Mri brain tumor disease detection usig ml

Mri brain tumor disease detection usig ml

1. Brain Tumour Classification Using Quantum Support Vector Machine Learning Algorithm PRESENTATION by

2. Agenda Abstract Introduction Existing Models Proposed system Proposed models Results Conclusion

3. Abstract This paper presents a novel approach to classifying brain tumors using Quantum Support Vector Machines (QSVM), an emerging quantum machine learning algorithm. As the volume and complexity of medical data grow, classical machine learning models encounter limitations in efficiency and accuracy. Quantum computing, with its ability to process vast amounts of data simultaneously, offers a promising solution. In this study, the QSVM model is applied to the Brats 2015 dataset of brain tumor MRI images, available on Kaggle, to classify tumors as either benign or malignant. The model's performance is compared with a classical SVM equivalent. 3

4. Results show that the QSVM model, implemented on a 32-qubit quantum simulator, was 188 times faster than the classical model and achieved a 1.60% higher accuracy at 95%. Additionally, when tested on a real-time 5-qubit superconducting quantum processor, the QSVM model demonstrated a 24.19% reduction in execution time with the same accuracy as the classical SVM. These findings highlight the potential of quantum machine learning in enhancing the speed and accuracy of medical data classification, specifically in brain tumor diagnosis. 4

5. Introduction  Quantum computing is revolutionizing various fields, offering solutions to computational problems that classical computers struggle to address efficiently. One of the key areas where quantum computing is making a significant impact is in machine learning, where massive datasets and complex algorithms often challenge conventional methods. As data continues to grow exponentially, the demand for more powerful computational models has led researchers to explore quantum machine learning (QML) approaches. In this context, Quantum Support Vector Machines (QSVM) emerge as a promising tool for classification tasks, offering the potential for faster and more accurate data processing. 5

6. In healthcare, accurate and timely diagnosis is crucial, particularly in the field of oncology, where early detection can significantly improve patient outcomes. Brain tumors, both benign and malignant, present a major diagnostic challenge due to the complexity of medical imaging data. Traditional machine learning algorithms, such as Support Vector Machines (SVM), have been used extensively for tumor classification tasks. However, their performance is often limited by the computational complexity involved in processing high-dimensional data. This is where quantum computing, with its ability to process multiple states simultaneously through the principles of superposition and entanglement, offers a unique advantage. 6

7. The results demonstrate that QSVM significantly outperforms classical SVM, offering both higher accuracy and reduced computational time. Specifically, the kernel-based QSVM model implemented on a 32-qubit quantum simulator achieved 95% accuracy, outperforming its classical counterpart by 1.60%. Furthermore, the execution time of the QSVM model was 188 times faster than the classical SVM model on the same dataset . The findings of this research highlight the potential of quantum machine learning in enhancing the efficiency and accuracy of medical diagnosis. As quantum computing continues to evolve, its integration into healthcare systems could lead to significant advancements in fields such as computer-aided diagnosis (CAD), providing healthcare professionals with more reliable tools for disease detection and treatment planning. 7

8. 8 (a) Normal Brain, (b) Benign Tumor, (c) Malignant Tumor

9. Literature Review SNO Title Year of Publish Technique/Protocol/Method/Algorithm Parameters Advantages 1 Brain Tumour Classification Using Quantum Support Vector Machine Learning Algorithm 2023 Quantum Support Vector Machine Quantum Kernel, Superconducting Processors 88x faster, 95% accuracy, reduced execution time 2 Quantum Machine Learning 2017 QSVM, k-means clustering, Grover's Algorithm Speedup, Kernel-based techniques Quantum speedup for classification and clustering 9

10. 3 Advances in Quantum Machine Learning 2015 QSVM, HHL Algorithm, k-Nearest Neighbor Quantum feature maps, dimensionality reduction Speedup in classification and clustering problems 4 Quantum Machine Learning: A Review and Current Status 2020 HHL, SVM, QSVM Superposition, Quantum Feature Mapping Faster computations, better data handling 5 Quantum Machine Learning for Breast Cancer Classification 2020 QSVM for Breast Cancer Dataset Quantum speedup, kernel tricks 234x faster, efficient handling of big data 10

11. 6 Experimental Evaluation of Quantum Machine Learning Algorithms 2023 Experimental Evaluation of Quantum Machine Learning Algorithms Accuracy improvement, Kernel-based models 3-4% better accuracy than classical models 7 Quantum Machine Learning for Support Vector Machine Classification 2022 QSVM, Quantum Feature Mapping Dimensionality Reduction, Speedup Faster than classical SVM, Reduced processing time 8 Implementation of Quantum Support Vector Machine Algorithm 2022 Quantum Variational Classifier, Deep Learning Ensemble Transfer Learning, Quantum Feature Mapping Detection score of 90.9% 11

12. 9 Implementation of Quantum Support Vector Machine Algorithm 2022 QSVM, MNIST Dataset Speedup, Quantum Feature Mapping 81.62% computational efficiency 10 The Role of Quantum-Enhanced SVM using Multiparametric MRI Parameters 2022 QSVM, Multiparametric MRI for tumor differentiation Quantum Kernel, MRI parameters 90% test score for brain tumor classification 12

13. EXISTING SYSTEM 13 The existing system for brain tumor classification using Quantum Support Vector Machine (QSVM) relies on the integration of machine learning techniques with quantum computing. Traditional machine learning, specifically classical Support Vector Machines (SVM), are effective for data classification, but face challenges when dealing with large, complex datasets . To address these challenges, QSVM is proposed as a solution. Quantum computers, which operate using qubits , leverage superposition and entanglement, enabling them to handle multiple states simultaneously and process large amounts of data faster than classical systems .

