Debates are raging on social media regarding explainable AI in healthcare. Geoffrey Hinton, one of the ‘godfathers of AI’ recently tweeted – “Suppose you have cancer and you have to choose between a black box AI surgeon that cannot explain how it works but has a 90% cure rate and a human surgeon with an 80% cure rate.Continue reading “Explainable artificial intelligence in healthcare by Dr. Johnson Thomas, MD, FACE, @JohnsonThomasMD”
In recent years, there has been an amplified focus on the use of artificial intelligence (AI) in various domains to resolve complex issues. Likewise, the adoption of artificial intelligence (AI) in healthcare is growing while radically changing the face of health- care delivery.Continue reading “Use of Artificial Intelligence in Healthcare Delivery by Dr. Sandeep Reddy, @docsunny50”
Has the cart been placed in front of the horse? The case of AI medical software regulation in developing countries.
Medical software is defined as the use of software for medical purposes. The uptake of medical software in healthcare has increased in line with increased application computation in healthcare delivery. Examples of medical software include software used in bedside monitors, MRIs, PACs, radiation therapy software, infusion pump rate devices, smartphone-based health applications. Etc.
The progress in research due to AI is facilitating Machine Learning (ML) techniques with potential to reduce costs, improve results, and increase patient satisfaction. However, applying ML techniques to create AI models is only a small portion in the overall realization of a clinical solution. Let me illustrate this with an example. Consider two cases – identifying a person from your personal photos vs. identifying a lesion in MR images of the brain. Both of these can be implemented using standard AI techniques but there would be a considerable variance in how quickly they could be deployed in use by the end-user.
So, what does it take to build a real-world AI-enabled clinical solution? First, let me draw attention to the principle challenges that exist today – the acquisition and aggregation of data. In the world of AI, it is often believed that one who has access to data is the king. But, in my view, one who has access to curated, annotated and consent managed data is the emperor of the world!
Unlike the large collections of public datasets that are available for creating general-purpose applications such as ImageNet, CIFAR etc., there is a significant shortage of high-quality medical datasets of specific clinical conditions. A vast majority of medical data is unstructured and dispersed across various systems, which makes data acquisition, extremely complex. Furthermore, region specific data residency and privacy laws also limit the sharing of patient data.
Once the problem of data acquisition is resolved, we are then confronted with the issues of quality and diversity. For example, to develop a lung-abnormality detection application, thousands of CXR images for each of the abnormalities – both in isolation and combination are required. Searching the hospital image archive for specific abnormalities in CXR images is highly error prone due to amount of manual effort required. In addition, for many clinical problems, a single data point is inadequate and needs more elements from the longitudinal record of the patient such as imaging priors, radiology reports, genomic data, lab reports etc.
Before the model creation, the data scientist has to spend a considerable amount of time and effort on the pre-processing steps. For example, to test the condition of the cells in Pathology, pre-processing steps like detection and segmentation of nuclei and glands are required, which themselves demand a highly accurate AI model. Although Deep Learning (DL) models are good at automatically extracting the features, a number of traditional features (like pathomics in pathology) improve the overall accuracy of a classifier. The challenging part is of course to find out which features are relevant and influential on the final prediction model.
Here comes the final and most appealing part of AI – the model development. It is only through close consultation with clinical experts that a data scientist can understand the problem statement for building the AI model. The more adept the data scientist is at choosing the analytical method, the higher will be accuracy of the model. Additionally, choosing a ML vs. DL technique holds significant importance in healthcare since the solution needs to justify the end result.
At Philips, we think of AI-enabled solutions as an extension of human capabilities, rather than a replacement. Any AI-enabled solution needs to demonstrate that it can augment the healthcare provider’s expertise in a meaningful and accurate manner. As the model is trained on a limited dataset, it is critical to be aware of its limitations when tested on datasets from diverse regions.
Despite the progress on the interoperability front, there are very few common representations of clinical pathways in healthcare. An AI model can’t be implemented in a hospital using the same parameters that were used to design it in the lab. Thus, it’s imperative to integrate it seamlessly with the existing clinical workflow of hospitals.
To ease the deployment of AI-enabled solutions in healthcare while keeping the aspects of hosting, performance and security in mind, we decided to build the HealthSuite Insights in Philips based on inputs from the data science and clinical communities.
Here are a few recommendations from us, which can simplify the overall process of creating AI-enabled clinical solutions. Firstly, churn out a cross-functional team of data scientists, clinicians, product managers, application specialists, software engineers, user-experience experts, DevOps and Q-and-R engineers. Then, study the clinical problem statement and role out the limitations of the existing solution. Further, establish clarity on the intended use of the proposed solution and clinical workflow integration. Institute a comprehensive data management strategy including acquisition and curation and acquire data from geographically distributed institutions.
Spend quality time in real-time testing to determine the overall accuracy of the model and focus on matters of privacy including consent management, security and regulatory. To eliminate issues pertaining to scalability, adopt an ML platform that can accelerate the development and deployment phases. Therefore, a comprehensive validation strategy that considers the intended use of the solution, interpretation variabilities, data and workflow diversity can help us confront the challenges and unlock the immense potential of AI in healthcare.
