Kailas Patil
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Kailas Patil has a total of 97 co-authors. Here are the top co-authors by number of shared publications:
- Prawit Chumchu
Kasetsart University
24 shared publications Yogesh Suryawanshi
Vishwakarma University, Pune, India
9 shared publicationsAmol Bhosle
Department of Computer Science and Engineering, School of Computing, MIT Art, Design and Technology University
7 shared publications- Shrikant Jadhav
San Jose State University
7 shared publications - Yogesh Suryawanshi
Vishwakarma University, Pune, India
7 shared publications - Alfa Nyandoro
Regent University
6 shared publications
Latest Publications
- 28 August 2026
- Published by Bentham Science Publishers Ltd. in Recent Patents on Engineering
- Vol. 20, 1-13
An Experimental Deep Learning Framework Integrating Contextual and Visual Features for Detecting Student Engagement in Virtual Classrooms
Background: Detection of student engagement in a virtual learning environment has become paramount. The current systems are mostly based on visual facial features, ignoring contextual demographic features, which can undermine generalization and fairness across different groups of learners. Objective: The study introduces a contextually sensible deep learning system that combines visual facial features with demographic embeddings (age, gender) to enhance the classification rate of engagement and achieve fairness and real-time capabilities. Methods: A training set of 16,000 labeled images was formed of 17 institutional ethics-approved volunteers aged 5-21+ years. Baseline performance was set by fine-tuning of 4 pretrained CNNs (MobileNet, Xception, DenseNet201, NasNet Large). A custom model, OnEduNet, was developed, which integrated visual and contextual representations. Performance of models was measured using macro F1-score, confidence intervals, significance testing (McNemar), fairness metrics (Demographic Parity, Equalized Odds), ablation testing, robustness testing, and Grad-CAM interpretability. Results: Results: OnEduNet achieved a statistically significant validation accuracy of 92%, outperforming the best baseline, MobileNet (89%). The model demonstrated a minor lack of demographic difference (EOD<0.05) and maintained a real-time inference speed (18ms/image). Conclusion: Combining contextual demographic features with visual representations could contribute greatly to engagement detection accuracy, fairness, interpretability, and deployment feasibility, which can be used to support scalable and patent-oriented educational AI systems.
- 28 August 2026
- Published by Bentham Science Publishers Ltd. in Recent Patents on Engineering
- Vol. 20, 1-13
An Experimental Deep Learning Framework Integrating Contextual and Visual Features for Detecting Student Engagement in Virtual Classrooms
Background: Detection of student engagement in a virtual learning environment has become paramount. The current systems are mostly based on visual facial features, ignoring contextual demographic features, which can undermine generalization and fairness across different groups of learners. Objective: The study introduces a contextually sensible deep learning system that combines visual facial features with demographic embeddings (age, gender) to enhance the classification rate of engagement and achieve fairness and real-time capabilities. Methods: A training set of 16,000 labeled images was formed of 17 institutional ethics-approved volunteers aged 5-21+ years. Baseline performance was set by fine-tuning of 4 pretrained CNNs (MobileNet, Xception, DenseNet201, NasNet Large). A custom model, OnEduNet, was developed, which integrated visual and contextual representations. Performance of models was measured using macro F1-score, confidence intervals, significance testing (McNemar), fairness metrics (Demographic Parity, Equalized Odds), ablation testing, robustness testing, and Grad-CAM interpretability. Results: Results: OnEduNet achieved a statistically significant validation accuracy of 92%, outperforming the best baseline, MobileNet (89%). The model demonstrated a minor lack of demographic difference (EOD<0.05) and maintained a real-time inference speed (18ms/image). Conclusion: Combining contextual demographic features with visual representations could contribute greatly to engagement detection accuracy, fairness, interpretability, and deployment feasibility, which can be used to support scalable and patent-oriented educational AI systems.
