Detection of Blast Cells in Microscopic Images Using Optimization Algorithms
M. Venkata Dasu1, P. Subbaiah2, K. Srilatha3
1M. Venkata Dasu, Research Scholar, Department of ECE, Rayalaseema University, Kurnool (Andhra Pradesh), India.
2Dr. P. Subbaiah, Professor, Department of ECE, Nalla Narasimha Reddy Engineering College, Ghatkesar, Hyderabad (Telangana), India.
3K. Srilatha, Assistant Professor, Department of CSE, NNRG, Hyderabad (Telangana), India.
Manuscript received on 15 October 2019 | Revised Manuscript received on 24 October 2019 | Manuscript Published on 02 November 2019 | PP: 2443-2446 | Volume-8 Issue-2S11 September 2019 | Retrieval Number: B12850982S1119/2019©BEIESP | DOI: 10.35940/ijrte.B1285.0982S1119
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Now a day’s in medical field, the most prevalent and hazardous disease is blood cancer. It starts in bone marrow where the blood is produced and prevents many of its regular functions. At an approximation for every three minutes one person in the world is diagnosed with the blood cancer. Early detection of the cancer is necessary for the proper treatment, so as to save the lives of mankind. In general the blood cancer diagnosis will be performed by visual examination of the blood samples of the patient under microscope. The blood cancer detection accuracy of this method depends on the technical skilling abilities of the operator and often leads non-standardized report. To improve patient diagnosis various image processing methods are developed to extract useful information from microscopic images. This could help Hematologists in their diagnostic process. In this paper pre-processing and post-processing methods are applied on microscopic images. These images will be acquired from either public or private database. In this work the images were collected from Microscopyu which is a public database. Pre-processing methods involves color conversion i.e. RGB to grayscale, removal of noise by median filter and an improved contrast enhancement technique CLAHE is implemented. Later Post-processing methods are applied. In this stage Otsu segmentation and optimization algorithms are combined for improving segmentation accuracy. Optimization algorithms used are Particle Swarm optimization (PSO) and Cuckoo Search algorithms (CSO) and Finally features of the segmented image can be extracted from d Scale Invariant Feature Transform (SIFT) . In this work existing method is PSO and proposed method is CSO. At the end the qualitative analysis of the work is done through the statistical parameters like segmentation accuracy, sensitivity, specificity, PSNR and CPU time.
Keywords: Leukemia, Enhancement, ACO, PSO, SIFT.
Scope of the Article: Image Security