As coined by one of its founding fathers, the renowned computer scientist John McCarthy, Artificial Intelligence (AI) refers to “the science and engineering of making intelligent machines, especially intelligent computer programs.” [1] With the rise of AI-enabled technologies in the 21st century, from OpenAI’s ChatGPT to speech recognition, to language processing, it is undeniable that the field has become an integral part of society and our daily lives, with countless companies seeking to leverage AI, and rightly so.
In regard to the increasing usage of AI-enabled technologies within existing industries, the medical science and healthcare industry is no exception. [2] Artificially Intelligent computer systems are able to perform tasks with greater efficiency compared to humans [2], and thus help to automate labour-intensive tasks and streamline processes, providing quick analysis, calculations, and processing of data like never before.
This article aims to explore the various applications of AI-assisted technologies in the medical science field, some of which Meticuly utilises for our implants and surgical guides, which help us deliver our precise and personalised medical solutions.
Predictive Analytics - Forecasting Outcomes for Decision-Making
Predictive analytics refers to the employment of various techniques including AI, modelling, and data mining, in order to discern patterns and relationships within large datasets. Through advanced statistical methods and AI algorithms, models can identify factors tied to certain health outcomes, such as health progression, readmission, and many more. With the digitization of healthcare, datasets are being generated in massive quantities, from electronic medical record (EMR) systems to health claims, to even quasi-health data from patient fitness trackers. [3]
By utilising these various sources of data, predictive analytics can aggregate and make sense of information that was previously hard to capture, enhancing patient care at an individual scale and allowing healthcare providers to identify growing trends as a whole, all the while with increased efficiency and reduced costs.
A prime example of predictive analytics in recent healthcare is the modelling of Covid-19 cases, in order to track infections and inform individuals on their risk of infection, based on geographical region and interactions with infected individuals. [4] This detection and prediction of disease spreading has also been developed for other epidemics such as Ebola. Moreover, researchers have also begun applying predictive analytics to resource management, anticipating patients’ length of stay and readmission rates. By doing so, staff can be allocated more efficiently, and the usage of facilities can be maximised, greatly reducing medical service costs.
Essentially, predictive analytics is a tool that serves to benefit all parties with the efficiency, cost-effectiveness, and insight that it offers. Medical professionals will be able to “deliver the right care, to the right patient, at the right time”. [3]
Computer vision - A picture is worth a thousand words
Computer vision is another key field of AI that enables computers to emulate human vision: by breaking images down into pixels, machines are then able to identify and extract features which can then be classified. [5] The advent of computer vision has allowed for greater precision and efficiency than ever, reducing the margin for human error, while being able to perform tasks within seconds.
In recent years, computer vision’s classification of objects has “achieved human-level performance”, and is increasingly used in medicine for detecting rare or early on-set diseases, many of which the anomalies are hard to discern with the human eye. This early identification can save time, money, and most importantly, lives.
Similarly, the segmentation of objects has also experienced exponential growth – computers are now able to precisely delineate structures within medical images such as MRIs or CT scans, as well as fusing them together to provide both structural and functional insights for comprehensive treatment planning.
In most cases, localising and quantifying morphological features requires extensive training and heavy costs, on top of being time-consuming. Computer vision serves as a solution to all these issues, automating the task and providing unprecedented levels of accuracy. With 90% of all medical data being image-based, the untapped potential of computer vision in healthcare is limitless. [6] Its versatility, precision, and efficiency are sure to continue empowering healthcare professionals with research, diagnosis, and treatment planning across all medical fields.
Machine learning - the core of AI
With the core of AI lying in emulating human intelligence, machine learning is a fundamental component - it essentially provides AI with its ‘learning’ capabilities. Machine learning centres around developing models that enable computers to make predictions or decisions based on data, while continually improving their performance by learning from this data. [7] From the previously mentioned predictive analytics to computer vision, these technologies all rely on machine learning models for complex datasets and automated processes.
A prominent subfield of machine learning, neural networks are a type of model inspired by the structure and function of the human brain. After data is input, it travels through successive layers wherein it is processed each layer by “neurons”, until it reaches the output. [7] Neural networks are highly effective in analysing complex patterns of data and thus are an important tool in predictive analytics, from disease diagnosis, to risk stratification, and even drug discovery, by analysing vast pharmaceutical datasets and the chemical properties of compounds, accelerating the drug development process.
