SuFMoFPA: A superpixel and meta-heuristic based fuzzy image segmentation approach to explicate COVID-19 radiological images

Abstract

Coronavirus disease 2019 or COVID-19 is one of the biggest challenges which are being faced by mankind. Researchers are continuously trying to discover a vaccine or medicine for this highly infectious disease but, proper success is not achieved to date. Many countries are suffering from this disease and trying to find some solution that can prevent the dramatic spread of this virus. Although the mortality rate is not very high, the highly infectious nature of this virus makes it a global threat. RT-PCR test is the only means to confirm the presence of this virus to date. Only precautionary measures like early screening, frequent hand wash, social distancing use of masks, and other protective equipment can prevent us from this virus. Some researches show that the radiological images can be quite helpful for the early screening purpose because some features of the radiological images indicate the presence of the COVID-19 virus and therefore, it can serve as an effective screening tool. Automated analysis of these radiological images can help the physicians and other domain experts to study and screen the suspected patients easily and reliably within the stipulated amount of time. This method may not replace the traditional RT-PCR method for detection but, it can be helpful to filter the suspected patients from the rest of the community that can effectively reduce the spread in the of this virus. A novel method is proposed in this work to segment the radiological images for the better explication of the COVID-19 radiological images. The proposed method will be known as SuFMoFPA (Superpixel based Fuzzy Modified Flower Pollination Algorithm). The type 2 fuzzy clustering system is blended with this proposed approach to get the better-segmented outcome. Obtained results are quite promising and outperforming some of the standard approaches which are encouraging for the practical uses of the proposed approach to screening the COVID-19 patients.

Contour-enhanced attention CNN for CT-based COVID-19 segmentation

“…Clustering techniques learn from unannotated datasets and generate patterns that can be used to better explicate COVID-19 CT images. Chakraborty et al applied a superpixel-based clustering approach to process spatial features in COVID-19 infected scans [14] . A modified flower pollination algorithm was designed to explore the search space for forming clusters.…”

Challenges to COVID-19 vaccine supply chain: Implications for sustainable development goals

“…The pandemic has implied huge changes in the way administration associations work ( Narayanamurthy and Tortorella, 2021 ). In the current scenario, a shortfall of vaccines due to failure of the vaccine supply chain (VSC) will make circumstances more confounded ( Chakraborty and Mali, 2020 ). A pandemic VSC is different than that of a traditional VSC because governments are directly procuring vaccines from the manufacturers bypassing the traditional chains of wholesalers and distributors ( Abbasi et al, 2020 ).…”

Metaheuristics based COVID-19 detection using medical images: A review

“…It has two parts, namely local pollination and global pollination. Chakraborty et al [ 81 ] proposed a modified form of FPA by mixing type 2 fuzzy system and named it Superpixel based Fuzzy Modified Flower Pollination Algorithm (SuFMoFPA). This modification improves the computation by selecting and processing the most relevant pixel of the image and ignoring the irrelevant pixels.…”

“…This process is repeated until the cluster centers stop changing. Mittal et al [ 8 ] and Chakraborty et al [ 81 ] adopted this process for the classification of COVID-19 images. The work [ 8 ] improved the performance of K-mean by adding gravitational search optimizer and applied K-mean after completing the pipeline of CNN.…”