It's only been 20 years since the twenty-first century and, certainly, one of the best momentous advancements and enabling impacts for the human culture of this century will be Artificial Intelligence. It is a grounded fact that AI and related operations and stages are set to reform the global point of view, impact productivity for the better, improve lifestyle, and generate huge amounts of wealth. For instance, the research firm Gartner expects worldwide AI-based economic activities to be increased from $1.2 trillion in 2018 to about $3.9 trillion in 2022 [1]. AI is still in a developing stage and therefore there are questions about its accountability but explainable AI is a technology that is helping in this matter because it is very important as in healthcare the stakes are very high [6].

AI in healthcare can be of great use in many fields but to analyse we have classified Artificial Intelligence in 3 phases:-

• Firstly, the AI can help address the low-level problem which is yet very effective and costs a major chunk of resources that is the administrative work that consumes a lot of time of the doctors, nurses, and other hospital staff and AI can come in handy to automate all this work and save the time of the staff to treat patients [5]. In the first scale, we can also include the AI applications based on imaging which are already being used in clinical fields such as radiology, pathology, etc [2]. • The second approach in which AI can be helpful is that it can be used in remote monitoring, that is if the patient is transferred from hospital to home their monitoring could be done effectively [5]. These can include AI-powered alerting systems or virtual assistants that if a patient is in need then gives an alert to the doctor. Furthermore, Natural Language Processing has a clinical use that it can use to convert the text to structured data which then helps in a better understanding of the patient’s medical history and also classify patients and summarise information. • The third phase witnessed a more refined approach towards the diagnostics part as it can be used to diagnose patients and predict/suggest a more practical and feasible cure based on the previous clinical trials and AI can also help in improving clinical decision making [5].

Artificial Intelligence in Medical Imaging

Traditional methods/techniques in the field of artificial intelligence largely depend on the predefined algorithms which use feature engineering based on the previous results and conclusions. These features are utilized to plan 3D imaging of growth or the intratumoral surface and dispersion of pixel powers (histogram). A subsequent assurance step ensures that main the applicable and significant elements are utilized for the displaying. Authentic AI models are then fit this data to perceive potential imaging-based biomarkers. Occurrences of these models fuse support vector machines(SVM) and irregular timberlands [3]. Recently, pixel/voxel-based has been discovered in medical imaging which does not require the explicit input of features which thus reducing the scope of error and giving more precise results [8]. Also, deep learning technologies such as convolutional neural networks have been showing great progress in the restructuring of images and improving the resolution which is being used in a more precise view of the human body scans [7].
Within healthcare, AI has been incorporated in many applications some of which include drug discovery, remote patient monitoring, medical diagnostics and imaging, risk management, wearables, virtual assistants and hospital management, etc.

Clinical Applications using AI:

• Thoracic Imaging --- Cellular breakdown in the lungs is quite possibly the most well-known and dangerous tumour. Cell breakdown in the lungs screening can help with perceiving aspiratory nodules, with early acknowledgment being lifesaving in various patients. AI can help in normally recognizing these nodules and characterizing them as harmless or dangerous [3].
• Colonoscopy --- Colonic polyps that are undetected or misclassified address a possible risk of colorectal malignant growth. But most polyps are from the outset innocuous, they can become perilous over time. Henceforth, early recognition and steady observing with hearty AI-based instruments would be very useful [3].
• Brain Imaging --- Whether malignant, benign, or asymptomatic, brain tumours display abnormal tissue growth; Artificial Intelligence makes it possible to predict their presence as well as their prognosis [3].
• DNA Sequencing --- With the increasing availability of sequencing data, genomic endpoints are becoming increasingly useful in cancer diagnosis and treatment. AI-based tools can analyse and extract high-level features to correlate somatic point mutations with cancer types as well can be useful in predicting the genetic mutations [3].

References

1. Tirthajyoti Sarkar : AI and machine learning for healthcare https://towardsdatascience.com/ai-and-machine-learning-for-healthcare-7a70fb3acb67
2. Mohammed Yousef Shaheen. AI in Healthcare: medical and socio-economic benefits and challenges. ScienceOpen Preprints.
DOI: 10.14293/S2199-1006.1.SOR-.PPRQNI1.v1
3. Hosny A, Parmar C, Quackenbush J, Schwartz LH, Aerts HJWL. Artificial intelligence in radiology. Nat Rev Cancer. 2018;18(8):500-510.
doi:10.1038/s41568-018-0016-5
4. Goldenberg, S., Nir, G. & Salcudean, S.E. A new era: artificial intelligence and machine learning in prostate cancer. Nat Rev Urol 16, 391–403 (2019).
https://doi.org/10.1038/s41585-019-0193-3>
5. Angela Spatharou, Solveigh Hieronimus, and Jonathan Jenkins. Transforming healthcare with AI: The impact on the workforce and organizations
https://www.mckinsey.com/industries/healthcare-systems-and-services/our-insights/transforming-healthcare-with-ai
6. U. Pawar, D. O’Shea, S. Rea and R. O’Reilly, "Explainable AI in Healthcare," 2020 International Conference on Cyber Situational Awareness, Data Analytics and Assessment (CyberSA), 2020, pp. 1-2,
doi: 10.1109/CyberSA49311.2020.9139655.
7. Kim, Mingyu et al. “Deep Learning in Medical Imaging.” Neurospine vol. 16,4 (2019): 657-668.
doi:10.14245/ns.1938396.198
8. Kenji Suzuki. 2012. Pixel-based machine learning in medical imaging.Journal of Biomedical Imaging 2012, Article 1 (January 2012), 1 pages.
DOI : https://doi.org/10.1155/2012/792079