14. 14 In the existing system described in the reference, QSVM is applied to classify brain tumors into benign or malignant categories based on MRI images. The dataset, Brats 2015, is used, and the QSVM model was implemented on both quantum simulators and real-time quantum machines. Results show that the quantum model achieved 95% accuracy and was up to 188 times faster than its classical counterpart. This system is a significant advancement, particularly in the healthcare sector, where quantum machine learning can assist in faster, more accurate diagnosis of brain tumors​

15. EXISTING SYSTEM DRAWBACKS 15 High Complexity : Implementing the QSVM algorithm, especially for large datasets, involves constructing complex quantum circuits that require significant computational resources ​ Increased Execution Time on Real Quantum Computers : While the QSVM algorithm provides exponential speed-up when run on simulators, it faces limitations when executed on real-time quantum processors due to quantum circuit complexity, increasing execution time ​ Quantum Feature Mapping Challenges : Selecting an optimal quantum feature map with appropriate repetitions is critical to ensure efficient data encoding, and incorrect configurations can lead to poor performance​

16. 16 Hardware Limitations : Real-time quantum computers, like the IBMQ backend, are still limited in terms of the number of qubits and processing speed, which constrains the ability to handle larger datasets Potential for Misclassification : The QSVM allows for a small degree of misclassification (slack variables) to prevent overfitting , but this can sometimes result in decreased accuracy for specific tasks​

17. Proposed system Data collection Data Loading and Exploration Data Preprocessing Model Building Model Comparison 17

18. Data collection: In this project I'm collecting the data in kaggle datasets 18

19. Data Loading and Exploration: Loading the kidney disease dataset Checking for missing values and dealing with them through random value imputation and mode imputation Exploratory data analysis (EDA) through visualizations like histograms, violin plots, scatter plots to understand data distributions 19

20. Data Preprocessing: Encoding categorical features using label encoding Splitting data into training and test sets 20

21. Model Building: Implementing various machine learning classification algorithms like QSVM, Random forest , logistic regression, svm , XG Boost Tuning hyperparameters of some models using techniques like GridSearchCV Evaluating models on test data using metrics like accuracy, confusion matrix, classification report 21

22. Model Comparison: Comparing the performance scores (accuracy) of the different models Identifying the best performing models like QSVM, Random forest , logistic regression, svm , XGBoost for Brain tumor disease prediction 22

23. Proposed block 

24. Proposed MODELS Quantum SVM Random Forest Logistic regression XGBoost

25. COMPARISION TABLE SNO MODEL SCORE 1 Extra Trees Classifier 0.991667 2 Gradient Boosting Classifier 0.983333 3 Stochastic Gradient Boosting 0.983333 4 XgBoost 0.983333 5 Cat Boost 0.983333 6 Decision Tree Classifier 0.975000 7 Random Forest Classifier 0.975000 8 Ada Boost Classifier 0.975000 9 KNN 0.716667

26. MODELS COMPARISON 26

27. Conclusion 27 In the classification of brain tumors, Quantum SVM demonstrated superior accuracy and computational efficiency compared to classical models like Random Forest, Logistic Regression, and XGBoost . The Quantum SVM achieved higher speed and precision, making it highly effective for processing complex datasets, highlighting quantum computing's potential in machine learning.

28. Feature scope 28 The feature scopes used by models such as Quantum SVM, Random Forest, Logistic Regression, and XGBoost are defined by their ability to handle complex data. Quantum SVM leverages quantum computing for efficient high-dimensional classification, Random Forest is effective for feature selection and handling imbalanced datasets, Logistic Regression is best for binary classification with interpretability, and XGBoost excels in feature interactions and handling missing values with gradient boosting.

29. REfrences [1] J. Biamonte , P. Wittek , N. Pancotti , P. Rebentrost , N. Wiebe , and S. Lloyd, “Quantum machine learning,” Nature, Vol. 549, no. 7671, pp. 195–02, 2017. DOI: 10.1038/nature23474 [2] . J. Adcock, et al. “Advances in quantum machine learning.” arXiv preprint arXiv:1512.02900, 2015. [3] N. Mishra , et al., “Quantum machine learning: A review and current status quantum machine learning,” Data Manag . Anal. Innov ., Vol. 2, pp. 101–45, 2020 [ 4 S. Saini , P. K. Khosla , M. Kaur , and G. Singh, “Quantum driven machine learning,” Int. J. Theor . Phys, Vol. 59, no. 12, pp. 4013–24, 2020. DOI: 10.1007/s10773-020-04656-1 [5 ] Dervla M. Connaughton et al. Monogenic causes of Chronic Kidney Disease in adults, Clinical investigation, 2019, volume 95, issue 4, p914-928.

30. Any Quires? 30

31. THANK YOU