@ksu.edu.sa
Department of Computer Science
King Saud University
Machine Learning; Biometrics; Biomedical Signal Analysis
Scopus Publications
Scholar Citations
Scholar h-index
Scholar i10-index
Nassim Ammour, Rami M. Jomaa, Md Saiful Islam, Yakoub Bazi, Haikel Alhichri, and Naif Alajlan
MDPI AG
The electrocardiogram (ECG) signal is shown to be promising as a biometric. To this end, it has been demonstrated that the analysis of ECG signals can be considered as a good solution for increasing the biometric security levels. This can be mainly due to its inherent robustness against presentation attacks. In this work, we present a deep contrastive learning-based system for ECG biometric identification. The proposed system consists of three blocks: a feature extraction backbone based on short time Fourier transform (STFT), a contrastive learning network, and a classification network. We evaluated the proposed system on the Heartprint dataset, a new ECG biometrics multi-session dataset. The experimental analysis shows promising capabilities of the proposed method. In particular, it yields an average top1 accuracy of 98.02% on a new dataset built by gathering 1539 ECG records from 199 subjects collected in multiple sessions with an average interval between sessions of 47 days.
Md Saiful Islam, Haikel Alhichri, Yakoub Bazi, Nassim Ammour, Naif Alajlan, and Rami M. Jomaa
MDPI AG
The electrocardiogram (ECG) signal produced by the human heart is an emerging biometric modality that can play an important role in the future generation’s identity recognition with the support of machine learning techniques. One of the major obstacles in the progress of this modality is the lack of public datasets with a long interval between sessions of data acquisition to verify the uniqueness and permanence of the biometric signature of the heart of a subject. To address this issue, we put forward Heartprint, a large biometric database of multisession ECG signals comprising 1539 records captured from the fingers of 199 healthy subjects. The capturing time for each record was 15 s, and recordings were made in resting and reading conditions. They were collected in multiple sessions over ten years, and the average interval between first session (S1) and third session (S3L) was 1572.2 days. The dataset also covers several demographic classes such as genders, ethnicities, and age groups. The combination of raw ECG signals and demographic information turns the Heartprint dataset, which is made publicly available online, into a valuable resource for the development and evaluation of biometric recognition algorithms.
Mona N. Alsaleem, Md Saiful Islam, Saad Al-Ahmadi, and Adel Soudani
MDPI AG
Atrial fibrillation (AF) is one of the most common cardiac arrhythmias, and it is an indication of high-risk factors for stroke, myocardial ischemia, and other malignant cardiovascular diseases. Most of the existing AF detection methods typically convert one-dimensional time-series electrocardiogram (ECG) signals into two-dimensional representations to train a deep and complex AF detection system, which results in heavy training computation and high implementation costs. In this paper, a multiscale signal encoding scheme is proposed to improve feature representation and detection performance without the need for using any transformation or handcrafted feature engineering techniques. The proposed scheme uses different kernel sizes to produce the encoded signal by using multiple streams that are passed into a one-dimensional sequence of blocks of a residual convolutional neural network (ResNet) to extract representative features from the input ECG signal. This also allows networks to grow in breadth rather than in depth, thus reducing the computing time by using the parallel processing capability of deep learning networks. We investigated the effects of the use of a different number of streams with different kernel sizes on the performance. Experiments were carried out for a performance evaluation using the publicly available PhysioNet CinC Challenge 2017 dataset. The proposed multiscale encoding scheme outperformed existing deep learning-based methods with an average F1 score of 98.54%, but with a lower network complexity.
