Broadband Networks,” International Journal of Computer Science Issues, Vol. 11, Issue 4, July 2014, pp. 68-74.
- J. A. Raja, Gulistan Raja and A. K. Khan, “Selective Compression of Medical Images using Multiple Regions of Interest,” Life Sciences Journal, vol. 10, no. 9, Sep. 2013, pp. 394-397.
- Muhammad Laiq-Ur-Rahman Shahid and Gulistan Raja, “Implementation of Modified Control Point Image Registration Method,” The Nucleus Journal, vol. 50, no. 1, 2013, pp. 53-60.
- Junaid Mir and Gulistan Raja, “Quad tree Fractal Compression for Brain MRI Images,” The Nucleus Journal, vol. 49, no. 1, 2012.
- Furqan Shoukat and Gulistan Raja, “Implementation of Rule-Based Medical Image Recognition for Lung CT Images,” Proceedings of 9th International Bhurban Conference on Applied Sciences & Technology, Islamabad, 9-12 Jan 2012.
- S. K. Hussain and Gulistan Raja, “A JPEG 2000 based Hybrid Image Compression Technique for Medical Images,” The Nucleus Journal, vol. 48, no. 4, Dec. 2011, pp. 287-293.
- Gulistan Raja, M. J. Mirza, and T. Song, “H.264/AVC Deblocking Filter based on Motion Activity in Video Sequences,” Journal of IEICE Electronics Express, Vol. 5, No. 19, 2008, pp. 809-814.
- Gulistan Raja and M.J. Mirza, “Evaluation of Loop Filtering for Reduction of Blocking Effects in Real Time Low Bit Rate Video Coding,” M.U. Research Journal of Engineering and Technology, Vol. 26, No. 3, 2007, pp. 211-218.
Research Projects
-
Title: EMeRALDS: Electronic Medical Records driven Automated Lung nodule Detection and cancer risk Stratification
Summary: Lung cancer has been one of the major threats to human life for decades. With the lowest survival rate following diagnosis, it is the leading cause of cancer deaths worldwide and is a matter of concern in developed and developing countries. Early detection of lung cancer can therefore play a pivotal role in reducing patient mortality. However, detecting lung cancer at an early stage is challenging due to a lack of symptoms in most patients until cancer has advanced to an incurable stage. Current clinical practice in most communities worldwide have limited, or no access to sophisticated computer-assisted detection (CADe) systems and relies on expert radiologists and clinical oncologists for detection and diagnosis. Given the large volume of data in thoracic computed tomography (CT) images (the standard imaging modality for lung cancer screening) and the typically small size of lung nodules, this leads to large intra- and inter-rater variability and a large number of false positives and negatives. This is further compounded in developing countries where the availability of expert radiologists is limited and consequently, misdiagnosis of lung cancer is a major issue. These issues could be resolved with the design of robust, generalizable and scalable CADe systems designed using state-of-the-art machine learning (ML) techniques.
The main goal of this project is to develop a fully automated, scalable, and robust CADe system for lung nodules for mass screening in developing communities in Pakistan, with limited access to high-quality medical facilities and expert personnel. In the primary phase of the project, we will utilize and extend the current state-of-the-art ML/deep learning (DL) algorithms for lung nodule detection in low-dose CT scans. Following the successful development of the proposed system, it will be deployed in a selected hospital in Pakistan, where its findings on real patient data will be evaluated and compared with clinical experts and a clinically validated tool available on the market. The primary outcome of this project will be a robust nodule detection that can be deployed in a clinical environment in Pakistan to meet current needs.
Principal Investigator: Dr. Furqan Shaukat
Co-Principal Investigator: Prof. Dr. Gulistan Raja, Dr. Junaid Mir.
Amount: 8340150/- (eight million three hundred forty thousand one hundred fifty)
Duration: 36 Months
-
Title: Generic Classification of Scene Contents using Multispectral Imaging, 2020-2021
Problem Statement: Multispectral Images are suitable for distinguishing between different generic scenes occurring in airborne picture using a multi-spectral camera mounted on Airplane/satellite. Such a capability is useful in autonomous robot applications to help negotiating the environment as well as, e.g. applications intended to create large scale inventories of assets in the proximity of roads. Many materials appearing similar if viewed by a common RGB camera, will show discriminating properties if viewed by a camera capturing a greater number of separated wavelengths. In this project, a GUI using a set of robust algorithms will be developed for broader classification of scene contents. The outcome would be a scene segmented into generic classes e.g. vegetation, concrete, water etc.
Principal Investigator: Dr. Gulistan Raja
Amount: PKR 0.20 Million
Funding Source: RAC
-
Title: An Effective Content based Image Retrieval System for Lung Diseases, 2017-2020
Problem Statement: Existing Content based Image Retrieval Systems (CBIR) mostly perform image retrieval using three fundamental units namely low-level features, similarity measure and semantic gap reduction. The performance of the CBIR system generally depends on feature extraction and similarity measurement technique. The main target of this project is to improve the performance of CBIR system for retrieval of Common imaging signs (CISs) images by using robust combination of multiple descriptors and feature selection scheme to obtain an optimal feature set.
