Santoshachandra Rao Karanam
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Santoshachandra Rao Karanam has a total of 15 co-authors. Here are the top co-authors by number of shared publications:
- Naresh Tangudu
Aditya Institute of Technology and Management
2 shared publications - Peddada Nagamani
Anurag University
2 shared publications S. Chakravarty
Centurion University of Technology and Management
2 shared publicationsY. Srinivas
GITAM University
2 shared publications- A. B. Pradeep Kumar
GITAM (Deemed to be university)
1 shared publication - Bhabani Sankar Satapathy
Department of Pharmaceutics, GITAM School of Pharmacy, GITAM Deemed to Be University, Hyderabad Campus, Telangana, India
1 shared publication
Latest Publications
- Open Access
- 21 January 2026
- Published by Sociedade Paranaense de Matemática in Boletim da Sociedade Paranaense de Matemática
- Vol. 44 (2) , 1-7
Enhancing Diabetic Retinopathy Detection with a Hybrid Model of DenseNet121 and Support Vector Machines
Diabetic Retinopathy is the most prevalent and common diabetes-linked ocular complication that remains one of the leading causes of vision impairment worldwide. The disease develops as a result of chronic hyperglycemia, which further promotes the development of retinal microvasculature, leading to eventual vision loss if not detected and addressed in a timely manner. In this case, it is essential to establish an effective and accurate diagnostic framework to facilitate the early detection and prevention of blindness. This paper outlines a hybrid framework for the classification of DR which integrates DenseNet121’s ability to extract features and the Machine learning group’s inherent classification with Support Vector Machines to diagnose DR into five phases: No DR, Mild, Moderate, Severe, and Proliferative DR. The presented model appropriately manages the most common issues associated with classification, including class imbalance, image noise, andimaging acquisition variance using machine learning pre-processing strategies and data augmentation. For the performance of the ten most popular pre-trained convolutional neural network architectures, a Stratified Whole Dataset was applied, conditioned on the source of the image: IDRiD, Kaggle, and AEH. DenseNet121 was identified as the best-performing pre-trained architecture, exhibiting a high sensitivity of 93.169 and an F1-score of 87.726, ahead of EfficientNetB0 and Xception. Taken together, the results provide strong evidence that strong pre-processing and model selection can significantly improve classification accuracy. The proposed methodology offers a valid, scalable, and patient-centered technique for self-regulating retinal image evaluation, possessing considerable potential for application in patient care. Subsequent experiments will be directed toincreasing the performance of the models with suboptimal performance and the Ensemble learning technique and fulfill the technique verification over bigger and more varied datasets for practical application.
- 21 January 2026
- Published by Sociedade Paranaense de Matemática in Boletim da Sociedade Paranaense de Matemática
- Vol. 44 (2) , 1-7
Open AccessEnhancing Diabetic Retinopathy Detection with a Hybrid Model of DenseNet121 and Support Vector Machines
Diabetic Retinopathy is the most prevalent and common diabetes-linked ocular complication that remains one of the leading causes of vision impairment worldwide. The disease develops as a result of chronic hyperglycemia, which further promotes the development of retinal microvasculature, leading to eventual vision loss if not detected and addressed in a timely manner. In this case, it is essential to establish an effective and accurate diagnostic framework to facilitate the early detection and prevention of blindness. This paper outlines a hybrid framework for the classification of DR which integrates DenseNet121’s ability to extract features and the Machine learning group’s inherent classification with Support Vector Machines to diagnose DR into five phases: No DR, Mild, Moderate, Severe, and Proliferative DR. The presented model appropriately manages the most common issues associated with classification, including class imbalance, image noise, andimaging acquisition variance using machine learning pre-processing strategies and data augmentation. For the performance of the ten most popular pre-trained convolutional neural network architectures, a Stratified Whole Dataset was applied, conditioned on the source of the image: IDRiD, Kaggle, and AEH. DenseNet121 was identified as the best-performing pre-trained architecture, exhibiting a high sensitivity of 93.169 and an F1-score of 87.726, ahead of EfficientNetB0 and Xception. Taken together, the results provide strong evidence that strong pre-processing and model selection can significantly improve classification accuracy. The proposed methodology offers a valid, scalable, and patient-centered technique for self-regulating retinal image evaluation, possessing considerable potential for application in patient care. Subsequent experiments will be directed toincreasing the performance of the models with suboptimal performance and the Ensemble learning technique and fulfill the technique verification over bigger and more varied datasets for practical application.
