Kumar Rajamani
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
Research Subjects
Kumar Rajamani has a total of 94 co-authors. Here are the top co-authors by number of shared publications:
Seemant Chaturvedi
Wayne State Univ, Sch Med, Detroit, MI USA
4 shared publicationsHanna Siebert
Institute of Medical Informatics, University of Lübeck, Luebeck, Germany
2 shared publications- Hanna Siebert
Universität zu Lübeck
2 shared publications - Mattias Paul Heinrich
Institute of Medical Informatics, University of Lübeck, Lübeck, Germany
2 shared publications Agostino Maselli
1 shared publication- Aiswariya Milan Kummaya
Amrita School of Computing, Amrita Vishwa Vidyapeetham
1 shared publication
Latest Publications
- Open Access
- 27 February 2025
- Published by MDPI AG in Applied System Innovation
- Vol. 8 (2) , 28
Fed-Hetero: A Self-Evaluating Federated Learning Framework for Data Heterogeneity
Federated learning (FL) enables deep learning models to be trained locally on devices without the need for data sharing, ensuring data privacy. However, when clients have uneven or imbalanced data distributions, it leads to data heterogeneity. Data heterogeneity can appear in different ways, often due to variations in label, data distributions, feature variations, and structural inconsistencies in the images. This can significantly impact FL performance, as the global model often struggles to achieve optimal convergence. To enhance training efficiency and model performance, a common strategy in FL is to exclude clients with limited data. However, excluding such clients can raise fairness concerns, particularly for smaller populations. To understand the influence of data heterogeneity, a self-evaluating federated learning framework for heterogeneity, Fed-Hetero, was designed to assess the type of heterogeneity associated with the clients and provide recommendations to clients to enhance the global model’s accuracy. Fed-Hetero thus enables the clients with limited data to participate in FL processes by adopting appropriate strategies that enhance model accuracy. The results show that Fed-Hetero identifies the client with heterogeneity and provides personalized recommendations.
- 27 February 2025
- Published by MDPI AG in Applied System Innovation
- Vol. 8 (2) , 28
Open AccessFed-Hetero: A Self-Evaluating Federated Learning Framework for Data Heterogeneity
Federated learning (FL) enables deep learning models to be trained locally on devices without the need for data sharing, ensuring data privacy. However, when clients have uneven or imbalanced data distributions, it leads to data heterogeneity. Data heterogeneity can appear in different ways, often due to variations in label, data distributions, feature variations, and structural inconsistencies in the images. This can significantly impact FL performance, as the global model often struggles to achieve optimal convergence. To enhance training efficiency and model performance, a common strategy in FL is to exclude clients with limited data. However, excluding such clients can raise fairness concerns, particularly for smaller populations. To understand the influence of data heterogeneity, a self-evaluating federated learning framework for heterogeneity, Fed-Hetero, was designed to assess the type of heterogeneity associated with the clients and provide recommendations to clients to enhance the global model’s accuracy. Fed-Hetero thus enables the clients with limited data to participate in FL processes by adopting appropriate strategies that enhance model accuracy. The results show that Fed-Hetero identifies the client with heterogeneity and provides personalized recommendations.
- Open Access
- 1 September 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 11, 95334-95345
Toward Detecting and Addressing Corner Cases in Deep Learning Based Medical Image Segmentation
Translating machine learning research into clinical practice has several challenges. In this paper, we identify some critical issues in translating research to clinical practice in the context of medical image segmentation and propose strategies to systematically address these challenges. Specifically, we focus on cases where the model yields erroneous segmentation, which we define as corner cases. One of the standard metrics used for reporting the performance of medical image segmentation algorithms is the average Dice score across all patients. We have discovered that this aggregate reporting has the inherent drawback that the corner cases where the algorithm or model has erroneous performance or very low metrics go unnoticed. Due to this reporting, models that report superior performance could end up producing completely erroneous results, or even anatomically impossible results in a few challenging cases, albeit without being noticed. We have demonstrated how corner cases go unnoticed using the Magnetic Resonance (MR) cardiac image segmentation task of the Automated Cardiac Diagnosis Challenge (ACDC) challenge. To counter this drawback, we propose a framework that helps to identify and report corner cases. Further, we propose a novel balanced checkpointing scheme capable of finding a solution that has superior performance even on these corner cases. Our proposed scheme leads to an improvement of 44.6 % for LV, 46.1 % for RV and 38.1 % for the Myocardium on our identified corner case in the ACDC segmentation challenge. Further, we establish the generalizability of our proposed framework by also demonstrating it’s applicability in the context of chest X-ray lung segmentation. This framework has broader applications across multiple deep learning tasks even beyond medical image segmentation.
