Research

Artificial Intelligence is not only about robotics assisting medical specialists in providing medicine and care in a faster and better way than before. It also entails other subject areas such as data science, artificial vision, financial technology and creative technologies in particular. I chose to focus my research on data science by using large data sets related to medical data of lung and breast cancers to provide a more efficient diagnosis to patients with similar symptoms.

In the past 10 years, artificial intelligence has been a priority in the development of Information Technology so much so that due to its development, we have come to a stage where large companies such as Microsoft, Facebook and Google have shifted their products to benefit the most from it. [1] [2] [3] In particular, the importance of data and its correct usage, has become extremely important to the modern world with new regulations recently launched on the conservation of data and its usage in relevance to the service being offered. [4] There have already been some studies on the use to machine learning techniques to diagnose Breast Cancer. Most of these use a recurrent neural network to be able to classify different features within a large data set, to then be able to correctly diagnose these types of cancer. The selected features would be age of initial presentation, tumor size, sentinel lymph node involvement, progesterone receptor expression, exogenous hormone use, and type of surgery. [5] [6] [7] Others such as Kim et al. [8] used normalized mutual information index for feature selection and supported vector machines.

The data sets are publicly available, and through machine learning algorithms, we can build a proof of concept to be used by local medical specialists. [8] Among the better designed and validated research studies, it becomes clear that machine learning methods can substantially improve the accuracy of predicting cancer susceptibility by 15–25%. It also becomes evident that machine learning can help improve the basic understanding of cancer development and progression in different patients.

This research is very much in line with the focus point of Horizon 2020; the EU strategy for Health Care for the years 2014 – 2020. Horizon 2020 is the biggest EU Research and Innovation program ever. It promises more breakthroughs, discoveries and world-firsts by taking great ideas from the lab to the market. [9] In particular, this program has set its goals for a healthier Europe, focusing on providing research to address chronic diseases; cancer in particular. [10] Since breast cancer is the most common cancer in women in the EU, the Commission’s Joint Research Centre, works towards a voluntary, evidence-based quality assurance scheme for breast cancer services. Having already existing vast data sets regarding the symptoms of this kind of cancer, puts this research as one of the most important points which Horizon 2020 is tackling. There have been already a number of research projects related to different kinds of cancer, which have benefited from the Horizon 2020 program.

[1] Facebook Press, “Open AI Frameworks, New AR/VR Advancements, and Other Highlights
from Day 2,” Facebook Press, 2 May 2018. [Online]. Available: https://newsroom.fb.com/news/2018/05/f8-2018-day-2/. [Accessed 20 July 2019].

[2] L. Goode, “AT GOOGLE I/O 2018, EXPECT ALL AI ALL THE TIME,” Wired.com, 07 May 2018. [Online]. Available: https://www.wired.com/story/google-io-2018-what-to-expect/. [Accessed 20 July 2019].

[3] Microsoft, “Artificial Intelligence,” Microsoft, 2016. [Online]. Available: https://www.microsoft.com/en-us/research/research-area/artificial-intelligence/. [Accessed 20 July 2019].

[4] C. R. Sven Jacobs, “Data privacy: AI and the GDPR,” Norton Rose Fulbright, 7 November 2017. [Online]. Available: https://www.aitech.law/blog/data-privacy-ai-and-the-gdpr. [Accessed 20 July 2019].

[5] D. S. W. Joseph A. Cruz, “Applications of Machine Learning in Cancer Prediction and Prognosis,” Cancer Inform., vol. 2006, no. 2, pp. 59 – 77, 2006.

[6] N. P. e. al., “Learning from data to predict future symptoms of oncology patients,” Plos One, 31 December 2018.

[7] IBM Watson Health, “Artificial Intelligence in medicine,” IBM Watson Health, [Online]. Available: https://www.ibm.com/watson-health/learn/artificial-intelligence-medicine. [Accessed 20 July 2019].

[8] F. Lab, “Fuchs Lab Breast Cancer Dataset,” 15 July 2019. [Online]. Available: http://thomasfuchslab.org/data/. [Accessed 20 July 2019].

[9] European Commission, “Health | Horizon 2020,” European Commission, [Online]. Available: https://ec.europa.eu/programmes/horizon2020/en/area/health. [Accessed 20 July 2019].

[10] European Commission, “Featured Projects | Horizon 2020,” European Commission, [Online]. Available: https://ec.europa.eu/programmes/horizon2020/en/newsroom/featured-projects/all. [Accessed 20 July 2019].