- Open Access
- 30 April 2025
- Published by Asian Publication Corporation in Asian Journal of Chemistry
- Vol. 37 (5) , 1006-1016
Innovative Non-Pharmacological Therapies for SARS-CoV-2 Infections: Advances, Challenges and Future Directions
The COVID-19 pandemic has necessitated diverse therapeutic approaches, including drug-free therapies to complement pharmaceutical interventions. This review examines the current landscape, challenges and future perspectives of non-drug therapies for COVID-19. Critical approaches include oxygen therapy, mechanical ventilation, prone positioning, physical therapy, nutritional support, psychological support, hydration management and alternative therapies such as acupuncture and phototherapy. Despite their potential, these therapies face significant challenges, including disease variability, resource limitations, patient-specific factors, infection control issues and the need for robust clinical evidence. Future perspectives highlight the technological advancements such as telehealth, wearable devices and artificial intelligence; improved respiratory support systems; integrated multidisciplinary care; extensive public health initiatives; stringent clinical trials; personalized medicine; global health equity; mental health integration; lifestyle modifications and preparedness for future pandemics. Addressing these challenges and leveraging advancements will enhance the effectiveness and accessibility of drug-free therapies, improving patient outcomes and bolstering healthcare resilience against future pandemics.
- 30 April 2025
- Published by Asian Publication Corporation in Asian Journal of Chemistry
- Vol. 37 (5) , 1006-1016
Open AccessInnovative Non-Pharmacological Therapies for SARS-CoV-2 Infections: Advances, Challenges and Future Directions
The COVID-19 pandemic has necessitated diverse therapeutic approaches, including drug-free therapies to complement pharmaceutical interventions. This review examines the current landscape, challenges and future perspectives of non-drug therapies for COVID-19. Critical approaches include oxygen therapy, mechanical ventilation, prone positioning, physical therapy, nutritional support, psychological support, hydration management and alternative therapies such as acupuncture and phototherapy. Despite their potential, these therapies face significant challenges, including disease variability, resource limitations, patient-specific factors, infection control issues and the need for robust clinical evidence. Future perspectives highlight the technological advancements such as telehealth, wearable devices and artificial intelligence; improved respiratory support systems; integrated multidisciplinary care; extensive public health initiatives; stringent clinical trials; personalized medicine; global health equity; mental health integration; lifestyle modifications and preparedness for future pandemics. Addressing these challenges and leveraging advancements will enhance the effectiveness and accessibility of drug-free therapies, improving patient outcomes and bolstering healthcare resilience against future pandemics.
- 20 August 2024
- Published by Springer Nature
- p. 229-244
A Systematic Review: How Computer Vision is Transforming Agriculture in Economic Growth
- 20 August 2024
- Published by Springer Nature
- p. 229-244
A Systematic Review: How Computer Vision is Transforming Agriculture in Economic Growth
- Retracted
- 10 January 2023
- Published by Taylor & Francis in International Journal of Healthcare Management
- Vol. 19 (5) , 1-12
RETRACTED ARTICLE: A statistical model approach based on the Gaussian Mixture Model for the diagnosis and classification of bone fractures
Medical imaging has significantly influenced the field of image processing in recent years. Several advancements have been made in medical imaging that is now being used, and all of them are geared at making disease diagnosis more accurate. Researchers recently explored how to diagnose, define, and classify bone fractures, and numerous approaches have been proposed. However, a uniform categorization has yet to be developed for all detected fractures. This article focuses on the methods for identifying realignments among the bone structures utilizing model-based approaches, and this methodology is highlighted here. The performance metrics are used to assess the findings, and the constructed model has shown substantial results in binary classification and multi-classification of bone fractures. The model that has been suggested is called the Finite Beta Gaussian Mixture Model (FBGMM), and its performance may be evaluated with the use of a confusion matrix. This project intends to construct an image processing system that can identify bone fractures promptly and reliably by including data from X-rays. FGBMM achieves good accuracy in both binary and multiclassification of fractures.
