Global Journal of Computer Science and Technology, C: Software & Data Engineering, Volume 22 Issue 2

Our experiments show that our system provides an optimal design for using deep learning models running on a multicore single machine for image analysis. We benchmarked our system with a baseline serial architecture and a baseline parallel architecture using standardized evaluation metrics: execution time, speedup, and CPU usage. These metrics are used to calculate the performance of a system. Our results indicate that the proposed system is approximately 78% faster than its baseline serial system counterpart and 12% faster than a baseline parallel system. As demonstrated by our evaluation criteria, our system exhibits an acceptable industrial performance compared to the other two presented baseline systems. This argument is further strengthened because our system is currently used as a product in an industrial setting to calculate and track capillary changes in patients with pancreatitis, COVID-19, and acute heart diseases. The clinical researchers welcomed using this system to analyze their medical images locally. This acceptance was mainly due to the system reducing analysis time and removing the risks of uploading the data to a thirdparty cloud provider. Our code has been made public as an open- source project in a GitHub repository for testing and usage by other clinical users. The users can import the package into their Python environment and immediately start using it. Moreover, users who can clone the code from GitHub can swap our algorithm with theirs, showing that our architecture can be generalized and utilized in the context of other use cases that require image analysis running on a CPU in near real-time. Thus, the generality of our approach can be justified by several other use cases that require image analysis. A cknowledgment The authors would like to thank the Research Council of Norway for providing the necessary funds for this project. The research carried out was funded under these two projects; Industrial Ph.D. project nr: 305716 and BIA project nr: 282213. We would also like to thank ODI Medical AS for providing the requirements, testing the system, and integrating it as part of their e-health application. R eferences R éférences R eferencias 1. R. Duncan, “A survey of parallel computer architectures, ”Computer, vol. 23, no. 2, pp. 5–16, 1990. 2. Q. Zhang, L. T. Yang, Z. Chen, and P. Li, “A survey on deep learning for big data,” Information Fusion, vol. 42, pp. 146–157, 2018. 3. N. C. Thompson, K. Greenewald, K. Lee, and G. F. Manso, “The computational limits of deep learning,” arXiv preprint arXiv:2007.05558, 2020. 4. S. Pumma, M. Si, W.-c. Feng, and P. Balaji, “Parallel i/o optimizations for scalable deep learning,” in 2017 IEEE 23rd International Conference on Parallel and Distributed Systems (ICPADS), IEEE, 2017, pp. 720– 729. 5. J. M. M. Rumbold and B. Pierscionek, “The effect of the general data protection regulation on medical research,” Journal of medical Internet research, vol. 19, no. 2, e47, 2017. 6. M. J.S. Parker and N. W. McGill, “The established and evolving role of nailfold capillaroscopy in Connective-Tissue disease,” in Connective Tissue Disease – Current State of the Art [Working Title], IntechOpen, 2019. 7. V. Nama, J. Onwude, I. T. Manyonda, and T. F. Antonios, “Is capillary rarefaction an independent risk marker for cardiovascular disease in south asians?” en, J. Hum. Hypertens., vol. 25, no. 7, pp. 465–466, Jul. 2011. 8. B. Fagrell and M. Intaglietta, “Microcirculation: Its significance in clinical and molecular medicine,” Journal of internal medicine, vol. 241, no. 5, pp. 349–362, 1997. 9. A. J. H. M. Houben, R. J. H. Martens, and C. D. A. Stehouwer, “Assessing microvascular function in humans from a chronic disease perspective,” en, J. Am. Soc. Nephrol., vol. 28, no. 12, pp. 3461–3472, Dec. 2017. 10. T. Wester, Z. A. Awan, T. S. Kvernebo, G. Salerud, and K. Kvernebo, “Skin microvascular morphology and hemodynamics during treatment with venoarterial extra-corporeal membrane oxygen- nation,” en, Clin. Hemorheol. Microcirc., vol. 56, no. 2, pp. 119–131, 2014. 11. A. L´opez et al., “Effects of early hemodynamic resuscitation on left ventricular performance and microcirculatory function during endotoxic shock,” en, Intensive Care Med Exp, vol. 3, no. 1, p. 49, Dec. 2015. 12. G. Hernandez, A. Bruhn, and C. Ince, “Microcirculation in sepsis: New perspectives,” Current vascular pharmacology, vol. 11, no. 2, pp. 161–169, 2013. 13. M. Helmy, A. Dykyy, T. T. Truong, P. Ferreira, and E. Jul, “Capillarynet: An automated system to analyze microcirculation videos from handheld vital microscopy,” arXiv preprint arXiv: 2104.11574, 2021. 14. A. Holmes, Hadoop in practice. Manning New York; 2012, vol. 3. 15. D. Borthakur et al., “Hdfs architecture guide,” Hadoop Apache Project, vol. 53, no. 1-13, p. 2, 2008. 16. K. Shvachko, H. Kuang, S. Radia, and R. Chansler, “The hadoop distributed file system,” in 2010 IEEE 26 th symposium on mass storage systems and technologies (MSST), Ieee, 2010, pp. 1–10. Capillary X: A Software Design Pattern for Analyzing Medical Images in Real-time using Deep Learning Global Journal of Computer Science and Technology Volume XXII Issue II Version I 21 Year 2022 ( ) C © 2022 Global Journals