Rise of the machines

News

  Posted by: The Probe      17th April 2020

Artificial intelligence (AI) has been a mainstay of science fiction for decades but today, AI is already influencing our lives in ways large and small. Across the internet, algorithms track and model user behaviour to serve content and advertisements. In the stock markets, algorithms now carry out the vast majority of trades, buying and selling in microseconds (a millionth of a second).[i] The blink of an eye used to be a byword for imperceptible quickness, but even that takes around a quarter to a third of a second – long enough for machines to carry out hundreds of thousands of trades.[ii]

Medicine and dentistry are also seeing the first impacts of the AI revolution. As you are no doubt aware, we face a burgeoning crisis of antibacterial resistance. We generally place the blame on suboptimal usage and prescribing patterns, as well as the incredible adaptability of bacteria. These are key-drivers of the problem, but economics has also contributed. In short, it is very difficult and expensive to develop broad-spectrum antibiotics and once developed, they only have a limited time span, during which they will remain effective. This limits how much recurring profit can be made on them. For antibiotics, phase I-III clinical trials alone cost upwards of £100,000,000, and only a vanishingly small proportion of clinical trials will ultimately yield a marketable antibiotic. These economic barriers to development have led to a dramatic drop in novel antibiotics reaching the market since 1990, even as antibiotic consumption has surged (increasing by 65% globally just between 2000 and 2015).[iii]

One area where algorithms excel is in crunching through mind-boggling quantities of data and permutations. But it’s not just raw computing power that is fuelling the latest developments. Computers are reliant on code and historically, it has been humans that write that code with all the short-comings that entails. More recently, we have developed techniques that essentially teach computers how to learn. How this is achieved is beyond the scope of this article, but one popular method essentially tests generation after generation of somewhat randomised code. Where code performs better, it is used as the template for the next generation, in a process reminiscent of biological evolution. When successful, this has produced algorithms that can be uncannily effective at their assigned task (though success is still far from guaranteed).

Recently, MIT researchers announced that they had successfully utilised a machine-learning algorithm to identify a powerful known antibiotic (alongside several other potential candidates), which they dubbed “halicin” – after the AI in 2001: A Space Odyssey. The new algorithm is capable of screening more than a hundred million chemical compounds in just days. Efforts to leverage machine learning by these and other researchers holds the promise of greatly reducing the cost barriers of antibiotic development.[iv], [v]

Within dentistry, machine learning is being looked into for diagnostic and predictive applications. This is still at an early stage, due in part to a paucity of large-scale standardised datasets, which deep learning techniques require. For example, there is a great deal of variability in CBCT imagery as a result of differences in equipment and exposure conditions.[vi] While this presents a challenge, efforts are already well under way to use machine learning in order to augment diagnoses, more accurately identify and weigh dental risk factors for select populations, better predict the occurrence of complications such as bisphosphonate-related osteonecrosis of the jaw following dental extraction, assess dental pulp stem cells and platelet derivatives, and enhance optical coherence tomography – among many other applications.[vii], [viii], [ix], [x], [xi]

Beyond AI, advances in robotics and cybernetics are reaching patients. The NHS is beginning to deploy robots for minimally invasive keyhole procedures. These robots are controlled by surgeons, who can operate from a comfortable seated position. For surgeons, this reduces fatigue and potentially career-shortening repetitive strain injuries. For patients, this technology promises higher degrees of precision, reduced pain, recovery time and instances of iatrogenic injury.[xii]

While AI and robotics promise to revolutionise dentistry and medicine, it by no means eliminates the need for skilled clinicians. AI can produce very impressive results for a given task, but it is a long way from possessing the adaptability and judgement that humans are capable of. Therefore, newer technologies should be examined for their potential benefits. For example, more and more clinicians are adopting solutions that enable the delivery of predictable and effective dental implant treatment. Supported by over 1,000 scientific studies, the Osstell Beacon™ from W&H can be used to assess the progress of osseointegration and determine the optimal time to load an implant. This ultimately facilitates effective management of patients with risk factors, ensuring unnecessarily long treatment times are avoided and more predictable outcomes are achieved.

