arXiv subjects

Manjunatha Mahadevappa

Publications and source records attributed to Manjunatha Mahadevappa.

7 recordsLinked to original sources

arXiv2026

An Attention-Based Framework for Alzheimers Disease Classification Using Resting-State fMRI

Accurate identification of Alzheimers disease (AD) using resting-state functional magnetic resonance imaging (rs-fMRI) remains challenging due to the high dimensionality, noise, and complex inter-regional dependencies inherent in functional brain connectivity, which limit the effectiveness of traditional approaches based on handcrafted connectivity features or conventional machine learning models. In this work, we present an attention-based deep learning framework for Alzheimers disease classification that operates directly on rs-fMRI functional connectivity matrices by treating brain regions as tokens and employing a Transformer-inspired self-attention mechanism to model long-range and global functional dependencies across distributed brain networks. The proposed framework learns discriminative functional representations without reliance on manual feature engineering and is evaluated on a longitudinal cohort from the Alzheimers Disease Neuroimaging Initiative (ADNI) comprising cognitively normal and Alzheimers disease subjects with multiple visits. A subject-wise evaluation protocol is adopted to prevent information leakage across visits, and class-weighted optimization is incorporated to address mild class imbalance. Experimental results for binary AD versus cognitively normal classification demonstrate that the proposed attention- based rs-fMRI model achieves an accuracy of 88.95% and a ROC-AUC of 0.90, along with a favorable precision-recall balance, highlighting the effectiveness of self-attention-driven functional connectivity modeling as a robust and interpretable approach for Alzheimers disease detection using resting-state fMRI.

eess.IV↗

arXiv2021

Geometric Solution of Image Degradation by Diffraction in Lensless Sensing and Microscopy

This paper proposes a non-computational method of counteracting the effect of image degradation introduced by the diffraction phenomenon in lensless microscopy. All the optical images (whether focused by lenses or not) are diffraction patterns, which preserve the visual information upto a certain extent determined by the size of the point spread functions, like airy disks in some cases. A highly diverging beam can be exploited to reduce the spatial extent of these point spread functions relatively in the transformed projective space, which can help us in the spatial unmixing of the visual information. The principle has been experimentally validated by the lensless imaging of red blood cells of diameter ~6-9 micrometers and a photolithography mask with features in micrometer scale. The important advantages of the proposed approach of non-computational shadow microscopy are the improved depth of field and a drastic increase in the sensor to sample working distance. The imaging method can also be used as a projection technique in the multi-angle optical computed tomography (CT).

physics.optics↗

arXiv2020

Detection and localization of Coronary Arterial Lesion with the Aid of Impedance Cardiography

In recent years, coronary artery disease is escalating and is likely to assume an epidemic proportion by 2030. Currently the reliable methods for detection of coronary arterial lesions are either conventional coronary angiogram (CAG) or MDCT (Multiple Detector Computed Tomography) coronary angiogram. Conventional CAG is an invasive procedure. Conventional CAG and CT (Computed Tomography) angiogram, both require expert supervision of either an interventional cardiologist or a radiologist. In this work, we have proposed a novel design and method for non-invasive detection and localization of coronary arterial lesion using Impedance Cardiography (ICG). The ICG signal recorded by the proposed device is used to extract feature points and compute augmentation index, amplitude and other time related parameters. The extracted features are used as input to a trained artificial neural network, for detection and prediction of coronary arterial lesions. The trained network generates specialized models, to be used for diagnosis of arterial lesions. The proposed methodology detects lesion in Left main coronary artery (LMCA), Left anterior descending artery (LAD), Diagonal branch, Left circumflex artery (LCX), and Right coronary artery (RCA) with an accuracy of 92%, 82%, 76%, 76%, 84% respectively. The proposed device could be also used by a common individual for detection of arterial lesion without any expert supervision, unassisted. The proposed algorithm eliminates the need of CAG for diagnosis of coronary arterial lesions (stenosis), and provides an insight into a new method for non-invasive monitoring of cardiovascular haemodynamics, detection and localization of coronary arterial lesion.