Rami M. Jomaa, Md Saiful Islam, Hassan Mathkour, and Saad Al-Ahmadi
Elsevier BV
Md Saiful Islam
Elsevier BV
Md Saiful Islam, Md. Abdul Awal, Jinnaton Nessa Laboni, Farhana Tazmim Pinki, Shatu Karmokar, Khondoker Mirazul Mumenin, Saad Al-Ahmadi, Md. Ashfikur Rahman, Md. Shahadat Hossain, and Seyedali Mirjalili
Elsevier BV
Kumar Debjit, Md Saiful Islam, Md. Abadur Rahman, Farhana Tazmim Pinki, Rajan Dev Nath, Saad Al-Ahmadi, Md. Shahadat Hossain, Khondoker Mirazul Mumenin, and Md. Abdul Awal
MDPI AG
A healthcare monitoring system needs the support of recent technologies such as artificial intelligence (AI), machine learning (ML), and big data, especially during the COVID-19 pandemic. This global pandemic has already taken millions of lives. Both infected and uninfected people have generated big data where AI and ML can use to combat and detect COVID-19 at an early stage. Motivated by this, an improved ML framework for the early detection of this disease is proposed in this paper. The state-of-the-art Harris hawks optimization (HHO) algorithm with an improved objective function is proposed and applied to optimize the hyperparameters of the ML algorithms, namely HHO-based eXtreme gradient boosting (HHOXGB), light gradient boosting (HHOLGB), categorical boosting (HHOCAT), random forest (HHORF) and support vector classifier (HHOSVC). An ensemble technique was applied to these optimized ML models to improve the prediction performance. Our proposed method was applied to publicly available big COVID-19 data and yielded a prediction accuracy of 92.38% using the ensemble model. In contrast, HHOXGB provided the highest accuracy of 92.23% as a single optimized model. The performance of the proposed method was compared with the traditional algorithms and other ML-based methods. In both cases, our proposed method performed better. Furthermore, not only the classification improvement, but also the features are analyzed in terms of feature importance calculated by SHapely adaptive exPlanations (SHAP) values. A graphical user interface is also discussed as a potential tool for nonspecialist users such as clinical staff and nurses. The processed data, trained model, and codes related to this study are available at GitHub.
Malak Abdullah Almarshad, Md Saiful Islam, Saad Al-Ahmadi, and Ahmed S. BaHammam
MDPI AG
Recent research indicates that Photoplethysmography (PPG) signals carry more information than oxygen saturation level (SpO2) and can be utilized for affordable, fast, and noninvasive healthcare applications. All these encourage the researchers to estimate its feasibility as an alternative to many expansive, time-wasting, and invasive methods. This systematic review discusses the current literature on diagnostic features of PPG signal and their applications that might present a potential venue to be adapted into many health and fitness aspects of human life. The research methodology is based on the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines 2020. To this aim, papers from 1981 to date are reviewed and categorized in terms of the healthcare application domain. Along with consolidated research areas, recent topics that are growing in popularity are also discovered. We also highlight the potential impact of using PPG signals on an individual’s quality of life and public health. The state-of-the-art studies suggest that in the years to come PPG wearables will become pervasive in many fields of medical practices, and the main domains include cardiology, respiratory, neurology, and fitness. Main operation challenges, including performance and robustness obstacles, are identified.
Yachao Yuan, Md. Saiful Islam, Yali Yuan, Shengjin Wang, Thar Baker, and Lutz Maria Kolbe
Institute of Electrical and Electronics Engineers (IEEE)
Road damages have caused numerous fatalities, thus the study of road damage detection, especially hazardous road damage detection and warning is critical for traffic safety. Existing road damage detection systems mainly process data at cloud, which suffers from a high latency caused by long-distance. Meanwhile, supervised machine learning algorithms are usually used in these systems requiring large precisely labeled data sets to achieve a good performance. In this article, we propose EcRD: an edge-cloud-based road damage detection and warning framework, that leverages the fast-responding advantage of edge and the large storage and computation resources advantages of cloud. There are three main contributions in this article: we first propose a simple yet efficient road segmentation algorithm to enable fast and accurate road area detection. Then, a light-weighted road damage detector is developed based on gray level co-occurrence matrix features at edge for rapid hazardous road damage detection and warning. Furthermore, a multitypes road damage detection model is introduced for long-term road management at cloud, embedded with a novel image generator based on cycle-consistent adversarial networks which automatically generates images with labels to further improve road damage detection accuracy. By comparing with the state-of-the-art, we demonstrate that the proposed EcRD can accurately detect both hazardous road damages at edge and multitypes road damages at cloud. Besides, it is around 579 times faster than cloud-based approaches without affecting users’ experience and requiring very low storage and labeling cost.