AI is an emerging technology. The healthcare institutions which will invest in open and interoperable standards for data management and clinical workflows today, will pioneer the implementation of AI-enabled solutions tomorrow. This will not only lead to better outcomes for patients but also significantly simplify the overall healthcare system.
In conclusion, I would like to quote an ancient Indian Sanskrit verse that keeps me personally motivated as we undertake this revolutionary journey:
Natural Language Processing: Giving doctors the freedom to write what they want
Electronic health records (EHRs) have clearly emerged as an innovative technology to facilitate the transition. However despite of the advancements, EHRs have not been able to achieve credible benefits in areas of population health management, health information exchange, patient care coordination and clinical analytics.
One of the biggest barriers in achieving success with EHRs has been the disparate forms of data which are difficult to aggregate and analyze. Doctors feel comfortable writing notes along with the flow of their clinical thoughts, however EHRs are not designed to capture medical information in a doctor’s natural language. This inability many a times leads to poor EHR usability. As a result, a lot of valuable information is left out from the ambit of analysis. With the advent of newer technologies, now it may be possible to plug such gaps. NLP (natural language processing) is one such technology which providers are now adopting with an anticipation to improve clinical outcomes and for the simplification of the daunting task of data entry in a computer.
Clinical data is not consistent, making analysis difficult
An EHR captures data in primarily four ways:
Clinical Data is directly entered in pre-structured templates
- Scanned documents are uploaded in the system
- Text reports are transcribed by speech recognition technology or by dictation and manual data entry.
- Data is purged into an EHR by interfacing it with other information systems like laboratory systems, radiology systems, or monitoring devices.
Clinical data is usually presented in a structured or unstructured format. Selective choices for capturing data in the form of templates like physician order sets, drop down menus, check boxes etc constitute structured data. Aggregation, analysis and reporting from structured data is easier but doesn’t provide an individualized, customized identity to an EHR. On the other hand, unstructured data constitutes free text narratives and clinical notes i.e doctor’s notes, patient encounters, patient health records etc and enable the physicians and patients to get their observations, complaints and concepts recorded in their own parlance. The unstructured data is a rich source of information about a patient’s health but it’s a challenge to transform it into structured and analyzable data that can be used for improving care outcomes. This challenge can be overcome with the technology of natural language processing.
Unstructured clinical notes are a mine of golden data; the wait to explore them ends with NLP
NLP is a data science based technology that can extract data from free text. NLP can be used by clinicians to convert medical notes into formats which are structured and standardized. Auto-processing of textual data can help providers in making use of clinical documentation data for a variety of purposes including but not limited to:
Improving communications between healthcare teams and thus help improve outcomes
- Reduce overhead costs of clinical documentation
- Improve revenues by automation of the coding and documentation
Computers can be given the ability to infer the intended meaning of words, thus enabling them to identify trends and patterns in huge datasets.
NLP can change the course of the way chronic diseases are managed:
One of the most promising area for exploring use cases of NLP in healthcare includes predictive analytics and risk scoring. Carefully deployed AI tools can be used for risk stratification and determination of hotspots in chronic diseases.
NLP can be used to tag socioeconomic terms hidden in free text notes to identify the social determinants of health. This can be augmented with machine learning to develop risk scores by proactive identification of trends from clinical and social data, laboratory reports, diagnoses etc. It is possible to create algorithms and train them on clinical record data to identify disease symptoms accurately.
Clinical records are a rich source of information regarding the symptoms of many diseases. Grouping of such similar symptoms can help in syndrome identification on the basis of disease presentation. As a result, it may be possible to unearth clusters which may otherwise not be suspected. Routinely available information in electronic health records, such as demographic and geographical location data and primary care free-text clinical records should be leveraged while making use of such algorithms.
Why off the shelf NLP engines may not be what the doctors want:
While it sounds easy, healthcare free text data comes with its own challenges. Word sense ambiguity is perhaps one of the most challenging problems in the noise of free text clinical notes. Accurate translation of the structured patient information pertaining to medical procedures, symptoms, tests etc depends on the algorithm’s ability to assign correct interpretations to the relevant medical words. For example, the acronym RA can be used in different contexts with different meaning by doctors. RA can be interpreted as right atrium, right arm or rheumatoid arthritis depending on the case presentation and clinical context.
Disambiguating the senses of acronyms, symbols and words that are used in a doctor’s clinical notes can significantly ease the burden on human effort needed to develop more accurate systems. A data-driven approach which involves development of any algorithm that infers patterns should consist of a supervised and unsupervised learning phase to yield benefits. In supervised learning every data item of the training data is labeled with the correct answer. Unsupervised learning on the other hand is a process where the computer recognizes patterns automatically. The true potential of an NLP and machine learning algorithm can only be harnessed when the data is trained in the provider’s environment.
Word sense disambiguation based NLP pays a significant role in improved analytics and patient outcomes:
Word sense ambiguation based language processing ability of the computer for accurate mining of clinical documents can bridge the gaps in documentation and aid clinical decision support and clinical documentation improvement programs.