- 1 September 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 11, 95334-95345
Open AccessToward Detecting and Addressing Corner Cases in Deep Learning Based Medical Image Segmentation
Translating machine learning research into clinical practice has several challenges. In this paper, we identify some critical issues in translating research to clinical practice in the context of medical image segmentation and propose strategies to systematically address these challenges. Specifically, we focus on cases where the model yields erroneous segmentation, which we define as corner cases. One of the standard metrics used for reporting the performance of medical image segmentation algorithms is the average Dice score across all patients. We have discovered that this aggregate reporting has the inherent drawback that the corner cases where the algorithm or model has erroneous performance or very low metrics go unnoticed. Due to this reporting, models that report superior performance could end up producing completely erroneous results, or even anatomically impossible results in a few challenging cases, albeit without being noticed. We have demonstrated how corner cases go unnoticed using the Magnetic Resonance (MR) cardiac image segmentation task of the Automated Cardiac Diagnosis Challenge (ACDC) challenge. To counter this drawback, we propose a framework that helps to identify and report corner cases. Further, we propose a novel balanced checkpointing scheme capable of finding a solution that has superior performance even on these corner cases. Our proposed scheme leads to an improvement of 44.6 % for LV, 46.1 % for RV and 38.1 % for the Myocardium on our identified corner case in the ACDC segmentation challenge. Further, we establish the generalizability of our proposed framework by also demonstrating it’s applicability in the context of chest X-ray lung segmentation. This framework has broader applications across multiple deep learning tasks even beyond medical image segmentation.
- 25 July 2022
- Published by Springer Nature in Signal, Image and Video Processing
- Vol. 17 (4) , 981-989
Attention-augmented U-Net (AA-U-Net) for semantic segmentation
Deep learning-based image segmentation models rely strongly on capturing sufficient spatial context without requiring complex models that are hard to train with limited labeled data. For COVID-19 infection segmentation on CT images, training data are currently scarce. Attention models, in particular the most recent self-attention methods, have shown to help gather contextual information within deep networks and benefit semantic segmentation tasks. The recent attention-augmented convolution model aims to capture long range interactions by concatenating self-attention and convolution feature maps. This work proposes a novel attention-augmented convolution U-Net (AA-U-Net) that enables a more accurate spatial aggregation of contextual information by integrating attention-augmented convolution in the bottleneck of an encoder–decoder segmentation architecture. A deep segmentation network (U-Net) with this attention mechanism significantly improves the performance of semantic segmentation tasks on challenging COVID-19 lesion segmentation. The validation experiments show that the performance gain of the attention-augmented U-Net comes from their ability to capture dynamic and precise (wider) attention context. The AA-U-Net achieves Dice scores of 72.3% and 61.4% for ground-glass opacity and consolidation lesions for COVID-19 segmentation and improves the accuracy by 4.2% points against a baseline U-Net and 3.09% points compared to a baseline U-Net with matched parameters.
- 25 July 2022
- Published by Springer Nature in Signal, Image and Video Processing
- Vol. 17 (4) , 981-989
Attention-augmented U-Net (AA-U-Net) for semantic segmentation
Deep learning-based image segmentation models rely strongly on capturing sufficient spatial context without requiring complex models that are hard to train with limited labeled data. For COVID-19 infection segmentation on CT images, training data are currently scarce. Attention models, in particular the most recent self-attention methods, have shown to help gather contextual information within deep networks and benefit semantic segmentation tasks. The recent attention-augmented convolution model aims to capture long range interactions by concatenating self-attention and convolution feature maps. This work proposes a novel attention-augmented convolution U-Net (AA-U-Net) that enables a more accurate spatial aggregation of contextual information by integrating attention-augmented convolution in the bottleneck of an encoder–decoder segmentation architecture. A deep segmentation network (U-Net) with this attention mechanism significantly improves the performance of semantic segmentation tasks on challenging COVID-19 lesion segmentation. The validation experiments show that the performance gain of the attention-augmented U-Net comes from their ability to capture dynamic and precise (wider) attention context. The AA-U-Net achieves Dice scores of 72.3% and 61.4% for ground-glass opacity and consolidation lesions for COVID-19 segmentation and improves the accuracy by 4.2% points against a baseline U-Net and 3.09% points compared to a baseline U-Net with matched parameters.