- 30 July 2026
- Published by Bentham Science Publishers Ltd. in Recent Patents on Engineering
- Vol. 20, 1-20
Sustainable Telecom Operations Support System (OSS) Architecture for Emerging Technologies in Healthcare, Agriculture, and Climate-Resilient Infrastructures
Objective: The main aim of this research has been to design a sustainable Telecom Operations Support System (OSS) architecture that will automate heavily, run on less power, and be compatible across different sectors. The article elaborates on the journey an average OSS framework takes when it is equipped with AI, ML, and green computing to produce smart decisions and carry out service orchestration in an eco-friendly way. Methods: Our proposal of the OSS domain layered model includes five subsystems: Service Catalog Management, Subscriber, Profile Management, Order Management, Workflow Engine, and Service Topology Management. Predictive analytics and AI-based optimization techniques were applied to evaluate sustainability and automation performance. Moreover, the proposed approach has a paradigm shift in the OSS theory patent by incorporating the sustainability-aware service templates, green SLAs, and carbon footprint analysis modules. results: The sustainable telecom OSS architecture demonstrated significant improvements in operational performance and resource optimization. Simulation-based evaluations revealed a 30–35% reduction in power utilization and a 25–28% improvement in service activation efficiency compared to baseline OSS models. The incorporation of AI/ML-driven decision loops enabled intelligent order prioritization and adaptive resource allocation, resulting in measurable environmental and operational benefits. Results: Simulation-based evaluation demonstrated a 30–35% reduction in total energy consumption and a 25–28% improvement in service activation efficiency compared to conventional OSS systems. Discussion: These findings clearly show that the need to shift towards sustainable-oriented, AIassisted OSS frameworks is important towards offering optimal telecom services that will make positive impacts in applications involving health care, agriculture, as well as climate-resilient solutions. Conclusion: Sustainable Telecom OSS Architecture is a scalable, environment-friendly, and holistic approach designed to address the challenge of digital transformation, and is also a relevant and patent- worthy basis for the development of green OSS technology.
- 30 July 2026
- Published by Bentham Science Publishers Ltd. in Recent Patents on Engineering
- Vol. 20, 1-20
Sustainable Telecom Operations Support System (OSS) Architecture for Emerging Technologies in Healthcare, Agriculture, and Climate-Resilient Infrastructures
Objective: The main aim of this research has been to design a sustainable Telecom Operations Support System (OSS) architecture that will automate heavily, run on less power, and be compatible across different sectors. The article elaborates on the journey an average OSS framework takes when it is equipped with AI, ML, and green computing to produce smart decisions and carry out service orchestration in an eco-friendly way. Methods: Our proposal of the OSS domain layered model includes five subsystems: Service Catalog Management, Subscriber, Profile Management, Order Management, Workflow Engine, and Service Topology Management. Predictive analytics and AI-based optimization techniques were applied to evaluate sustainability and automation performance. Moreover, the proposed approach has a paradigm shift in the OSS theory patent by incorporating the sustainability-aware service templates, green SLAs, and carbon footprint analysis modules. results: The sustainable telecom OSS architecture demonstrated significant improvements in operational performance and resource optimization. Simulation-based evaluations revealed a 30–35% reduction in power utilization and a 25–28% improvement in service activation efficiency compared to baseline OSS models. The incorporation of AI/ML-driven decision loops enabled intelligent order prioritization and adaptive resource allocation, resulting in measurable environmental and operational benefits. Results: Simulation-based evaluation demonstrated a 30–35% reduction in total energy consumption and a 25–28% improvement in service activation efficiency compared to conventional OSS systems. Discussion: These findings clearly show that the need to shift towards sustainable-oriented, AIassisted OSS frameworks is important towards offering optimal telecom services that will make positive impacts in applications involving health care, agriculture, as well as climate-resilient solutions. Conclusion: Sustainable Telecom OSS Architecture is a scalable, environment-friendly, and holistic approach designed to address the challenge of digital transformation, and is also a relevant and patent- worthy basis for the development of green OSS technology.