Principal Investigator: Dr. Gulistan Raja
Amount: PKR 0.30 Million
Funding Source: Directorate of ASR&TD UET Taxila
-
Title: Automated Detection of Diabetic Retinopathy Using Fundus Images, 2017-2020
Problem Statement: Diabetic retinopathy (DR) is a severe condition due to diabetes which causes damage to vision and even led to blindness in the patient. DR is one of the most basic reasons for vision loss in the recent-age. If not taken proper measures, this eye problem will increase in future diabetic patients due to diabetes incidence increase in humans. The problem to be addressed in this project is the automated detection between DR and normal images without the need for segmentation.
Principal Investigator: Dr. Gulistan Raja
Amount: PKR 0.30 Million
Funding Source: Directorate of ASR&TD UET Taxila
-
Title: Image Enhancement Techniques (Night-Time), 2019-2020
Problem Statement: Infrared imagery is produced by detecting the radiations emitted by objects. The temperature of objects determines the wavelength of radiation. On the other hand, visible imagery is produced by the rays of light reflected off of objects. This is why, visible images captured in low light or bad weather conditions are especially degraded. Since the amount of light has minimum impact on infrared images, they are effective during the day, at night, and in all weather conditions. This project investigates for the effective method to dehaze and enhance the outdoor infrared images.
Principal Investigator: Dr. Gulistan Raja
Amount: PKR 0.10 Million
Funding Source: RAC
-
Title: Brain Activity Analysis in Response to High Dynamic Range (HDR) Content Using Electroencephalography (EEG), 2018-2019.
Problem Statement: High Dynamic Range (HDR) imaging technologies, the next frontier in digital multimedia, is a step-forward towards the non-trivial challenging tasks of acquisition and representing the high-fidelity representation of the real-world scene on display devices. Through Wide Colour Gamut (WCG) and high contrast, HDR displays can invoke true-to-life visual sensations, which have effect on human perception and brain activity due to closer seamless and immersive real-world experience and needs to be assessed. In this project, human brain activity is analysed in response to HDR and low dynamic range content using electroencephalograph (EEG) signals.
Principal Investigator: Dr. Junaid Mir
Amount: PKR 0.47 Million
Funding Source: Higher Education Commission (HEC), Pakistan.
-
Title: Biomedical Signals Classification using Interactive Visual Analysis Tool, 2017-2018.
Problem Statement: Biomedical signals provide a myriad of information regarding the health of a patient. By examining changes from normal signals, medical experts can determine several different diseases. However, detection of patients from healthy subjects has not been performed using visual analysis tools. The visual analysis approach will help the user to visually analyse the data and get a better understanding of the data. Therefore, automatic classification of the biomedical signals is required.
Principal Investigator: Dr. Laiq-Ur-Rahman Shahid
Amount: PKR 0.48 Million
Funding Source: Higher Education Commission (HEC), Pakistan.
-
Title: Upper Airway Segmentation.
Problem Statement: Over the last few years, the scientists and researchers have put their efforts to study the gravitational effects on the vocal tract of subjects in singing phonation. The analysis method of dynamic processes of the vocal tract (including the oral cavity and pharynx), such as swallowing, singing voice and speech has been improved significantly due to Magnetic Resonance Imaging (MRI). To fully understand upper airway anatomy, we need to examine the volume of the airway and surrounding upper airway structures. To study upper airway anatomy, it is important to develop a segmentation technique to extract the anatomy from medical images. The first step towards this endeavour is to establish a reliable segmentation of the pharynx from 3D MR images. As manual segmentation is a laborious, observer-dependent, and time-consuming process, full automation of the three-dimensional analysis of the pharynx is required.
Principal Investigator: Dr. Laiq-Ur-Rahman Shahid
-
Title: An Efficient Scheme for Lung Nodule Detection, 2015.
Problem Statement: Lung cancer has been one of the major threats to human life for decades in both developed and under developed countries with the smallest rate of survival after diagnosis. The survival rate can be increased by early nodule detection. Computer Aided Detection (CAD) can be an important tool for early lung nodule detection and preventing the deaths caused by the lung cancer. Current lung nodule detection schemes show their lack of ability to detect all nodules while maintaining the same precision in terms of sensitivity and reduced number of false positives per scan. Most of the algorithms are optimized and limited to a particular set of data which limits the generalization of the results. In addition, the current schemes have not been evaluated on sufficiently large datasets to achieve more robustness. Therefore, methods evaluated having lesser number of nodules are not guaranteed to present the same performance in all circumstances. Moreover, since feature extraction is very important for the characterization of the nodules from other anatomic structures present in the lung region, the choice of optimum feature set for nodule detection via conventional feature-based approaches or convolutional neural networks is still an unresolved issue. Thus, the real challenge is to make more accurate systems in terms of sensitivity and reduced FP/scan with increased nodule diversity.
Principal Investigator: Dr. Furqan Shaukat
Amount: PKR 1.74 Million
Funding Source: Directorate of ASR&TD UET Taxila
|
|