Nayeem Hasan, Md Saiful Islam, Wenyu Chen, Muhammad Ashad Kabir, and Saad Al-Ahmadi
MDPI AG
This paper proposes an encryption-based image watermarking scheme using a combination of second-level discrete wavelet transform (2DWT) and discrete cosine transform (DCT) with an auto extraction feature. The 2DWT has been selected based on the analysis of the trade-off between imperceptibility of the watermark and embedding capacity at various levels of decomposition. DCT operation is applied to the selected area to gather the image coefficients into a single vector using a zig-zig operation. We have utilized the same random bit sequence as the watermark and seed for the embedding zone coefficient. The quality of the reconstructed image was measured according to bit correction rate, peak signal-to-noise ratio (PSNR), and similarity index. Experimental results demonstrated that the proposed scheme is highly robust under different types of image-processing attacks. Several image attacks, e.g., JPEG compression, filtering, noise addition, cropping, sharpening, and bit-plane removal, were examined on watermarked images, and the results of our proposed method outstripped existing methods, especially in terms of the bit correction ratio (100%), which is a measure of bit restoration. The results were also highly satisfactory in terms of the quality of the reconstructed image, which demonstrated high imperceptibility in terms of peak signal-to-noise ratio (PSNR ≥ 40 dB) and structural similarity (SSIM ≥ 0.9) under different image attacks.
Dalal A. AlDuwaile and Md Saiful Islam
MDPI AG
The electrocardiogram (ECG) signal has become a popular biometric modality due to characteristics that make it suitable for developing reliable authentication systems. However, the long segment of signal required for recognition is still one of the limitations of existing ECG biometric recognition methods and affects its acceptability as a biometric modality. This paper investigates how a short segment of an ECG signal can be effectively used for biometric recognition, using deep-learning techniques. A small convolutional neural network (CNN) is designed to achieve better generalization capability by entropy enhancement of a short segment of a heartbeat signal. Additionally, it investigates how various blind and feature-dependent segments with different lengths affect the performance of the recognition system. Experiments were carried out on two databases for performance evaluation that included single and multisession records. In addition, a comparison was made between the performance of the proposed classifier and four well-known CNN models: GoogLeNet, ResNet, MobileNet and EfficientNet. Using a time–frequency domain representation of a short segment of an ECG signal around the R-peak, the proposed model achieved an accuracy of 99.90% for PTB, 98.20% for the ECG-ID mixed-session, and 94.18% for ECG-ID multisession datasets. Using the preprinted ResNet, we obtained 97.28% accuracy for 0.5-second segments around the R-peaks for ECG-ID multisession datasets, outperforming existing methods. It was found that the time–frequency domain representation of a short segment of an ECG signal can be feasible for biometric recognition by achieving better accuracy and acceptability of this modality.
K. M. Azharul Hasan, Sajal Das Shovon, Naimul Hoque Joy, and Md. Saiful Islam
IEEE
The COVID-19 has started expanding through the world and has become a pandemic since january 2020 . With the rise of new cases daily along with mass death, nation and society are becoming fearful of it. People from all over the world are expressing their thoughts and views about this pandemic in many social media platforms. Nowadays social media is one of the most common ways to express idea or verdict on something. With the improvement of modern computing technology, machines are constantly being conducted to interpret what people express in social media like Twitter, Facebook, Instagram etc. These thoughts or views can be categorized and analysed based on sentiment. In this paper, we have analysed the sentiment of people what they express in social media by using tweets gathering from Twitter. We have categorized the sentiment of the tweets into five classes namely ’Strongly Negative’, ’Negative’, ’Neutral’, ’Positive’ and ’Strongly Positive’. Initially we use the Textblob of python for classification. This classification does not show good results and needs massive change as there lies many new terms related to COVID-19 which effects the sentiment of tweets. We have labeled automatically by creating Regular Expression rules with our new corrected word library which is created by analysing the tweets manually. We have trained a model with the updated labeled dataset with Long short-term memory(LSTM), Logistic Regression(LR), Multinomial Naive Bayes(MNB) and analysed the performance. We found that our data labelling shows better performance comparing to standard dataset.
