Artificial intelligence (AI) can be defined as a branch of computer science that deals with developing devices or software that are capable of processing a set of information just like humans would; to think and learn from it. AI in medicine refers to the use of artificial intelligence technology/automated processes in the diagnosis and treatment of patients who require care.
In recent years, many advancements have been made in this sector that make it capable of analyzing medical data and playing an impactful role in medicine – now and in the future. All of these advances have been made possible due to the focus on imitating human thought processes on the machines being developed. The involvement of AI technology is seeming like a necessity in a time where the amount of data being gathered and processed is ever-growing. This is not to pit man versus machine, but to establish a more reliable working process that involves man and machine. The development of medical artificial intelligence has been related to the development of AI programs intended to help the clinician in the formulation of a diagnosis, the making of therapeutic decisions and the prediction of outcome.
The ultimate goal is not to replace physicians, it’s instead to provide all clinicians with tools to make faster, more accurate imaging diagnoses and treatment plans — something that could one day benefit patients across the world.
127 our experts are availableright now
The ability to establish an accurate and timely explanation of a patient’s health problem is a key aspect of healthcare yet it has been reported that diagnostic errors contribute to approximately 10 percent of patient deaths, and medical record reviews suggest that they account for 6 to 17 percent of adverse events in hospitals.
The need to reduce the amount of human errors in therapeutic and diagnostic areas can be massively aided by the involvement of deep learning machines as the advancements in this area continue to expand. An AI system can assist physicians by providing up-to-date medical information from journals, textbooks and clinical in order to give the required diagnosis to patients.
How do AI systems work?
Machines are getting better and better at processing large loads of complex medical data to return outputs that make the work of doctors much easier than it currently is. AI systems that are taking up roles in the world of medicine are expected to accurately learn from the data provided to them and whether or not they successfully carry out this responsibility is down to the algorithms that their functionality is based on. There are many popular techniques used to develop efficient AI-based medical machines; below are some of these techniques: Fuzzy Logic: A technique for dealing with sources of imprecision and uncertainty, it handles fuzzy concepts, inexact information and approximate reasoning in expert systems.
Based on the idea of partial truth, where information is non-numerical or imprecise, the truth value of this logic system can vary between the ranges of true (1) or false (0).
The term fuzzy logic was introduced with the proposal of fuzzy set theory by Lotfi Zadeh in 1965.15 Artificial Neural Networks (ANNs): One of the main tools used in machine learning, ANNs are brain-inspired systems which are developed with the goal of replicating how we humans think. They consist of input, output, and hidden layers that process data and find patterns that are otherwise far too complex for humans to notice.
Neural networks adjust their hidden layers of neurons if the outcome isn’t the desired one by the user, this technique called ‘backpropagation’ has made ANNs vital role players in artificial intelligence.
Applications of AI in MedicineEven though AI in medicine is generally still a work in progress, there are functional technologies that have been implemented and seen as a step forward in what awaits in the future structure of healthcare, below are some of those technologies:Chatbots: Applications that use speech recognition to receive information from patients about their symptoms and give back diagnosis or prevention measures based on their set algorithms. An example of such chatbots is Babylon, developed by Babylon Health, a UK based health subscription service. It’s used for prevention and diagnosis of diseases by first receiving information regarding the patients’ symptoms through its speech recognition feature, then comparing the symptoms to a database of diseases to narrow down what the cause is.
This application is not to be considered a dedicated substitute for actual physicians but as an option to find solutions on what the next move of a patient should be; whether it is to go consult a doctor in the case of something serious or just to get over the shelf treatments for minor issues that are common enough not to require a visit to a hospital.
Enlitic
A start-up system that applies deep learning to find an underlying pattern, without human intervention, in the huge set of medical data that is provided to it. It works with datasets such as radiology and pathology images, blood tests, EKG’s, genomics, patient histories, and electronic health records. The insights and final decisions come about after systematic comparison against billions of clinical cases from various data sources.
Enlitic technology is said to be able to interpret a given medical data image within a fraction of milliseconds; which is about 10,000 times faster than the average radiologists.8 As if the tremendous speed of action isn’t impressive enough, in a test to accurately classify malignant tumors, Enlitic performed 50% better than a panel of radiology experts.8 This system makes the work of physicians much easier as was seen in one internal study where radiologists where functioning at a rate 21% faster than usual and with less false positives(wrong diagnosis) with the help of Enlitic models.
Proscia
A platform for digital pathology that is used by thousands of researchers and pathologists to access data beyond their physical reach of a microscope and share their findings with others.
It’s software aims to perfect cancer diagnosis and is capable of high quality image analysis, extracting data from tissue with its efficient quantification algorithms letting it learn from millions of slides to recognize cancer’s patterns faster and more consistently than normal.
Proscia offers a lab that is accessible from any location via the cloud. It is a platform that offers global access to digital pathological content gained from a seamless collaboration with experts all around the world.
Insilico Medicine
A company that uses powerful, high-performance AI-based computer solutions to discover and repurpose drugs for various diseases including cancer, fibrosis, age-related diseases, and neurodegenerative diseases.
It connects genomic, metabolomic, proteomic, and other clinical databases to analyze and find pathologically activated pathways which aid the process of developing actionable targets.
Techniques such as Generative Adversarial Networks and Reinforcement Learning (GAN-RL) and ‘in silico Pathway Activation Network Decomposition Analysis (iPANDA) are used for the functioning of their state-of-the-art technological innovations that have given hope for bigger and better advancements in the world of medicine.
DXplain
A decision support system developed at the Laboratory of Computer Science at the Massachusetts General Hospital (MGH), DXplain is used as both an electronical medical textbook and a medical reference system. Its case reference and analysis mode is powered by its knowledge base that currently includes over 2400 diseases and over 5000 clinical findings. As a decision support system, it accepts a set of clinical data consisting of signs, symptoms, laboratory data and the likes to give an output of an ordered list of diagnoses which could justify the existence of the clinical manifestation being studied. Justifications as to why each of the resulting diagnoses have been returned are also given by DXplain. DXplain goes one step further in suggesting what type of additional clinical data should be added in case a narrower list of diagnoses is required by the practitioner. Access to DXplain via the Internet is provided only after executing a license with MGH.
Conclusion
Artificial Intelligence is becoming highly involved in our day today lives and medicine is one of the areas where AI has been behind when compared to other fields, but lately researches have increased in number and AI use in medicine is being practiced in some areas. The increase in growth in AI uses in medicine in the future the use of AI technology will revolutionize how medicine is practiced.
