BKBaiju Karun

Baiju Karun

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Baiju Karun has a total of 14 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Lightweight Blockchain Enabled Privacy-Preserving IoT Framework for Secure Parkinson’s Disease Fall Detection and Healthcare Monitoring

    by M. Rakhee, .M. Sudheep Elayidom, Baiju Karun

    The increased number of PD patients has increased the adoption of the Internet of Things (IoT)-based wearable devices for remote monitoring and automatic fall detection, which helps to notify caregivers and clinicians immediately. Unfortunately, the continuous transmission of patients' physiological and movement data through wireless sensor networks increases their vulnerability to risks of unauthorized access, poor authentication, and privacy loss. Most importantly, current automatic fall detection systems offer low levels of data verification and protection. To overcome this issue, the proposed study introduces the Cross-Layer Lightweight Starvation Blockchain Security scheme with the Provable Master Seeding Authentication Policy (LWSBS-PMSAP) for preserving privacy in the PD monitoring system. The data collected from the IoT wearable devices is first encrypted with the Advanced Shuffle Standard Starvation Encryption Cryptography (ASS-SEC) and packaged in the form of blockchain blocks, which are then stored on the encrypted servers; a controller node performs the verification and validation of the policy to prevent any unauthorized interference, while a zero-knowledge proof-based Provable Master Seeding Authentication (PMSA) protocol verifies devices and users. The framework was tested on the PD monitoring dataset of 100 patients having 1,500 sensor samples. LWSBS-PMSAP obtained a 98.7% throughput rate, 180 ms latency, 99.2% security rate, and 96.8% privacy preservation, and performed better than Ethereum-based, Hyperledger-based, and general lightweight blockchain frameworks. In the context of fall detection, the proposed framework attained 98.7% accuracy, 97.4% precision, 99.0% recall, and 98.2% F1-Score, performing better than CNN-based, LSTM-based, and ensemble-DNN-based models. The ablation study also proved that the contribution of each component in terms of security and privacy gain is measurable. Further, the statistical validation (mean, standard deviation, 95% confidence interval) revealed that there is low variation in performance of the proposed framework among the multiple trials. It can be concluded that LWSBS-PMSAP provides a computationally lightweight, secure, and privacy-preserving solution for real-time fall detection and remote monitoring of Parkinson's Disease, with clear potential for extension to broader IoT-based healthcare applications.

  • Open Access

    Advanced Hybrid Brain Tumor Segmentation in MRI: Elephant Herding Optimization Combined with Entropy-Guided Fuzzy Clustering

    by Baiju Karun, Arunprasath Thiyagarajan, Pallikonda Rajasekaran Murugan, Natarajan Jeyaprakash, Kottaimalai Ramaraj, Rakhee Makreri

    Accurate and early detection of brain tumors is essential for improving clinical outcomes and guiding effective treatment planning. Traditional segmentation techniques in MRI often struggle with challenges such as noise, intensity variations, and complex tumor morphologies, which can hinder their effectiveness in critical healthcare scenarios. This study proposes an innovative hybrid methodology that integrates advanced metaheuristic optimization and entropy-based fuzzy clustering to enhance segmentation precision in brain tumor detection. This method combines the nature-inspired Elephant Herding Optimization (EHO) algorithm with Entropy-Driven Fuzzy C-Means (EnFCM) clustering, offering significant improvements over conventional methods. EHO is utilized to optimize the clustering process, enhancing the algorithm’s ability to delineate tumor boundaries, while entropy-based fuzzy clustering accounts for intensity inhomogeneity and diverse tumor characteristics, promoting more consistent and reliable segmentation results. This approach was evaluated using the BraTS challenge dataset, a benchmark in the field of brain tumor segmentation. The results demonstrate marked improvements across several performance metrics, including Dice similarity, mean squared error (MSE), peak signal-to-noise ratio (PSNR), and the Tanimoto coefficient (TC), underscoring this method’s robustness and segmentation accuracy. By managing image noise and reducing computational demands, the EHO-EnFCM approach not only captures intricate tumor structures but also facilitates efficient image processing, making it suitable for real-time clinical applications. Overall, the findings reveal the potential of this hybrid approach to advance MRI-based tumor detection, offering a promising tool that enhances both accuracy and computational efficiency for medical imaging and diagnosis.


  • Glioma detection using EHO based FLAME clustering in MR brain images

    by Baiju Karun, T. Arun Prasath, M. Pallikonda Rajasekaran, Rakhee Makreri

    MRI is a popular imaging method for examining brain tumours. The ability to precisely segment tumours from MRI is absolutely essential for medical diagnostics and surgical planning. Manual tumour segmentation might be unrealistic for more comprehensive studies. Deep learning is the most widely used technique in medical diagnosis. For effective tumour dissection from brain MRI, this paper proposed a novel combination of FLAME and EHO Algorithm. FLAME is a type of clustering method that groups the most similar pixels in to a single cluster. EHO algorithm is one of the nature‐inspired metaheuristic optimization algorithms based on the social herding behaviour of elephants and swimming search methods. The proposed methodology's efficiency is validated through testing on various BraTS challenge datasets. The average computational time, mean squared error, peak signal to noise ratio, tanimoto coefficient, and dice score ‐ obtained are 23.3775 s, 0.213, 54.9669 dB, 54.6148%, and 84.053%, respectively.


  • Application of fuzzy logic and machine learning techniques to improve inherently safer design in process safety management: A brief study

    by Baiju Karun, Renjith V. R., Sudheep Elayidom

    Over the last decade, one of the most significant areas under focus in process safety management was developing an inherently safer design. The main objective of having an inherently safer design is to avoid hazards and risks from developing in the first place, rather than to reduce them after they have already occurred. A number of strategies, including index‐based and other types, are used in today's process industries. This paper provides a brief overview of the current inherent design methods used in the process industries. This study also details how new technologies such as fuzzy logic and machine learning are used in the improvement of inherently safer designs. Traditional safety evaluation methods have flaws such as poor accuracy, large human element influence, which can affect the degree of safety. Inherently safer design prediction was modeled using various machine learning techniques like random forest, support vector machine (SVM), and K‐neighborhood algorithm. Accuracy obtained for the sample prediction of upper flammability limit while using random forest algorithm was found to be more efficient while comparing with K‐neighborhood and support vector machine algorithms. Accuracy obtained was in the range of 90%–95% for each epoch. The accuracy of the model will always be dependent on the type of parameters that we select for prediction. By considering more safety parameters and efficient machine learning algorithms for training models, we can develop systems with high accuracy predictions for inherently safer process plants.


  • Open Access

    Enhancing the Precision of Walsh Wavelet Based Approach for Color and Texture Feature Extraction in CBIR by Including a Shape Feature

    by M. Rakhee, V. K. Govindan, Baiju Karun

    This paper proposes a method for enhancing the performance of Content Based Image Retrieval, employing a shape feature along with color and texture of Walsh wavelet. The color and texture features of the images are extracted using Walsh Wavelet and the shape feature is extracted by edge detection using any of Roberts, Sobel, Prewitt or Canny Operator. The performance of the approach is tested based on the precision values on a database containing 44 images. The results show that the precision of retrieval is increased when a shape feature is employed in the second stage of a two-stage retrieval process. Adding the shape as a third feature in a single stage retrieval process does not provide any improvement in retrieval performance with respect to precision and recall. Performance comparison was also carried out with other existing approaches, namely Walshlet and Walsh transform. The experimental results show that Walsh Wavelet has higher precision than Walshlet and Walsh transform. Also, shape extraction with Sobel and Prewitt operators provides better performance when compared to Canny and Roberts.