2020 Third International Conference on Smart Systems and Inventive Technology (ICSSIT) · 2020 · 11 citations · 13 references
MedicineOncologyDiagnostic ImagingMachine LearningNeuro-oncologyBrain TumorsCancer DetectionComputer-aided DiagnosisMedical ImagingTumor ClassificationTumor DetectionAutomatic ClassificationBrain Tumor DiagnosisClassification System DesignClassification System EvaluationComputer-aided Tumor DiagnosisConvolutionConvolutional Neural Network DetectionK-nearest Neighbors
As per reports, Brain tumour is among the top in deaths in comparison with all kinds of tumours. Accordingly in 2019, a total of 65000 brain tumour cases were reported including children and adults. Among them 20% of deaths occurred and it may increase in further time. MRI (Medical Resonance Imaging) scan is the used to identify the tumours in the brain. As tumour cells are of different sizes, it is very difficult for the neurosurgeons to identify small-sized tumours through MRI. In order to avoid stress on neurosurgeons, an automated brain tumour detection system is developed by employing FCM (Fuzzy C-Means) and Clustering-based CNN (Convolution Neural Network) classification system is proposed to deliver an accuracy of 91%. Comparison with KNN (K- Nearest Neighbour) and BPNN (Back propagation Neural Network) is done, where CNN gives enhanced accuracy in comparison with other classifiers.
13