Mandeep Singh
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Mandeep Singh has a total of 25 co-authors. Here are the top co-authors by number of shared publications:
- Assif Assad
Islamic University of Science and Technology
9 shared publications - Asifa Nazir
Department of Computer Science & Engineering Islamic University of Science & Technology Kashmir Jammu and Kashmir India
5 shared publications - Muzafar Rasool Bhat
Department of Computer Science, Islamic University of Science and Technology, Kashmir, Awantipora, India
5 shared publications - Farooq Hussain Bhat
Department of Physics, Islamic University of Science and Technology, Kashmir, Awantipora 192122, Jammu and Kashmir, India
4 shared publications - Shabir Ahmad
Department of Physics, Islamic University of Science and Technology, Awantipora, Kashmir-192122, India
4 shared publications - Ahsan Hussain
Department of Computer Science & Engineering, Islamic University of Science & Technology, Kashmir, India
3 shared publications
Latest Publications
- 22 October 2026
- Published by Wiley in Concurrency and Computation: Practice and Experience
- Vol. 38 (21)
An Explainable Attention‐Driven EfficientNetB7 Framework for Early Detection of Head and Neck Cancer Using Digital Holographic Microscopy
This study presents a deep learning (DL) framework for head and neck cancer (HNC) detection using holographic imaging data. A self‐curated dataset of holographic and corresponding bright‐field images is used for comparative evaluation, demonstrating the novelty of the approach. The framework employs an EfficientNetB7 backbone integrated with a convolutional block attention module (CBAM) to enhance feature representation via channel and spatial attention mechanisms. To address class imbalance, class weighting and focal loss are applied to improve learning for minority classes. Experimental results demonstrated strong performance, achieving 96.93% validation accuracy, 97.80% test accuracy, and 99.31% area under the curve (AUC), outperforming bright‐field imaging data. The attention‐guided model improved classification performance by up to 12.72% over baseline models. These findings demonstrate the effectiveness of the proposed framework for accurate and early detection of HNC. Furthermore, explainability is incorporated using gradient‐weighted class activation mapping (Grad‐CAM) variants and local interpretable model‐agnostic explanations (LIME), providing qualitative visual insights into the model's decision‐making process, thereby facilitating its potential use in future computer‐aided decision support.
- 22 October 2026
- Published by Wiley in Concurrency and Computation: Practice and Experience
- Vol. 38 (21)
An Explainable Attention‐Driven EfficientNetB7 Framework for Early Detection of Head and Neck Cancer Using Digital Holographic Microscopy
This study presents a deep learning (DL) framework for head and neck cancer (HNC) detection using holographic imaging data. A self‐curated dataset of holographic and corresponding bright‐field images is used for comparative evaluation, demonstrating the novelty of the approach. The framework employs an EfficientNetB7 backbone integrated with a convolutional block attention module (CBAM) to enhance feature representation via channel and spatial attention mechanisms. To address class imbalance, class weighting and focal loss are applied to improve learning for minority classes. Experimental results demonstrated strong performance, achieving 96.93% validation accuracy, 97.80% test accuracy, and 99.31% area under the curve (AUC), outperforming bright‐field imaging data. The attention‐guided model improved classification performance by up to 12.72% over baseline models. These findings demonstrate the effectiveness of the proposed framework for accurate and early detection of HNC. Furthermore, explainability is incorporated using gradient‐weighted class activation mapping (Grad‐CAM) variants and local interpretable model‐agnostic explanations (LIME), providing qualitative visual insights into the model's decision‐making process, thereby facilitating its potential use in future computer‐aided decision support.
- 1 June 2026
- Published by Elsevier in Optics Communications
- Vol. 617
High-throughput holographic sensing: A reconstruction-free framework via Vision Transformer foundation models
- 1 June 2026
- Published by Elsevier in Optics Communications
- Vol. 617
High-throughput holographic sensing: A reconstruction-free framework via Vision Transformer foundation models
- 26 March 2026
- Published by Springer Nature in Signal, Image and Video Processing
- Vol. 20 (4) , 211
Focal-AttentionNet: A novel AI-driven framework integrating holographic imaging for head and neck cancer diagnosis
- 26 March 2026
- Published by Springer Nature in Signal, Image and Video Processing
- Vol. 20 (4) , 211
Focal-AttentionNet: A novel AI-driven framework integrating holographic imaging for head and neck cancer diagnosis
- 14 January 2026
- Published by IOP Publishing in Journal of Optics
- Vol. 28 (1)
Complex-valued convolutional neural networks for disease detection utilizing digital holographic wavefronts
Phase carries crucial morphological and compositional information and is central to many optical imaging modalities. Digital holographic microscopy (DHM), interferometric technique, captures the complex object field of a specimen, containing both phase and amplitude. Conventional deep learning (DL) networks process complex data as paired real channels, neglecting intrinsic phase–amplitude coupling. In this reported work, we introduce complex-valued convolutional neural networks (CV-CNNs) for binary classification of normal and cancerous cervical cells using the DHM complex object field. Four CV-CNN variants (ModReLU, CReLU, Cardioid, zReLU) were evaluated via five-fold cross-validation against an equivalent real-valued CNN (RV-CNN). The ModReLU CV-CNN achieved the highest performance (Accuracy = 0.89 ± 0.03, F1 = 0.89 ± 0.04, Precision = 0.89 ± 0.04, Recall = 0.89 ± 0.05, AUC = 0.96 ± 0.02), demonstrating that phase-aware complex convolution significantly enhances the classification performance. By eliminating phase unwrapping, it also reduces associated computational cost and ambiguity. The methodology is also applicable to other phase-sensitive imaging modalities.