Oishee Bintey Hoque, Mohammad Imrul Jubair, Al-Farabi Akash, and Md. Saiful Islam
Springer International Publishing
Bangladeshi Sign Language (BdSL) is a commonly used medium of communication for the hearing-impaired people in Bangladesh. A real-time BdSL interpreter with no controlled lab environment has a broad social impact and an interesting avenue of research as well. Also, it is a challenging task due to the variation in different subjects (age, gender, color, etc.), complex features, and similarities of signs and clustered backgrounds. However, the existing dataset for BdSL classification task is mainly built in a lab friendly setup which limits the application of powerful deep learning technology. In this paper, we introduce a dataset named BdSL36 which incorporates background augmentation to make the dataset versatile and contains over four million images belonging to 36 categories. Besides, we annotate about 40, 000 images with bounding boxes to utilize the potentiality of object detection algorithms. Furthermore, several intensive experiments are performed to establish the baseline performance of our BdSL36. Moreover, we employ beta testing of our classifiers at the user level to justify the possibilities of real-world application with this dataset. We believe our BdSL36 will expedite future research on practical sign letter classification. We make the datasets and all the pre-trained models available for further researcher.
Dalal Alduwaile and Md Saiful Islam
IEEE
Biometrics plays a crucial role in information security to identify and constantly validate individuals using physiological characteristics. During the last decade, Electrocardiogram (ECG) signal has emerged as a biometric modality due to its desirable characteristics for a reliable recognition system. However, the duration of the signal required for the recognition is long, and it is still one of the limitations of existing biometric recognition methods for their acceptability as a biometric modality. In this paper, a method is proposed to use the single heartbeat ECG signal for biometric recognition of a person with the help of deep machine learning technique. We investigate the use of a light and a pre-trained convolutional neural network for the classification of single heartbeat ECG signal segmented based on the R-peak and transformed used continuous wavelet transformation. Different scenarios of segmentations experimented; Fixed length, variable length, blind, and feature depending segmentations. The performance of the proposed method was tested with a landmark dataset available online. We obtained 99.94% and 99.83% recognition accuracy for a window of ECG signal for a single heartbeat outperforming existing methods.
Shahriar Hossain, Rafeed Rahman, Md.Sabbir Ahmed, and Md.Saiful Islam
IEEE
Rafsanjany Kushol, , Md. Hasanul Kabir, M. Abdullah-Al-Wadud, Md Saiful Islam, , , and
American Institute of Mathematical Sciences (AIMS)
The improper circulation of blood flow inside the retinal vessel is the primary source of most of the optical disorders including partial vision loss and blindness. Accurate blood vessel segmentation of the retinal image is utilized for biometric identification, computer-assisted laser surgical procedure, automatic screening, and diagnosis of ophthalmologic diseases like Diabetic retinopathy, Age-related macular degeneration, Hypertensive retinopathy, and so on. Proper identification of retinal blood vessels at its early stage assists medical experts to take expedient treatment procedures which could mitigate potential vision loss. This paper presents an efficient retinal blood vessel segmentation approach where a 4-D feature vector is constructed by the outcome of Bendlet transform, which can capture directional information much more efficiently than the traditional wavelets. Afterward, a bunch of ensemble classifiers is applied to find out the best possible result of whether a pixel falls inside a vessel or non-vessel segment. The detailed and comprehensive experiments operated on two benchmark and publicly available retinal color image databases (DRIVE and STARE) prove the effectiveness of the proposed approach where the average accuracy for vessel segmentation accomplished approximately 95%. Furthermore, in comparison with other promising works on the aforementioned databases demonstrates the enhanced performance and robustness of the proposed method.
Mona Alsaleem and Md Saiful Islam
IEEE
Atrial fibrillation (AF) is the most common heart disorder manifested as an abnormal rhythm of irregular heartbeats that could lead to strokes and death. In this paper, we propose a double-layer bi-directional long short term memory (LSTM) neural network to classify a short segment of ECG signal transformed into spectrogram. We also use a preprocessing step to augment the dataset to achieve better classification performance. We conducted different experiments on different segment lengths and different network parameters using PhysioNet Challenge 2017 dataset and we achieved a total accuracy of 91.4% of classifying AF signals outperforming existing methods.