- 13 July 2026
- Published by Taylor & Francis
- p. 67-87
Quantum Support Vector Machines (QSVMs) and Their Applications
Quantum Support Vector Machines (QSVMs) blend the interesting worlds of quantum computing and machine learning, which offers fresh approaches for tackling complex classification problems in high-dimensional and non-linear data spaces. This chapter provides a comprehensive exploration of QSVMs, beginning with the fundamentals of quantum computing concepts such as qubits, superposition, and entanglement, which enable quantum algorithms to surpass classical computational limits. Next, we look at how QSVMs are mathematically expressed, highlighting the use of kernel techniques and quantum feature maps to embed classical data into exponentially huge Hilbert spaces.The implementation of QSVMs using frameworks like Qiskit, PennyLane, and TensorFlow Quantum is discussed in detail, accompanied by code examples and case studies, including binary classification on the two-moons dataset. Applications of QSVMs across domains such as healthcare, finance, cybersecurity, natural language processing, materials science, and climate sustainability are highlighted, demonstrating their broad relevance. The chapter also addresses challenges related to noise, scalability, and hybrid integration in the NISQ era, and explores future directions involving error correction, deep learning integration, and cloud-based deployment. Together, these insights illustrate QSVMs’ potential as a transformative tool in next-generation machine learning.
- 13 July 2026
- Published by Taylor & Francis
- p. 67-87
Quantum Support Vector Machines (QSVMs) and Their Applications
Quantum Support Vector Machines (QSVMs) blend the interesting worlds of quantum computing and machine learning, which offers fresh approaches for tackling complex classification problems in high-dimensional and non-linear data spaces. This chapter provides a comprehensive exploration of QSVMs, beginning with the fundamentals of quantum computing concepts such as qubits, superposition, and entanglement, which enable quantum algorithms to surpass classical computational limits. Next, we look at how QSVMs are mathematically expressed, highlighting the use of kernel techniques and quantum feature maps to embed classical data into exponentially huge Hilbert spaces.The implementation of QSVMs using frameworks like Qiskit, PennyLane, and TensorFlow Quantum is discussed in detail, accompanied by code examples and case studies, including binary classification on the two-moons dataset. Applications of QSVMs across domains such as healthcare, finance, cybersecurity, natural language processing, materials science, and climate sustainability are highlighted, demonstrating their broad relevance. The chapter also addresses challenges related to noise, scalability, and hybrid integration in the NISQ era, and explores future directions involving error correction, deep learning integration, and cloud-based deployment. Together, these insights illustrate QSVMs’ potential as a transformative tool in next-generation machine learning.
- Open Access
- 13 April 2026
- Published by Frontiers Media SA in Frontiers in Artificial Intelligence
- Vol. 9, 1787508
Artificial intelligence for early endometrial cancer diagnosis using multimodal clinical data: integrating deep learning, explainability, and data privacy
Introduction: Early diagnosis of endometrial cancer is critical for improving survival. However, most AI-based methods are developed using single-modality data, and most of them are uninterpretable and do not offer privacy protection, or they completely ignore privacy issues. We propose a multimodal AI framework that utilizes histopathology whole-slide images (WSIs) and clinical data and incorporates both explainability and privacy-aware learning. Methods: Five hundred and twenty-nine patients’ data (354 early-stage and 175 advanced-stage) with 794 WSIs and 208,000 image patches were collected for the study. By using a convolutional neural network (CNN), morphological features were extracted from WSIs, and clinical variables were encoded by a multi-layer perceptron, respectively. These two modalities of information (the learned representations) were combined to make the final category prediction. Interpretability was enabled through Grad-CAM and clinical feature attribution, and privacy-aware training was supported by secure parameter aggregation. Results: The multimodal model obtained an accuracy of 0.91 and an AUC of 0.95, thus it exceeded the performance of clinical-only (accuracy = 0.78, AUC = 0.81) and histopathology-only (accuracy = 0.85, AUC = 0.89) models with a considerable increase in sensitivity (0.89) and specificity (0.93). The performance was preserved by privacy-aware learning as well. The net clinical benefit was maximum as per the Decision Curve Analysis. Conclusion: This framework offers a solution that is accurate, interpretable, and privacy-preserving, thus it can act as a diagnostic aid for early endometrial cancer. However, since the model was developed and evaluated using only the TCGA-UCEC cohort, external multi-center validation is required to confirm generalizability across diverse clinical populations, imaging protocols, and laboratory conditions.
