Two unlikely interweaving sciences, artificial intelligence and biomedicine, have changed our health and lives. These two sciences have now intertwined further, aiding scientists, medical professionals, and, ultimately, all of us to improve our ongoing health so we can live better lives. This article will introduce some of the ways these two sciences are working together to solve medical mysteries and problems that have plagued us for generations.
The field of biomedical sciences is quite broad, dealing with several disciplines of scientific and medical research, including genetics, epidemiology, virology, and biochemistry. It also incorporates scientific disciplines whose fundamental aspects are the biology of health and diseases.
In addition, biomedical sciences also aim at relevant sciences that include but are not limited to cell biology and biochemistry, molecular and microbiology, immunology, anatomy, bioinformatics, statistics, and mathematics. Because of this wide breadth of areas that biomedical sciences touches, the research, academic, and economic significance it spans are broader than that of hospital laboratory science alone.
Artificial intelligence, applied to biomedical science, uses software and algorithms with complex structures designed to mirror human intelligence to analyse medical data. Specifically, artificial intelligence provides the capability of computer-trained algorithms to estimate results without the need for direct human interactions.
Some critical applications of AI to biomedical science are clinical text mining, retrieval of patient-centric information, biomedical text evaluation, assisting with diagnosis, clinical event forecasting, precision medicine, data-driven prognosis, and human computation.
The Massachusetts Institute of Technology has developed an AI model that can automate the critical step of medical decision-making. This process is generally a task for experts to identify essential features found in massive datasets by hand.
The MIT project automatically identified the voicing patterns of patients with vocal cord nodules (see graphic below). These features were used to predict which patients had or did not have the nodule disorder.
Vocal nodules may not seem like a critical medical condition to identify. However, the field of predictive analytics has increasing promise, allowing clinicians to diagnose and treat patients. For example, AI models can be trained to find patterns in patient data. AI has been utilised in sepsis care, in the design of safer chemotherapy regimens, to predict a patient’s risk of dying in the ICU or having breast cancer, among many others.
At the University of Zurich, academics use artificial intelligence to create biomedical imaging using machine learning methods that improve optoacoustic imaging. This technique can study brain activity, visualise blood vessels, characterise skin lesions, and diagnose cancer.
The quality of the images rendered depends on the number of sensors used by the apparatus and their distribution. This novel technique developed by Swiss scientists allows for a noteworthy reduction in the number of sensors needed without reducing the image quality. This allows for a reduction in the costs of the device and increases the imaging speed allowing for improved diagnosis.
To accomplish this, researchers started with a self-developed top-of-the-end optoacoustic scanner with 512 sensors, which produced the highest-quality images. Next, they discarded most of the sensors, leaving between 32 and 128 sensors.
This had a detrimental effect on the resulting image quality. Due to insufficient data, different distortions appeared on the images. However, a previously trained neural network was able to correct for these distortions and could produce images closer in quality to the measurements obtained with the 512-sensor device. The scientists stated that other data sources could be used and enhanced similarly.
Scientists at the University of Central Florida’s Computer Vision Center designed and trained a computer how to detect tiny particles of lung cancer seen on CT scans. These were so small that radiologists were unable to identify them accurately. The AI system could identify 95% of the microtumors, while the radiologists could only identify 65% with their eyes.
This AI approach for tumour identification is similar to algorithms used in facial recognition software. It will scan thousands of faces, looking for a matching pattern. The University group was provided with more than 1000 CT scans supplied by the National Institutes of Health with the Mayo Clinic collaboration.
The software designed to identify cancer tumours used machine learning to ignore benign tissues, nerves, and other masses encountered in the CT scans while analysing the lung tissue.
With an always-increasing supply of electronic data being collected in the healthcare space, scientists realise new uses for the subfield of AI. Machine learning can improve medical care and patient outcomes. The analysis made by machine learning algorithms has contributed to advancements in plastic surgery.
Machine learning algorithms have been applied to historical data to evolve algorithms for increased knowledge acquisition. IBM’s Watson Health cognitive computing system has been working on healthcare applications related to plastic surgery. The IBM researchers designated five areas where machine learning could improve surgical efficiency and clinical outcomes:
The IBM researchers also expect a practical application of machine learning to improve surgical training. The IBM team is concentrating on measures that ensure surgeries are safe and their results have clinical relevance–while always remembering that computer-generated algorithms cannot yet replace the trained human eye.
The researchers also stated that the tools could not only aid in decision making, but they may also find patterns that could be more evident in minor data set analysis or anecdotal experience.
Machine learning has identified one of the common causes of dementia and stroke in the most widely used brain scan (CT) with more accuracy than current methods. This is small vessel disease (SVD), a common cause of stroke and dementia. Experts at the University of Edinburgh and Imperial College London have developed advanced AI software to detect and measure small vessel disease severity.
Testing showed that the software had an 85% accuracy in predicting the severity of SVD. As a result, the scientists assert that their technology can help physicians carry out the most beneficial treatment plans for patients, swiftly aiding emergency settings and predicting a patient’s likelihood of developing dementia.
AI has helped humans in many facets of life, and now it is becoming an aid to doctors, helping them identify ailments sooner and determine the best pathways to tackle diseases. AI performs best with larger data sets, and as the volume of data increases, the effectiveness of AI models will continue to improve.
The current generation of machine models uses specific images and data to solve defined problems. More abstract use of big data will be possible in the future, meaning that more extensive data sets of disorganised data will be combined, and high-quality computers (potentially quantum computers) will be allowed to make new inferences from those data sets.
For example, when multiple tests like blood pressure, pulse-ox, EKG, bloodwork, and other tests, including CT and MRI scans, are all combined, the models may see things that doctors did not piece together. This is when machine learning will take medicine to the next level, providing even more helpful information to doctors to help us live longer and healthier lives.
Disclaimer: The information provided in this article is solely the author’s opinion and not investment advice – it is provided for educational purposes only. By using this, you agree that the information does not constitute any investment or financial instructions. Do conduct your own research and reach out to financial advisors before making any investment decisions.
The author of this text, Jean Chalopin, is a global business leader with a background encompassing banking, biotech, and entertainment. Mr. Chalopin is Chairman of Deltec International Group, www.deltecbank.com.
The co-author of this text, Robin Trehan, has a bachelor’s degree in economics, a master’s in international business and finance, and an MBA in electronic business. Mr. Trehan is a Senior VP at Deltec International Group, www.deltecbank.com.
The views, thoughts, and opinions expressed in this text are solely the views of the authors, and do not necessarily reflect those of Deltec International Group, its subsidiaries, and/or its employees.
Thank you for your interest in the Deltec Initiatives Foundation. You can contact us using the form below.
Before sending a message to us, kindly note that we do not accept requests for funding. Thank you for your understanding.