- 14 January 2026
- Published by IOP Publishing in Journal of Optics
- Vol. 28 (1)
Complex-valued convolutional neural networks for disease detection utilizing digital holographic wavefronts
Phase carries crucial morphological and compositional information and is central to many optical imaging modalities. Digital holographic microscopy (DHM), interferometric technique, captures the complex object field of a specimen, containing both phase and amplitude. Conventional deep learning (DL) networks process complex data as paired real channels, neglecting intrinsic phase–amplitude coupling. In this reported work, we introduce complex-valued convolutional neural networks (CV-CNNs) for binary classification of normal and cancerous cervical cells using the DHM complex object field. Four CV-CNN variants (ModReLU, CReLU, Cardioid, zReLU) were evaluated via five-fold cross-validation against an equivalent real-valued CNN (RV-CNN). The ModReLU CV-CNN achieved the highest performance (Accuracy = 0.89 ± 0.03, F1 = 0.89 ± 0.04, Precision = 0.89 ± 0.04, Recall = 0.89 ± 0.05, AUC = 0.96 ± 0.02), demonstrating that phase-aware complex convolution significantly enhances the classification performance. By eliminating phase unwrapping, it also reduces associated computational cost and ambiguity. The methodology is also applicable to other phase-sensitive imaging modalities.
- 13 November 2025
- Published by IOP Publishing in Modelling and Simulation in Materials Science and Engineering
- Vol. 33 (8) , 085005
Exploring feature generation and engineering techniques for machine learning prediction of ferromagnetic Curie temperatures in perovskites
Rapid estimation of Curie temperature is essential for discovering advanced functional materials like ABO 3 -type ferromagnetic perovskites used in spintronics, sensors, and multiferroics. This study applies machine learning (ML) to predict Curie temperatures directly from chemical formulae, eliminating the need for computationally intensive simulations. A manually curated dataset of 300 perovskites with known Curie temperatures is used. Features are generated by separately considering A-site and B-site cation properties and also by treating the compound as a whole. Three strategies, feature selection, recursive elimination, and principal component analysis are employed for feature engineering. Multiple models are trained and evaluated using cross-validation, with coefficient of determination R 2 as the performance metric. Shapley Additive Explanations highlights mean absolute deviation of A-site ground state volume and B-site space-group number as key factors influencing Curie temperature. The best model achieves an R 2 of 0.8 and is applied to predict Curie temperatures for 21,467 ABO 3 compositions, demonstrating the potential of ML in accelerating magnetic materials discovery.
- 13 November 2025
- Published by IOP Publishing in Modelling and Simulation in Materials Science and Engineering
- Vol. 33 (8) , 085005
Exploring feature generation and engineering techniques for machine learning prediction of ferromagnetic Curie temperatures in perovskites
Rapid estimation of Curie temperature is essential for discovering advanced functional materials like ABO 3 -type ferromagnetic perovskites used in spintronics, sensors, and multiferroics. This study applies machine learning (ML) to predict Curie temperatures directly from chemical formulae, eliminating the need for computationally intensive simulations. A manually curated dataset of 300 perovskites with known Curie temperatures is used. Features are generated by separately considering A-site and B-site cation properties and also by treating the compound as a whole. Three strategies, feature selection, recursive elimination, and principal component analysis are employed for feature engineering. Multiple models are trained and evaluated using cross-validation, with coefficient of determination R 2 as the performance metric. Shapley Additive Explanations highlights mean absolute deviation of A-site ground state volume and B-site space-group number as key factors influencing Curie temperature. The best model achieves an R 2 of 0.8 and is applied to predict Curie temperatures for 21,467 ABO 3 compositions, demonstrating the potential of ML in accelerating magnetic materials discovery.