- 13 April 2026
- Published by Frontiers Media SA in Frontiers in Artificial Intelligence
- Vol. 9, 1787508
Open AccessArtificial intelligence for early endometrial cancer diagnosis using multimodal clinical data: integrating deep learning, explainability, and data privacy
Introduction: Early diagnosis of endometrial cancer is critical for improving survival. However, most AI-based methods are developed using single-modality data, and most of them are uninterpretable and do not offer privacy protection, or they completely ignore privacy issues. We propose a multimodal AI framework that utilizes histopathology whole-slide images (WSIs) and clinical data and incorporates both explainability and privacy-aware learning. Methods: Five hundred and twenty-nine patients’ data (354 early-stage and 175 advanced-stage) with 794 WSIs and 208,000 image patches were collected for the study. By using a convolutional neural network (CNN), morphological features were extracted from WSIs, and clinical variables were encoded by a multi-layer perceptron, respectively. These two modalities of information (the learned representations) were combined to make the final category prediction. Interpretability was enabled through Grad-CAM and clinical feature attribution, and privacy-aware training was supported by secure parameter aggregation. Results: The multimodal model obtained an accuracy of 0.91 and an AUC of 0.95, thus it exceeded the performance of clinical-only (accuracy = 0.78, AUC = 0.81) and histopathology-only (accuracy = 0.85, AUC = 0.89) models with a considerable increase in sensitivity (0.89) and specificity (0.93). The performance was preserved by privacy-aware learning as well. The net clinical benefit was maximum as per the Decision Curve Analysis. Conclusion: This framework offers a solution that is accurate, interpretable, and privacy-preserving, thus it can act as a diagnostic aid for early endometrial cancer. However, since the model was developed and evaluated using only the TCGA-UCEC cohort, external multi-center validation is required to confirm generalizability across diverse clinical populations, imaging protocols, and laboratory conditions.
- 24 March 2026
- Published by MyJove Corporation in Journal of Visualized Experiments
- No. 229
PreventativeTestPro: A Scalable Hybrid Testing Framework Utilizing Observability and Generative AI for Proactive Software Quality Engineering
This paper introduces a sophisticated, scalable testing system that integrates observability-driven automation with AI-augmented proactive quality engineering to tackle contemporary software delivery difficulties. The suggested system enhances PreventativeTestPro, an open-source, hybrid testing platform that combines black-box and white-box methodologies, by incorporating an innovative observability-based test orchestration layer. The platform utilizes logs, metrics, events, and traces alongside browser and server-side monitoring to promptly identify anomalies, enhance test case selection, and automate the creation of functional, performance, and security test suites. A distinctive characteristic is the incorporation of large language models (LLMs) to provide root cause insights and autonomously construct new test cases based on production behaviors and identified abnormalities, thus providing adaptive regression coverage and intelligent remediation. The system facilitates concurrent test execution with instantaneous AI-driven log analysis, fostering a continuous feedback loop between operations and testing. It has been validated in several enterprise scenarios, including microservices-based SaaS platforms and SAP BTP ecosystems. Empirical findings from four production deployments and a beta group of 49 engineers indicate a decrease of up to 30% in mean time to resolution, over 95% compliance with SLAs, and substantial improvements in both test coverage and defect traceability. The effortless connection with industry-standard tools illustrates its plug-and-play capability. This research presents a comprehensive, tool-independent, and forward-looking quality engineering methodology consistent with agile and DevOps principles. Future endeavors encompass dynamic anomaly classification through machine learning, extension to mobile and user experience-oriented systems, and augmented large language model capabilities for domain-specific test development and failure forecasting.