Rami M. Jomaa, Hassan Mathkour, Yakoub Bazi, and Md Saiful Islam
MDPI AG
Although fingerprint-based systems are the commonly used biometric systems, they suffer from a critical vulnerability to a presentation attack (PA). Therefore, several approaches based on a fingerprint biometrics have been developed to increase the robustness against a PA. We propose an alternative approach based on the combination of fingerprint and electrocardiogram (ECG) signals. An ECG signal has advantageous characteristics that prevent the replication. Combining a fingerprint with an ECG signal is a potentially interesting solution to reduce the impact of PAs in biometric systems. We also propose a novel end-to-end deep learning-based fusion neural architecture between a fingerprint and an ECG signal to improve PA detection in fingerprint biometrics. Our model uses state-of-the-art EfficientNets for generating a fingerprint feature representation. For the ECG, we investigate three different architectures based on fully-connected layers (FC), a 1D-convolutional neural network (1D-CNN), and a 2D-convolutional neural network (2D-CNN). The 2D-CNN converts the ECG signals into an image and uses inverted Mobilenet-v2 layers for feature generation. We evaluated the method on a multimodal dataset, that is, a customized fusion of the LivDet 2015 fingerprint dataset and ECG data from real subjects. Experimental results reveal that this architecture yields a better average classification accuracy compared to a single fingerprint modality.
Rakib Ul Haque, M. F. Mridha, Md. Abdul Hamid, M. Abdullah-Al-Wadud, and Md. Saiful Islam
Springer Science and Business Media LLC
Though plenty of research works have been done on stop word/phrase detection, there is no work done on Bengali stop words and stop phrases. This research innovates the definition and classification of Bengali stop words and phrases and implements two approaches to identify them. First one is a corpus-based approach, while the second one is based on the finite-state automaton. Performance of both approaches is measured and compared. Result analysis shows that corpus-based method outperforms the finite-state automaton-based method. The corpus-based and finite-state automaton-based method shows 90% and 80% of accuracy, respectively, for stop word detection and 80% and 70% accuracy, respectively, for stop phrase detection.
Sarah Alharbi, Md Saiful Islam, and Saad Alahmadi
Springer International Publishing
Cardiac signal (also known as ECG signal) attracted researchers for using it in generating cryptographic keys due to its availability and its intrinsic nature of every individual. However, the intra-individual variance of ECG signal decreases the possibility of getting a time-invariant key for each individual and increases decryption errors in case of using it in symmetric cryptography. In this paper, we propose a time-invariant cryptographic key generation approach (TICK) that uses a novel method for reducing the intra-individual variance in the real-valued ECG features of multiple sessions. Also, it uses a quantization method for converting the improved ECG features to binary sequences with high randomness. We have tested the approach on a multi-session database. Experimental results show its viability to improve the reliability of keys up to 96.80% using across-sessions data and up to 98.69% using within-session data. We verified the randomness using five of U.S. National Institute of Standards and Technology statistical tests and the generated keys passed all tests. Also, we verified the randomness using min-entropy, and the generated keys offer entropy of ~1.
Subbroto Saha, S.M. Islam, M. Abdullah-AL-Wadud, Saiful Islam, Farman Ali, and Kyoung Park
MDPI AG
Kidney-type glutaminase (GLS) and liver-type glutaminase (GLS2) are dysregulated in many cancers, making them appealing targets for cancer therapy. However, their use as prognostic biomarkers is controversial and remains an active area of cancer research. Here, we performed a systematic multiomic analysis to determine whether glutaminases function as prognostic biomarkers in human cancers. Glutaminase expression and methylation status were assessed and their prominent functional protein partners and correlated genes were identified using various web-based bioinformatics tools. The cross-cancer relationship of glutaminases with mutations and copy number alterations was also investigated. Gene ontology (GO) and pathway analysis were performed to assess the integrated effect of glutaminases and their correlated genes on various cancers. Subsequently, the prognostic roles of GLS and GLS2 in human cancers were mined using univariate and multivariate survival analyses. GLS was frequently over-expressed in breast, esophagus, head-and-neck, and blood cancers, and was associated with a poor prognosis, whereas GLS2 overexpression implied poor overall survival in colon, blood, ovarian, and thymoma cancers. Both GLS and GLS2 play oncogenic and anti-oncogenic roles depending on the type of cancer. The varying prognostic characteristics of glutaminases suggest that GLS and GLS2 expression differentially modulate the clinical outcomes of cancers.