- 1 September 2021
- Published by Elsevier in Journal of Stroke and Cerebrovascular Diseases
- Vol. 30 (9) , 105541
The Cerebro-Renal System- Anatomical and Physiological Considerations
- 1 September 2021
- Published by Elsevier in Journal of Stroke and Cerebrovascular Diseases
- Vol. 30 (9) , 105541
The Cerebro-Renal System- Anatomical and Physiological Considerations
- 20 May 2021
- Published by Elsevier in Journal of Biomedical Informatics
- Vol. 119, 103816
Dynamic deformable attention network (DDANet) for COVID-19 lesions semantic segmentation
- 20 May 2021
- Published by Elsevier in Journal of Biomedical Informatics
- Vol. 119, 103816
Dynamic deformable attention network (DDANet) for COVID-19 lesions semantic segmentation
- Open Access
Dynamic deformable attention (DDANet) for semantic segmentation
Deep learning based medical image segmentation is an important step within diagnosis, which relies strongly on capturing sufficient spatial context without requiring too complex models that are hard to train with limited labelled data. Training data is in particular scarce for segmenting infection regions of CT images of COVID-19 patients. Attention models help gather contextual information within deep networks and benefit semantic segmentation tasks. The recent criss-cross-attention module aims to approximate global self-attention while remaining memory and time efficient by separating horizontal and vertical selfsimilarity computations. However, capturing attention from all non-local locations can adversely impact the accuracy of semantic segmentation networks. We propose a new Dynamic Deformable Attention Network (DDANet) that enables a more accurate contextual information computation in a similarly efficient way. Our novel technique is based on a deformable criss-cross attention block that learns both attention coefficients and attention offsets in a continuous way. A deep segmentation network (in our case a U-Net [1]) that employs this attention mechanism is able to capture attention from pertinent non-local locations and also improves the performance on semantic segmentation tasks compared to criss-cross attention within a U-Net on a challenging COVID-19 lesion segmentation task. Our validation experiments show that the performance gain of the recursively applied dynamic deformable attention blocks comes from their ability to capture dynamic and precise (wider) attention context. Our DDANet achieves Dice scores of 73.4% and 61.3% for Ground-Glass-Opacity and Consolidation lesions for COVID-19 segmentation and improves the accuracy by 4.9% points compared to a baseline U-Net.
Open AccessDynamic deformable attention (DDANet) for semantic segmentation
Deep learning based medical image segmentation is an important step within diagnosis, which relies strongly on capturing sufficient spatial context without requiring too complex models that are hard to train with limited labelled data. Training data is in particular scarce for segmenting infection regions of CT images of COVID-19 patients. Attention models help gather contextual information within deep networks and benefit semantic segmentation tasks. The recent criss-cross-attention module aims to approximate global self-attention while remaining memory and time efficient by separating horizontal and vertical selfsimilarity computations. However, capturing attention from all non-local locations can adversely impact the accuracy of semantic segmentation networks. We propose a new Dynamic Deformable Attention Network (DDANet) that enables a more accurate contextual information computation in a similarly efficient way. Our novel technique is based on a deformable criss-cross attention block that learns both attention coefficients and attention offsets in a continuous way. A deep segmentation network (in our case a U-Net [1]) that employs this attention mechanism is able to capture attention from pertinent non-local locations and also improves the performance on semantic segmentation tasks compared to criss-cross attention within a U-Net on a challenging COVID-19 lesion segmentation task. Our validation experiments show that the performance gain of the recursively applied dynamic deformable attention blocks comes from their ability to capture dynamic and precise (wider) attention context. Our DDANet achieves Dice scores of 73.4% and 61.3% for Ground-Glass-Opacity and Consolidation lesions for COVID-19 segmentation and improves the accuracy by 4.9% points compared to a baseline U-Net.