- 24 March 2026
- Published by MyJove Corporation in Journal of Visualized Experiments
- No. 229
PreventativeTestPro: A Scalable Hybrid Testing Framework Utilizing Observability and Generative AI for Proactive Software Quality Engineering
This paper introduces a sophisticated, scalable testing system that integrates observability-driven automation with AI-augmented proactive quality engineering to tackle contemporary software delivery difficulties. The suggested system enhances PreventativeTestPro, an open-source, hybrid testing platform that combines black-box and white-box methodologies, by incorporating an innovative observability-based test orchestration layer. The platform utilizes logs, metrics, events, and traces alongside browser and server-side monitoring to promptly identify anomalies, enhance test case selection, and automate the creation of functional, performance, and security test suites. A distinctive characteristic is the incorporation of large language models (LLMs) to provide root cause insights and autonomously construct new test cases based on production behaviors and identified abnormalities, thus providing adaptive regression coverage and intelligent remediation. The system facilitates concurrent test execution with instantaneous AI-driven log analysis, fostering a continuous feedback loop between operations and testing. It has been validated in several enterprise scenarios, including microservices-based SaaS platforms and SAP BTP ecosystems. Empirical findings from four production deployments and a beta group of 49 engineers indicate a decrease of up to 30% in mean time to resolution, over 95% compliance with SLAs, and substantial improvements in both test coverage and defect traceability. The effortless connection with industry-standard tools illustrates its plug-and-play capability. This research presents a comprehensive, tool-independent, and forward-looking quality engineering methodology consistent with agile and DevOps principles. Future endeavors encompass dynamic anomaly classification through machine learning, extension to mobile and user experience-oriented systems, and augmented large language model capabilities for domain-specific test development and failure forecasting.
- Open Access
- 3 March 2026
- Published by Elsevier in Software Impacts
- Vol. 28
QUALITY: Quick Unified Automation Leveraging Intelligent Test Yield
- 3 March 2026
- Published by Elsevier in Software Impacts
- Vol. 28
Open AccessQUALITY: Quick Unified Automation Leveraging Intelligent Test Yield
- 25 February 2026
- Published by MyJove Corporation
- No. 228,p. e69744
JoVE Video Dataset
Maintaining optimal water quality is essential in aquaculture, particularly for ornamental species such as Kois, where color clarity and environmental balance are directly linked to health and aesthetics. This article presents a complete protocol for designing, deploying, and operating a real-time water quality monitoring system based on underwater image classification using embedded deep learning. The proposed system integrates a low-cost underwater camera with a Raspberry Pi-based controller network in an autonomous water surface robot, enabling the capture, transmission, and classification of aquatic images through a lightweight neural network model. Five water condition categories -- ranging from clear to turbid -- are visually distinguished, with classification accuracy exceeding 99% on a custom dataset. The system is configured to run autonomously, offering continuous assessment and visual alerts for deteriorating conditions. This visual protocol covers device assembly, system calibration, model deployment, and real-time monitoring demonstrations. The method supports sustainable aquaculture practices by enabling non-invasive, continuous evaluation of aquatic environments using affordable, scalable technologies. The approach can be adapted to other aquaculture or environmental monitoring contexts with minimal modification.
- 25 February 2026
- Published by MyJove Corporation
- No. 228,p. e69744
JoVE Video Dataset
Maintaining optimal water quality is essential in aquaculture, particularly for ornamental species such as Kois, where color clarity and environmental balance are directly linked to health and aesthetics. This article presents a complete protocol for designing, deploying, and operating a real-time water quality monitoring system based on underwater image classification using embedded deep learning. The proposed system integrates a low-cost underwater camera with a Raspberry Pi-based controller network in an autonomous water surface robot, enabling the capture, transmission, and classification of aquatic images through a lightweight neural network model. Five water condition categories -- ranging from clear to turbid -- are visually distinguished, with classification accuracy exceeding 99% on a custom dataset. The system is configured to run autonomously, offering continuous assessment and visual alerts for deteriorating conditions. This visual protocol covers device assembly, system calibration, model deployment, and real-time monitoring demonstrations. The method supports sustainable aquaculture practices by enabling non-invasive, continuous evaluation of aquatic environments using affordable, scalable technologies. The approach can be adapted to other aquaculture or environmental monitoring contexts with minimal modification.