Md Saiful Islam, Mohamed Maher Ben Ismail, Ouiem Bchir, Mohammed Zakariah, and Yousef Ajami Alotaibi
Institute of Electrical and Electronics Engineers (IEEE)
Atrial fibrillation (AF) is the most common cardiac arrhythmia. It increases the risk of stroke, dementia, and death; therefore, its timely diagnosis at an initial stage is crucial. Often wearable mobile devices are recommended for the primary detection of this life-threatening arrhythmia. Irregularity of the heartbeat duration, often measured through R-R intervals (RRI), has been intensively investigated during the past four decades for automatic detection of AF. However, little improvement has been made when the input signal (RRI tachogram) contains different types of arrhythmic rhythms. In this paper, we propose a neighborhood component analysis (NCA) based linear transformation of a window of RRI tachogram to improve the robustness of AF detection. Several state-of-the-art classification models are trained and tested using transformed signals, and AF detection performance are evaluated using the challenging MIT-BIH Arrhythmia Database containing various types of arrhythmic rhythms. The experimental results show significant improvement in AF detection performance using the transformed signals compared to those for signals in the original space and after linear-discriminant-analysis-based transformation. In particular, for the Naïve Bayesian classification of the transformed signals, we obtained 98.59% sensitivity, 99.91% specificity, 99.16% positive predictive value, and 99.79% accuracy. The proposed AF detection method outperforms the existing methods reported in the past four decades. Owing to the use of a short window of RRI tachogram (15 consecutive RRIs), the proposed method can be incorporated into a deployable mobile screening device for robust detection of AF.
Rami M. Jomaa, Md Saiful Islam, and Hassan Mathkour
IEEE
Biometrics is one of the most encouraging authentication systems in the recent years. However, spoof attack is one of the main problems with a biometric system. Spoof attack falls within a subset of what is called presentation attack. The heart is an emerging biometric modality which is getting attention for its robustness against presentation attacks. Introducing heart-signal into a fingerprint biometric system can yield promising results showing its robustness against spoof attacks with increasing the authentication accuracy. In this work, a sequential fusion method is improved for anti-spoofing capability. The idea behind the proposed system is the utilization of the natural liveness property of heart-biometrics in addition to boosting the heart-signal scores to increase the anti-spoofing of a multimodal biometric system. We have evaluated our proposed method with public databases of fingerprint biometric and heart-signal (ECG signal). The obtained results are very encouraging for the development of a robust anti-spoofing multimodal authentication system.
Uzzal Kumar Dutta, Md Abdur Razzaque, M. Abdullah Al-Wadud, Md Saiful Islam, M. Shamim Hossain, and B. B. Gupta
Institute of Electrical and Electronics Engineers (IEEE)
In this paper, we design self-adaptive scheduling (SAS) algorithms for base transceiver stations (BTSs) of 5G networks to improve energy efficiency, reduce carbon footprint, and develop a self-sustainable green cellular network. In the SAS algorithm, a BTS switches among its operating states (active, turned-off, and sleep), thereby exploiting the traffic loads of the BTS and the single-hop neighbor BTSs thereof. The dynamic settings of traffic thresholds help the SAS system in achieving a high degree of cooperation among the neighborhood BTSs, which in turn increases the energy savings of the network. Each active SAS BTS independently and dynamically decides in determining its operation state, thus make our proposed SAS algorithms fully distributed. Results from a simulation conducted in network simulator version 3 show that BTS scheduling significantly influences cellular networks, and the proposed SAS algorithm can significantly increase the energy savings compared with state-of-the-art protocols.
Saba Noor, Mansoor Ahmed, Malik Najmus Saqib, M. Abdullah-Al-Wadud, Md Saiful Islam, and Fazal-e-Amin
American Scientific Publishers