- Open Access
- 25 August 2020
- Published by Open Engineering Inc in engrXiv
Dynamic deformable attention (DDANet) for semantic segmentation
Deep learning based medical image segmentation is an important step within diagnosis, which relies strongly on capturing sufficient spatial context without requiring too complex models that are hard to train with limited labelled data. Training data is in particular scarce for segmenting infection regions of CT images of COVID-19 patients. Attention models help gather contextual information within deep networks and benefit semantic segmentation tasks. The recent criss-cross-attention module aims to approximate global self-attention while remaining memory and time efficient by separating horizontal and vertical self-similarity computations. However, capturing attention from all non-local locations can adversely impact the accuracy of semantic segmentation networks. We propose a new Dynamic Deformable Attention Network (DDANet) that enables a more accurate contextual information computation in a similarly efficient way. Our novel technique is based on a deformable criss-cross attention block that learns both attention coefficients and attention offsets in a continuous way. A deep segmentation network (in our case a U-Net \cite{Jo2019}) that employs this attention mechanism is able to capture attention from pertinent non-local locations and also improves the performance on semantic segmentation tasks compared to criss-cross attention within a U-Net on a challenging COVID-19 lesion segmentation task. Our validation experiments show that the performance gain of the recursively applied dynamic deformable attention blocks comes from their ability to capture dynamic and precise (wider) attention context. Our DDANet achieves Dice scores of 73.4\% and 61.3\% for Ground-Glass-Opacity and Consolidation lesions for COVID-19 segmentation and improves the accuracy by 4.9\% points compared to a baseline U-Net.
- 25 August 2020
- Published by Open Engineering Inc in engrXiv
Open AccessDynamic deformable attention (DDANet) for semantic segmentation
Deep learning based medical image segmentation is an important step within diagnosis, which relies strongly on capturing sufficient spatial context without requiring too complex models that are hard to train with limited labelled data. Training data is in particular scarce for segmenting infection regions of CT images of COVID-19 patients. Attention models help gather contextual information within deep networks and benefit semantic segmentation tasks. The recent criss-cross-attention module aims to approximate global self-attention while remaining memory and time efficient by separating horizontal and vertical self-similarity computations. However, capturing attention from all non-local locations can adversely impact the accuracy of semantic segmentation networks. We propose a new Dynamic Deformable Attention Network (DDANet) that enables a more accurate contextual information computation in a similarly efficient way. Our novel technique is based on a deformable criss-cross attention block that learns both attention coefficients and attention offsets in a continuous way. A deep segmentation network (in our case a U-Net \cite{Jo2019}) that employs this attention mechanism is able to capture attention from pertinent non-local locations and also improves the performance on semantic segmentation tasks compared to criss-cross attention within a U-Net on a challenging COVID-19 lesion segmentation task. Our validation experiments show that the performance gain of the recursively applied dynamic deformable attention blocks comes from their ability to capture dynamic and precise (wider) attention context. Our DDANet achieves Dice scores of 73.4\% and 61.3\% for Ground-Glass-Opacity and Consolidation lesions for COVID-19 segmentation and improves the accuracy by 4.9\% points compared to a baseline U-Net.
- 1 September 2018
- Published by Elsevier in Neuroscience Letters
- Vol. 683, 144-149
Asymmetries in force matching are related to side of stroke in right-handed individuals
- 1 September 2018
- Published by Elsevier in Neuroscience Letters
- Vol. 683, 144-149
Asymmetries in force matching are related to side of stroke in right-handed individuals
- 18 April 2017
- Published by Wolters Kluwer Health in Neurology
- Vol. 88 (16_supplem)
Apixaban for the treatment of Cerebral Venous Thrombosis: A Report of Two Cases (P1.281)
- 18 April 2017
- Published by Wolters Kluwer Health in Neurology
- Vol. 88 (16_supplem)
Apixaban for the treatment of Cerebral Venous Thrombosis: A Report of Two Cases (P1.281)
- 18 April 2017
- Published by Wolters Kluwer Health in Neurology
- Vol. 88 (16_supplem)
Successful Thrombolysis After Protamine Reversal Of Heparin (P3.287)
- 18 April 2017
- Published by Wolters Kluwer Health in Neurology
- Vol. 88 (16_supplem)
Successful Thrombolysis After Protamine Reversal Of Heparin (P3.287)