- 24 February 2026
- Published by MyJove Corporation in Journal of Visualized Experiments
- No. 228,p. e69744
Real-Time Pond Water Assessment via Embedded Deep Learning and Visual Data Acquisition: A Practical Monitoring Approach for Aquaculture
Maintaining optimal water quality is essential in aquaculture, particularly for ornamental species such as Kois, where color clarity and environmental balance are directly linked to health and aesthetics. This article presents a complete protocol for designing, deploying, and operating a real-time water quality monitoring system based on underwater image classification using embedded deep learning. The proposed system integrates a low-cost underwater camera with a Raspberry Pi-based controller network in an autonomous water surface robot, enabling the capture, transmission, and classification of aquatic images through a lightweight neural network model. Five water condition categories -- ranging from clear to turbid -- are visually distinguished, with classification accuracy exceeding 99% on a custom dataset. The system is configured to run autonomously, offering continuous assessment and visual alerts for deteriorating conditions. This visual protocol covers device assembly, system calibration, model deployment, and real-time monitoring demonstrations. The method supports sustainable aquaculture practices by enabling non-invasive, continuous evaluation of aquatic environments using affordable, scalable technologies. The approach can be adapted to other aquaculture or environmental monitoring contexts with minimal modification.
- 24 February 2026
- Published by MyJove Corporation in Journal of Visualized Experiments
- No. 228,p. e69744
Real-Time Pond Water Assessment via Embedded Deep Learning and Visual Data Acquisition: A Practical Monitoring Approach for Aquaculture
Maintaining optimal water quality is essential in aquaculture, particularly for ornamental species such as Kois, where color clarity and environmental balance are directly linked to health and aesthetics. This article presents a complete protocol for designing, deploying, and operating a real-time water quality monitoring system based on underwater image classification using embedded deep learning. The proposed system integrates a low-cost underwater camera with a Raspberry Pi-based controller network in an autonomous water surface robot, enabling the capture, transmission, and classification of aquatic images through a lightweight neural network model. Five water condition categories -- ranging from clear to turbid -- are visually distinguished, with classification accuracy exceeding 99% on a custom dataset. The system is configured to run autonomously, offering continuous assessment and visual alerts for deteriorating conditions. This visual protocol covers device assembly, system calibration, model deployment, and real-time monitoring demonstrations. The method supports sustainable aquaculture practices by enabling non-invasive, continuous evaluation of aquatic environments using affordable, scalable technologies. The approach can be adapted to other aquaculture or environmental monitoring contexts with minimal modification.
- Open Access
- 23 February 2026
- Published by Frontiers Media SA in Frontiers in Artificial Intelligence
- Vol. 9, 1767612
Explainable multi-modal deep learning for transparent cancer diagnosis: integrating radiology, clinical features, and decision visualization
Introduction: Although artificial intelligence–based cancer diagnostic models have demonstrated strong predictive performance, their lack of transparency and reliance on single-modality data continue to limit clinical trust and adoption. Effectively integrating multi-modal data with interpret-able decision-making remains a key challenge. Methods: We propose an explainable multi-modal deep learning framework that integrates radiological imaging and structured clinical features using attention-based fusion. Image-level explanations are generated using Grad-CAM++, while SHAP is employed to quantify clinical feature contributions, enabling unified and cross-modal aligned interpretation rather than independent uni-modal explanations. The framework was evaluated on publicly available datasets, including CBIS-DDSM mammography, Duke Breast Cancer MRI, and TCGA cohorts (BRCA, LUAD, and GBM), comprising a total of 3,842 images from 2,917 patients. Results: The proposed model consistently outperformed uni-modal approaches and simple fusion baselines, achieving an improved balance between sensitivity and specificity. Attention-based fusion demonstrated superior performance compared with feature concatenation, and the integration of explainability did not compromise predictive accuracy. Visual and clinical explanations highlighted diagnostically relevant tumor regions and established oncological risk factors. Stable performance across datasets indicates strong generalization capability. Discussion: These results demonstrate that explainable multi-modal learning can effectively combine accuracy, interpret-ability, and robustness, supporting the development of reliable AI-based decision-support systems for cancer diagnosis.
