Artificial Intelligence in Healthcare
Dr. Amit Ray, explains how artificial intelligence is changing the landscape of healthcare and modern personalized precision medicine. He also explains the current trends, scopes and concerns of AI in healthcare. Artificial intelligence (AI) quickly became an exponential technology and revolutionizing every aspect of life. The core of artificial intelligence is the ability to learn from data and the ability to adapt in changing situation and environment with high precision, accuracy and speed.
With the increasing availability of healthcare data and rapid progress of machine learning algorithms and analysis techniques AI is gradually enabling doctors for better diagnosis, disease surveillance, facilitating early detection, uncovering novel treatments, and creating an era of truly personalized medicine. Artificial intelligence in healthcare is going play a significant role in solving the issues like drug-interaction, false alarms, over-diagnosis, over-treatment.
There is also growing concern about how tech giants are making profit by selling personal data, including medical information. However, AI with new technologies of IoT and Blockchain has tremendous scope for better medical treatment with data security. Earlier I discussed nine key challenges of Artificial Intelligence in healthcare.
What is Artificial Intelligence?
Artificial intelligence is defined as the branch of science and technology that concerned with the study of software and hardware to provide machines the ability to learn insights from data and environment, and the ability to adapt in changing situation with high precision, accuracy and speed.
Generally, AI systems include machine learning algorithms for structured data, such as the classical support vector machine and neural network, and the modern deep learning, reinforcement learning as well as natural language processing for unstructured data. AI has already exceeded human performance in many areas such as Chess, Go Game, visual tasks, image recognition and voice recognition, voice based chatbots for customer services.
Primary Areas of AI in Healthcare Services:
The main areas of AI applications in healthcare are; providing personalized precision medicine, analysis and interpretation of radiology images, automated diagnosis, prescription preparation, clinical workflow monitoring, patient monitoring and care, discovery of new drugs, predicting the impact of gene edits, treatment protocol development. Artificial intelligence are used mostly for the diagnosis of cancer, nervous system disease and cardiovascular disease. All three diseases are leading causes of death; therefore, early diagnoses of these three diseases are crucial to prevent the deterioration of patients’ health status[1].
Artificial Intelligence in Radiology
Artificial intelligence in radiology has an enormous scope. With artificial intelligence it is possible to analyze and interpret large amounts of radiological images efficiently. Late detection of disease significantly increases treatment costs and reduces survival rates. Often visual human interpretation of an isolated image are time-consuming, difficult and expensive. Recently, AI for radiology uses deep learning, reinforcement learning, and other machine learning algorithms to systematically assess x-ray, CT Scan and MRI images of and instantly provide detailed reports on their findings.
AI algorithms read medical images like a radiologist. They identify the hidden patterns in the image and relate them with medical data. The AI systems are trained using vast numbers of images like CT scans, magnetic resonance imaging (MRI), ultrasound or nuclear imaging.
Studies have shown that computer aided screening can decrease false negatives by ~45% [5]. The AI machines reading radiology studies correctly, reaching around 95 percent accuracy [6]. The AI tools can rapidly review a number of images, prior images, patient history and other medical data, and then extract the most meaningful insights, which can then be verified by the radiologist.
AI in Precision Medicine
Precision medicine incorporate each individual’s DNA, microbiome, biochemistry, and lifestyle into treatment plans. As defined by the National Institute of Health (NIH), precision medicine is an emerging approach for disease treatment and prevention that takes into account individual variability in genes, environment, and lifestyle.
Big data is a major driver in the development of precision medicine. In other words, precision medicine offers a path to help people to recover from illness faster – and stay healthy longer.
Currently, “cognitive overload” is a real phenomenon for physicians. During diagnosis considering genomic data is not possible. However, with the advent of gene sequencing and improved biological understanding AI based systems will help to target and develop personalized therapies, drugs and treatment policies.
Artificial intelligence in precision medicine combines genetic, proteomic, structural, and computational methods to proceed from patient-based systems data, such as genome-wide association studies, to functional complexes, to pathways, and ultimately to predictive networks.
AI for Diagnosis and Monitoring Health Data
Today, one of the principal problems facing physicians and clinicians in general is the overload of huge patient information to examine through. With the advent of electronic medical records (EMRs) there is rapid accumulation of electronic. This includes lab reports, pathology reports, waveforms, radiology imaging data, exam and procedure reports. There is also data from admission, discharge and transfer (ADT), hospital information system (HIS) and billing software.
In the coming years there will be a further data explosion with the use of bidirectional patient portals, where patients can upload their own data and images from their smartphones and other health devices like home monitoring test results, and activity tracking from apps, wearables and the evolving Internet of things (IoT) to aid in keeping patients healthy. Gradually it is becoming very difficult or impossible to go through the large volumes of data to pick out what is clinically relevant or actionable. This is where artificial intelligence will play a crucial role in the next couple years. AI will quickly scrutinize through massive amounts of big health data and offer immediate clinical decision support system for appropriate action to make a diagnosis or even offer differential diagnoses.
AI with Blockchain Technology to Share Medical Data
Blockchain is an open distributed ledger that can record transactions between two parties efficiently and in the verifiable and permanent way. Blockchain with AI will help to develop algorithms that run in a decentralized and distributed manner that processes that data.
There has been an enormous development in collecting data from personal health monitoring devices and mobile apps, from electronic health records (EHR) in clinical settings and, to a lesser extent, from IoT designed to assist with medical procedures and hospital operations. Blockchain can secure the data input for AI, make it possible to observe every step taking by an AI for its learning and decision making process and create data sharing places, vastly reducing the cost of AI. The use of blockchain technology prevents data leakage, reselling, data ownership and data usage rights. AI with Blockchain is perfect combination for clinical workflow monitoring.
Machine Learning Algorithms for Healthcare Systems
The primary machine learning algorithms used for AI based healthcare systems are include linear regression, logistic regression, naive Bayes, decision tree, nearest neighbour, random forest, discriminant analysis, support vector machine (SVM) and neural network. However, among these support vector machine (SVM) and neural network are most popular and effective[1]. Currently, deep learning and deep reinforcement algorithms are used for analysis and interpretation of radiological images like CT scan, MRI and nuclear images.
Summary:
AI leverages deep reinforcement learning algorithms to overcome the obstacles inherent in large data sets and unstructured data in medical setup. In clinical settings, AI based systems functions as an assistant that helps clinicians work more efficiently and make more accurate diagnosis, which helps improve the productivity of healthcare professionals and provide better experiences for the patients.
Appropriate protection of patient‘s data confidentiality is a central issue yet to be resolved. Deep AI research in healthcare will propel our understanding of diseases — their origins and mechanisms, and opportunities for prevention and treatment more efficiently and cost effectively.
Source
Compassionate Artificial Superintelligene AI-5.0, Dr. Amit Ray, 2018
References:
- Artificial intelligence in healthcare: past, present and future, BMJ, 2017
- Artificial intelligence powers digital medicine, Nature, 2018
- AI diagnostics need attention, Nature, 2018
- AI Will Change Radiology, but It Won’t Replace Radiologists, Harvard Business Review, March 2018
- Evaluation of a Real-time Interactive Pulmonary Nodule Analysis System on Chest Digital Radiographic Images van Beek, Edwin J.R. et al. Academic Radiology , Volume 15 , Issue 5 , 571 – 575
- How Artificial Intelligence Will Change Medical Imaging Dave Fornell, 2017
- Phronesis of AI in radiology: Superhuman meets natural stupidity Judy Gichoya et al.
- A New Initiative on Precision Medicine, The New England Journal of Medicine, Francis S. Collins, Harold Varmus. February 26, 2015