Lung Cancer Detection Using Fusion of Medical Knowledge and Content Based Image Retrieval for LIDC Dataset

Semantic gap is proved to be major gap in the field of Content-based medical image retrieval (CBMIR). The contribution of this work is to fuse medical knowledge with CBMIR approach in the findings of lung nodules using LIDC database by National Cancer Institute (NCI), USA. It includes the development of lung nodule database with available biopsy reports as well as radiologist's annotations along with the computer generated features. This fusion resulted in an average precision of 92.8%, classification accuracy of 88.6%, sensitivity of 89% and specificity of 95% using 80 biopsy reports. The results have been further validated with PGIMER, Chandigarh dataset and classify the nodules in three classes: Benign, Malignant and Metastasis which is the novelty of this work. Experimental results show that the proposed parameters and analysis improves the semantic performance while reducing the computational complexity as well as physician's diagnostic time using the slice selection mechanism. The research also gives a message to medical community that by preserving the already diagnosed cases in hospitals along with physician's diagnostic details can certainly bring a revolution in better diagnosis of new cases at an early stage.