- 10 January 2023
- Published by Taylor & Francis in International Journal of Healthcare Management
- Vol. 19 (5) , 1-12
RetractedRETRACTED ARTICLE: A statistical model approach based on the Gaussian Mixture Model for the diagnosis and classification of bone fractures
Medical imaging has significantly influenced the field of image processing in recent years. Several advancements have been made in medical imaging that is now being used, and all of them are geared at making disease diagnosis more accurate. Researchers recently explored how to diagnose, define, and classify bone fractures, and numerous approaches have been proposed. However, a uniform categorization has yet to be developed for all detected fractures. This article focuses on the methods for identifying realignments among the bone structures utilizing model-based approaches, and this methodology is highlighted here. The performance metrics are used to assess the findings, and the constructed model has shown substantial results in binary classification and multi-classification of bone fractures. The model that has been suggested is called the Finite Beta Gaussian Mixture Model (FBGMM), and its performance may be evaluated with the use of a confusion matrix. This project intends to construct an image processing system that can identify bone fractures promptly and reliably by including data from X-rays. FGBMM achieves good accuracy in both binary and multiclassification of fractures.
- Retracted
- 20 July 2022
- Published by Taylor & Francis in International Journal of Healthcare Management
- Vol. 19 (5) , 1-12
RETRACTED ARTICLE: A systematic approach to diagnosis and categorization of bone fractures in X-Ray imagery
Researchers have proposed several ways of diagnosing, classifying and categorizing bone fractures. Nevertheless, no standardized classification has yet been established for all identified fractures. In recent times Machine learning and deep learning are becoming more popular. Deep Neural Networks (DNN) are well-known models for their image classification capabilities and capacity to tackle complex problems. To categorize (normal, comminute, oblique, spiral, greenstick, impacted, and transverse) and recognize fragmented images, the -Speeded Up Robust Features (SURF) and Scale Invariant Feature Transform (SIFT) feature extraction approaches were utilized for several algorithms: K-Nearest Neighbor (KNN), Support Vector Machine (SVM), Random Forest (RF), Inception V3 and ResNeXt101. This research aims to construct an image processing system, including information from X-ray and Computer Tomography (CT) scans, classifying bone fractures rapidly and precisely. Pre-processing, quality enhancement, and extraction techniques are utilized to process X-ray images of the shattered bone collected from the hospital. The images are then separated and classified into fractured and unfractured bones and are compared with the accuracy of other methodologies like KNN, SVM, RF, InceptionV3, and ResNeXt101. According to the findings, the bone fracture detection and classification system perform well-using ResNeXt101, with an accuracy rate of 93.75%.
- 20 July 2022
- Published by Taylor & Francis in International Journal of Healthcare Management
- Vol. 19 (5) , 1-12
RetractedRETRACTED ARTICLE: A systematic approach to diagnosis and categorization of bone fractures in X-Ray imagery
Researchers have proposed several ways of diagnosing, classifying and categorizing bone fractures. Nevertheless, no standardized classification has yet been established for all identified fractures. In recent times Machine learning and deep learning are becoming more popular. Deep Neural Networks (DNN) are well-known models for their image classification capabilities and capacity to tackle complex problems. To categorize (normal, comminute, oblique, spiral, greenstick, impacted, and transverse) and recognize fragmented images, the -Speeded Up Robust Features (SURF) and Scale Invariant Feature Transform (SIFT) feature extraction approaches were utilized for several algorithms: K-Nearest Neighbor (KNN), Support Vector Machine (SVM), Random Forest (RF), Inception V3 and ResNeXt101. This research aims to construct an image processing system, including information from X-ray and Computer Tomography (CT) scans, classifying bone fractures rapidly and precisely. Pre-processing, quality enhancement, and extraction techniques are utilized to process X-ray images of the shattered bone collected from the hospital. The images are then separated and classified into fractured and unfractured bones and are compared with the accuracy of other methodologies like KNN, SVM, RF, InceptionV3, and ResNeXt101. According to the findings, the bone fracture detection and classification system perform well-using ResNeXt101, with an accuracy rate of 93.75%.