- Open Access
- 30 September 2025
- Published by Royal Society of Chemistry (RSC) in Nanoscale Advances
- Vol. 7 (23) , 7570-7578
Boosting the electrical properties of nano-structured (Ga 10 Se 90 )-MWCNT bilayer thin films
Selenium-based chalcogenide semiconductors are of growing interest for optoelectronic applications due to their high DC conductivity, photoconductivity, and carrier mobility. In this work, Ga10Se90-MWCNT nanocomposite thin films were fabricated via thermal evaporation and subsequently modified using swift heavy ion (SHI) irradiation using 70 MeV Ni7+ ions at fluences of 1 × 1012 to 1 × 1013 ions per cm2. SHI irradiation promoted the homogeneous mixing of CNTs within the GaSe matrix, enhancing the electrical properties through defect engineering and interfacial charge transfer. Characterization via FESEM, EDX, and FTIR confirmed uniform nanoparticle dispersion and strong GaSe-CNT interactions. Post-irradiation, DC conductivity increased from 2.11 × 10−4 to 2.97 × 10−3 Ω−1 cm−1, photoconductivity from 5.16 × 10−5 to 2.90 × 10−3 Ω−1 cm−1, and carrier mobility from 50.9 to 475.1 cm2 V−1 s−1. These enhancements are attributed to improved charge transport via CNT networks and SHI-induced acceptor states. The results highlight the tunability of GaSe-CNT composites via SHI, supporting their potential use in advanced optoelectronic devices.
- 30 September 2025
- Published by Royal Society of Chemistry (RSC) in Nanoscale Advances
- Vol. 7 (23) , 7570-7578
Open AccessBoosting the electrical properties of nano-structured (Ga 10 Se 90 )-MWCNT bilayer thin films
Selenium-based chalcogenide semiconductors are of growing interest for optoelectronic applications due to their high DC conductivity, photoconductivity, and carrier mobility. In this work, Ga10Se90-MWCNT nanocomposite thin films were fabricated via thermal evaporation and subsequently modified using swift heavy ion (SHI) irradiation using 70 MeV Ni7+ ions at fluences of 1 × 1012 to 1 × 1013 ions per cm2. SHI irradiation promoted the homogeneous mixing of CNTs within the GaSe matrix, enhancing the electrical properties through defect engineering and interfacial charge transfer. Characterization via FESEM, EDX, and FTIR confirmed uniform nanoparticle dispersion and strong GaSe-CNT interactions. Post-irradiation, DC conductivity increased from 2.11 × 10−4 to 2.97 × 10−3 Ω−1 cm−1, photoconductivity from 5.16 × 10−5 to 2.90 × 10−3 Ω−1 cm−1, and carrier mobility from 50.9 to 475.1 cm2 V−1 s−1. These enhancements are attributed to improved charge transport via CNT networks and SHI-induced acceptor states. The results highlight the tunability of GaSe-CNT composites via SHI, supporting their potential use in advanced optoelectronic devices.
- 29 September 2025
- Published by Springer Nature in Journal of Optics
- p. 1-15
Integrating quantitative phase imaging with deep learning for enhanced cervical cancer detection
- 29 September 2025
- Published by Springer Nature in Journal of Optics
- p. 1-15
Integrating quantitative phase imaging with deep learning for enhanced cervical cancer detection
- 12 May 2025
- Published by Springer Nature in Signal, Image and Video Processing
- Vol. 19 (7) , 1-30
Integrating artificial intelligence and holographic imaging for advanced cervical cancer diagnosis
- 12 May 2025
- Published by Springer Nature in Signal, Image and Video Processing
- Vol. 19 (7) , 1-30
Integrating artificial intelligence and holographic imaging for advanced cervical cancer diagnosis
- Open Access
Recent Advances of Guided Mode Resonant Sensors Applied to Cancer Biomarker Detection
Guided mode resonance (GMR)-based sensors have emerged as a promising technology for the early screening of cancer, offering advantages such as sensitivity, specificity, low cost, non-invasiveness, and portability. This review article provides a comprehensive overview of the latest advancements in GMR technology and its applications in biosensing, with a specific focus on cancer. The current state of cancer diagnosis and the critical need for point-of-care (POC) devices to address these challenges are discussed in detail. Furthermore, the review systematically explores various strategies employed in GMR-based cancer detection including design principles and the integration of advanced technologies. Additionally, it aims to provide researchers valuable insights for developing GMR sensors capable of detecting cancer biomarkers outside the laboratory environment.
Open AccessRecent Advances of Guided Mode Resonant Sensors Applied to Cancer Biomarker Detection
Guided mode resonance (GMR)-based sensors have emerged as a promising technology for the early screening of cancer, offering advantages such as sensitivity, specificity, low cost, non-invasiveness, and portability. This review article provides a comprehensive overview of the latest advancements in GMR technology and its applications in biosensing, with a specific focus on cancer. The current state of cancer diagnosis and the critical need for point-of-care (POC) devices to address these challenges are discussed in detail. Furthermore, the review systematically explores various strategies employed in GMR-based cancer detection including design principles and the integration of advanced technologies. Additionally, it aims to provide researchers valuable insights for developing GMR sensors capable of detecting cancer biomarkers outside the laboratory environment.