Convolutional Neural Networks (CNN) is a subset of neural networks designed specifically for processing structured grid data, such as medical images. [8] As opposed to traditional computer vision techniques that require painstakingly handcrafted image-processing algorithms, CNN are able to extract hierarchical features directly from raw data and can work with complex high-dimensional datasets. Hence, they are highly suitable for tasks like object detection, image classification, image segmentation, and more. Nowadays, CNN has shown success in detecting, localising, and segmenting a plethora of anatomical features, including but not limited to tumours, lesions, fractures, organ structures, cavities, and cancerous cells.
Machine learning and its various subfields, particularly neural networks, are at the forefront of healthcare innovation, serving as a backbone to most, if not all, AI-assisted technologies. Just as how learning is crucial for human intelligence, machine learning underlies AI functions, and is a critical factor towards the development of artificial intelligence.
Meticuly’s application of AI-assisted technologies
Meticuly’s design process is a testament to the future of the healthcare industry with AI-assisted technologies. Custom-made AI algorithms have been developed with extensive clinical and cadaveric data, automating the labour-intensive modelling tasks originally performed by design engineers. Post-operative results are also compared with pre-op planning, continually improving the predictive capabilities of the algorithms. The AI-assisted engines have significantly streamlined the implant creation process, allowing the 3D implant to be delivered within just 7 days. This efficiency and speed in designing has helped maintain the affordability of Meticuly’s implants.
With a strong foundation in AI-assisted technologies and 3D printing, Meticuly provides personalised medical solutions, delivering precision, reliability, and biocompatibility at the bespoke level. Through powerful technologies and a trusted and meticulous team, we strive for and ensure patient satisfaction and outcomes.
Written by Eclair Sakdibhornssup
References:
(1) Joshi, A.; Mishra, G. Artificial Intelligence. In Proceedings of the International Conference and Workshop on Emerging Trends in Technology; ACM: Mumbai Maharashtra India, 2010; pp 1023–1023. https://doi.org/10.1145/1741906.1742236.
(2) Basu, K.; Sinha, R.; Ong, A.; Basu, T. Artificial Intelligence: How Is It Changing Medical Sciences and Its Future? Indian J Dermatol 2020, 65 (5), 365–370. https://doi.org/10.4103/ijd.IJD_421_20.
(3) Predictive Analytics and the Future of Healthcare. https://www.intel.com/content/www/us/en/healthcare-it/predictive-analytics.html.
(4) Villanustre, F.; Chala, A.; Dev, R.; Xu, L.; LexisNexis, J. S.; Furht, B.; Khoshgoftaar, T. Modeling and Tracking Covid-19 Cases Using Big Data Analytics on HPCC System Platform. Journal of Big Data 2021, 8 (1), 33. https://doi.org/10.1186/s40537-021-00423-z.
(5) Firth-Butterfield, K.; Ammanath, B. How computer vision is set to change healthcare, retail and more. World Economic Forum. https://www.weforum.org/agenda/2022/03/how-computer-vision-change-healthcare/ (accessed 2023-09-13).
(6) Marr, B. 7 Amazing Examples Of Computer And Machine Vision In Practice. Forbes. https://www.forbes.com/sites/bernardmarr/2019/04/08/7-amazing-examples-of-computer-and-machine-vision-in-practice/ (accessed 2023-09-13).
(7) Kufel, J.; Bargieł-Łączek, K.; Kocot, S.; Koźlik, M.; Bartnikowska, W.; Janik, M.; Czogalik, Ł.; Dudek, P.; Magiera, M.; Lis, A.; Paszkiewicz, I.; Nawrat, Z.; Cebula, M.; Gruszczyńska, K. What Is Machine Learning, Artificial Neural Networks and Deep Learning?—Examples of Practical Applications in Medicine. Diagnostics 2023, 13 (15), 2582. https://doi.org/10.3390/diagnostics13152582.
(8) Sarvamangala, D. R.; Kulkarni, R. V. Convolutional Neural Networks in Medical Image Understanding: A Survey. Evol. Intel. 2022, 15 (1), 1–22. https://doi.org/10.1007/s12065-020-00540-3.
Comments