It’s easy to get carried away thinking about where technology might take us, but that can lead to us overlooking the practical advances that are available today. To the outside observer, the latest dental equipment can look like something out of science fiction, but it is tried and tested technology that is being used with outstanding results right now.

 

To find out more visit www.wh.com/en_uk, call 01727 874990 or email office.uk@wh.com

 

[i] Peters M. Algorithmic capitalism in the epoch of digital reason. Fast Capitalism. 2017; 14(1): 65-74. https://doi.org/10.32855/fcapital.201701.012 February 21, 2020.

[ii] Golan T., Grossman S., Deouell L., Malach R. Widespread suppression of high-order visual cortex during blinks and external predictable visual interruptions. bioRxiv. 2018. https://doi.org/10.1101/456566 February 21, 2020.

[iii] Renwick M., Mossialos E. What are the economic barriers of antibiotic R&D and how can we overcome them? Expert Opinion on Drug Discovery. 2018; 13;10: 889-892. https://doi.org/10.1080/17460441.2018.1515908 February 21, 2020.

[iv] Trafton A. Artificial intelligence yields new antibiotic. MIT News. 2020. http://news.mit.edu/2020/artificial-intelligence-identifies-new-antibiotic-0220 February 20, 2020.

[v] Yoshida M., Hinkley T., Tsuda S., Abul-Haija Y., McBurney R., Kulikov V., Mathieson J., Reyes S., Castro M., Cronin L. Using evolutionary algorithms and machine learning to explore sequence space for the discovery of antimicrobial peptides. Chem. 2018; 4(3): 533-543. https://doi.org/10.1016/j.chempr.2018.01.005 February 21, 2020.

[vi] Hwang J., Jung Y., Cho B., Heo M. An overview of deep learning in the field of dentistry. Imaging Science in Dentistry. 2019; 49(1): 1-7. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6444007/ February 21, 2020.

[vii] Lee J., Kim D., Jeong S., Choi S. Detection and diagnosis of dental caries using a deep learning-based convolutional neural network algorithm. Journal of Dentistry. 2018; 77: 106-111. https://doi.org/10.1016/j.jdent.2018.07.015 February 21, 2020.

[viii] Yoon S., Choi T., Odlum M., Mitchell D., Kronish I., Davidson K., Finkelstein J. Machine learning to identify behavioral determinants of oral health in inner city older Hispanic adults. Studies in Health Technology and Infomatics. 2018; 251: 253-256. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6211787/ February 21, 2020.

[ix] Kim D., Kim H., Nam W., Kim HJ., Cha I. Machine learning to predict the occurrence of bisphosphonate-related osteonecrosis of the jaw associated with dental extraction: a preliminary report. Bone. 2018; 116: 207-214. https://doi.org/10.1016/j.bone.2018.04.020 February 21, 2020.

[x] Bindal P., Bindal U., Kazemipoor M., Kazemipoor M., Jha S. Hybrid machine learning approaches in viability assessment of dental pulp stem cells treated with platelet-rich concentrates on different periods. Applied Medical Informatics. 2019; 41(3): 93-191. https://eprints.utas.edu.au/32137/ February 21, 2020.

[xi] Salehi H., Mahdian M., Murshid M., Judex S., Tadinada A. Deep learning-based quantitative analysis of dental careis using optical coherence tomography: an ex vivo study. Proc. SPIE. 2019; 10857. https://doi.org/10.1117/12.2510076 February 21, 2020.

[xii] Chowdhury H. Robot surgeons to begin work in NHS hospitals. The Telegraph. 2020. https://www.telegraph.co.uk/technology/2020/02/20/british-start-up-cmr-surgical-launches-robot-surgeons-nhs/ February 21, 2020.


No Comments


No comments yet.


Sorry, the comment form is closed at this time.