physics.med-ph↗

arXiv2020

Lensless in-line holographic microscopy with light source of low spatio-temporal coherence

Lensless microscopy with coherent or partially coherent light sources is a well known imaging technique, commonly referred as digital in-line holographic microscopy. In the established methods, both the spatial and temporal coherence of light play a crucial role in determining the resolution of reconstructed object. We report lensless microscopy with a spatially extended white LED, a light source of low spatial and very low temporal coherence. The wave-field propagation between two parallel planes can be obtained using a convolution operation, where the convolution kernel depends on the object-sensor distance and the characteristics of the light source. For a light source of unknown characteristics, this kernel is an unknown function. In the proposed reconstruction method, we decompose an unknown convolution kernel of very large size (128 X 128) into a small unknown light-source-specific kernel (size 9 X 9) and a known light-source-independent kernel (size 128 X 128). This drastically reduces the number of parameters to be estimated at the system identification step, which has been performed here by one time imaging of the known microscopic objects. Final unknown object estimation has been performed using the upper-bound constrained deconvolution. A lateral resolution of ~1-2 micrometer has been demonstrated.

physics.optics↗

arXiv2020

EM based Framework for Single-shot Compressive Holography

Lensless in-line holography is a simple, portable, and cost-effective method of imaging especially for the biomedical microscopy applications. We propose a multiplicative gradient descent optimization based method to obtain multi-depth imaging from a single hologram acquired in this imaging system. We further extend the method to achieve phase imaging from a single hologram. Negative-log-likelihood functional with the assumption of poisson noise has been used as the cost function to be minimized. The ill-posed nature of the problem is handled by the sparse regularization and the upper-bound constraint. The gradient descent optimization requires calculation of the partial derivative of the cost function with respect to a given estimate of the object. A method of obtaining this quantity for holography in both the cases of real object and complex object has been shown. The reconstruction method has been validated using extensive simulation and experimental studies. The comparison with the previously established iterative shrinkage/thresholding algorithm based compressive holography shows that the proposed method has the following advantages: significantly faster convergence rate, better reconstructed image quality and the ability to perform phase imaging.

eess.IV↗

arXiv2020

Photonic crystal fiber for high resolution lensless in-line holographic microscopy

We propose to use high numerical aperture single mode optical fibers like photonic crystal fiber for lensless in-line holographic microscopy. Highly divergent beam helps to overcome the spatial sampling limitation of the image sensor. In this paper, a submicron lateral resolution has been demonstrated, with an imaging sensor of pixel pitch 1.12 micrometer and a photonic crystal fiber of mode field diameter 1.8 micrometer. In earlier methods of single-shot lensless imaging, submicron resolution has been obtained at very small working distance and field of view. The proposed method improves the resolution without compromising the working distance. A working distance of (but not limited to) ~1.7 mm with a field of View ~1.4 mm has been demonstrated.

physics.optics↗

arXiv2019

An Ensemble Learning Based Classification of Individual Finger Movement from EEG

Brain computer interface based assistive technology are currently promoted for motor rehabilitation of the neuromuscular ailed individuals. Recent studies indicate a high potential of utilising electroencephalography (EEG) to extract motor related intentions. Limbic movement intentions are already exhaustively studied by the researchers with high accuracy rate. But, capturing movement of fingers from EEG is still in nascent stage. In this study, we have proposed an ensemble learning based approach for EEG in distinguishing between movements of different fingers, namely, thumb, index, and middle. Six healthy subjects participated in this study. Common spatial patterns (CSP) were extracted as features to classify with the extra tree or extremely randomized tree binary classifier. The average classification accuracy of decoding a finger from rest condition was found to be $74\%$, wheres in discriminating of movement of pair of fingers average accuracy was $60\%$. Furthermore, error correcting output coding (ECOC) was added to the binary classifier to use it in multiclass classification. The proposed algorithm achieved a maximum kappa value of 0.36 among the subjects.

eess.SP↗