- 23 February 2026
- Published by Frontiers Media SA in Frontiers in Artificial Intelligence
- Vol. 9, 1767612
Open AccessExplainable multi-modal deep learning for transparent cancer diagnosis: integrating radiology, clinical features, and decision visualization
Introduction: Although artificial intelligence–based cancer diagnostic models have demonstrated strong predictive performance, their lack of transparency and reliance on single-modality data continue to limit clinical trust and adoption. Effectively integrating multi-modal data with interpret-able decision-making remains a key challenge. Methods: We propose an explainable multi-modal deep learning framework that integrates radiological imaging and structured clinical features using attention-based fusion. Image-level explanations are generated using Grad-CAM++, while SHAP is employed to quantify clinical feature contributions, enabling unified and cross-modal aligned interpretation rather than independent uni-modal explanations. The framework was evaluated on publicly available datasets, including CBIS-DDSM mammography, Duke Breast Cancer MRI, and TCGA cohorts (BRCA, LUAD, and GBM), comprising a total of 3,842 images from 2,917 patients. Results: The proposed model consistently outperformed uni-modal approaches and simple fusion baselines, achieving an improved balance between sensitivity and specificity. Attention-based fusion demonstrated superior performance compared with feature concatenation, and the integration of explainability did not compromise predictive accuracy. Visual and clinical explanations highlighted diagnostically relevant tumor regions and established oncological risk factors. Stable performance across datasets indicates strong generalization capability. Discussion: These results demonstrate that explainable multi-modal learning can effectively combine accuracy, interpret-ability, and robustness, supporting the development of reliable AI-based decision-support systems for cancer diagnosis.
- 14 January 2026
- Published by MyJove Corporation
- No. 227,p. e69909
JoVE Video Dataset
Congestion in IoT-enabled wireless sensor networks (WSNs) degrades packet delivery, latency, and energy usage, impairing the network, especially under bursty and heterogeneous traffic conditions. This protocol illustrates an intelligent congestion control technique that combines hybrid data aggregation, adaptive scheduling, and a neuro-fuzzy decision engine to efficiently handle network load. The method involves first generating simulation data, creating topologies of different node densities, and setting up traffic patterns using NS-2.35. Packet traces are obtained for each scenario to allow reproducible evaluation. The protocol workflow refers to the combination of two mechanisms: (1) hybrid aggregation, which combines packets in time- and count-based windows while retaining priority labels, and (2) adaptive scheduling, which handles dual priority queues via weighted round robin. A neuro-fuzzy controller always evaluates buffer occupancy, link quality, channel utilization, residual energy, and traffic priority. Taking these inputs, it regulates aggregation depth, queue weights, and transmission decisions by fuzzy inference and neuro-adaptive learning. Performance measurement tasks encompass the calculation of packet delivery ratio, end-to-end latency, throughput, node-level energy consumption, and network lifetime. Statistical analyses are performed across multiple runs to check the reliability of the results. The approach reveals better performance in the simulation compared to the baseline schemes. This protocol offers a reproducible framework for exploring hybrid congestion control methods that enable energy-efficient, scalable, and QoS-aware operation in IoT-enabled WSN environments.
- 14 January 2026
- Published by MyJove Corporation
- No. 227,p. e69909
JoVE Video Dataset
Congestion in IoT-enabled wireless sensor networks (WSNs) degrades packet delivery, latency, and energy usage, impairing the network, especially under bursty and heterogeneous traffic conditions. This protocol illustrates an intelligent congestion control technique that combines hybrid data aggregation, adaptive scheduling, and a neuro-fuzzy decision engine to efficiently handle network load. The method involves first generating simulation data, creating topologies of different node densities, and setting up traffic patterns using NS-2.35. Packet traces are obtained for each scenario to allow reproducible evaluation. The protocol workflow refers to the combination of two mechanisms: (1) hybrid aggregation, which combines packets in time- and count-based windows while retaining priority labels, and (2) adaptive scheduling, which handles dual priority queues via weighted round robin. A neuro-fuzzy controller always evaluates buffer occupancy, link quality, channel utilization, residual energy, and traffic priority. Taking these inputs, it regulates aggregation depth, queue weights, and transmission decisions by fuzzy inference and neuro-adaptive learning. Performance measurement tasks encompass the calculation of packet delivery ratio, end-to-end latency, throughput, node-level energy consumption, and network lifetime. Statistical analyses are performed across multiple runs to check the reliability of the results. The approach reveals better performance in the simulation compared to the baseline schemes. This protocol offers a reproducible framework for exploring hybrid congestion control methods that enable energy-efficient, scalable, and QoS-aware operation in IoT-enabled WSN environments.