More insightful extraction of data is possible with a decreased ambiguity in clinical data. When the computer has the ability to infer the intended meaning of words, it can find useful patterns in heaps of data easily. IBM’s Watson Supercomputer technology is an apt example of how NLP can facilitate meaningful analytics, by identifying such patterns. IBM’s content analytics process is used for collection and analysis of structured and unstructured data, and its similarity analytics makes use of NLP and machine learning technology for analysis of a large number of variables in a patient’s medical history and present condition to identify patterns and draw a comparison with similar conditions and potential outcomes.
There is no doubt that word sense disambiguation enabled NLP technology can have a potentially huge on impact clinical data analytics with its superior ability to infer meanings of extracted data more accurately. Data analytics for improved patient outcomes is not the only benefit of this technology, it can also support accuracy of billing. With its ability to support clinical documentation improvement programs, it can also help in improving clinical workflows.
SymptomAI by “PredictDisease” is a healthcare analytics platform that is driven by artificial intelligence, NLP and machine learning to assist patients and primary care physicians by measuring the potential risk of a chronic disease that starts with minor symptoms. The platform leverages data from lifestyle activities, social media/website forums, scientific research papers, and family history, matching these with known signs/symptoms and other demographic characteristics for the early detection of the chronic disease. It takes into account, social and biologic determinants of health to predict the risk score. Visit us at www.predictdisease.com or write to us at email@example.com for more info.
. Auto Coding and NLP:
: Dooling, Julie A. “Advancing Technology Connects Transcription and Coding: The Developing Role of NLP, NLU, and CAC in HIM.” Journal of AHIMA 83, no.7 (July 2012): 52-53
: Goldberg, Michael. “IBM Makes New Health Care Push with Predictive Analytics, Process Management.” Data Informed. http://data-informed.com/ibm-makes-new-health-care-push-with-predictive-analytics-process-management/
This comment was picked up by a journalist of an online news site and reported as the headline of a news item covering the conference. However, the journalist also mentioned I hadn’t provided any details to qualify my comments. To my defence, the focus of my speech was not about this topic. Subsequently, in a conversation a week ago with the founder of an US-based AI health start-up, we discussed the multitude of opportunities in using AI enabled health services in developing countries and how few are aware of these opportunities. These circumstances and a reminder from the editor of this site about my promise to contribute an article has now led me to articulate the benefits of AI in health in the context of developing countries. Here, I use India as a typical developing country but many of the processes I discuss can potentially be used in other developing countries.
First, let us discuss some of the common issues that health services in developing countries face. A common grievance of health services is the lack of qualified workforce to treat and manage patients. Where health services have qualified personnel, they are overloaded with patients affecting the quality of the service they provide. The other common issue is the urban-rural maldistribution of qualified physicians. The preference of physicians to practice in urban health centres has led to a skewed distribution favouring urban centres and disadvantaging rural communities. In spite of government initiatives to push quality rural health services, the urban-rural divide is to stay. Another prominent issue is the variability in the quality of health services provided in different parts of the country and sometimes within the same region or city. This inconsistency is because of poor monitoring of health services by national accreditation bodies or poor compliance with quality standards by health services. Further, outbreaks of infectious diseases because of mainly environmental reasons has become alarmingly frequent in developing countries. Poor surveillance infrastructure means these outbreaks can progress to epidemics in a span of days. These aren’t the only health system concerns in developing countries but are areas that I think will benefit from the application of AI techniques.
India with its massive population of 1.35 billion (2018 population estimate) is in dire need of strong health infrastructure and government policy to service the nation’s health needs. India, just like many developing countries, has significant challenges in delivering this requirement. A combination of increasing burden from chronic diseases, a large ageing population, qualified personnel shortage, urban-rural divide, low government investment in health, inadequate health insurance coverage and variable quality of health service delivery have contributed to this state of affairs. However, the Government of India has been lately active in firming up the health policy and strengthening the health infrastructure. One of the major initiatives of the government was the release of the National Health Policy last year with an aim to reinvigorate the healthcare delivery in India by increasing health spending, establishing national quality standards, promoting evidence-based healthcare and introduction of digital health initiatives. With regards to the latter objective, the intention to set up a National Digital Health Authority and promote interoperable Electronic Health Record systems across India will create a strong foundation for digital health innovations to be applied. This digital platform will also provide opportunities for Foreign Direct Investment (FDI) and contribute to further growth of digital health in India. As digital health initiatives ramp up in India, opportunities for application of AI will also open up.
So how would AI applications help the Indian health system? Earlier, I discussed the healthcare delivery challenges developing countries face. The same difficulties apply to India too. AI systems driven by deep neural networks and computer vision have matched accuracy levels of human clinicians in interpreting radiological, fundoscopic and histopathological images. Intelligent agents are also being used to mine data and analyse electronic health records to assist clinicians in the medical diagnosis and predicting mortality of patients. Also, machine learning and natural language processing driven mobile applications are being used to communicate with patients and aid medication adherence, healthy lifestyles and schedule visits to doctors. Further, AI applications are being used in hospitals to predict the length of stay of patients and formulate treatment plans for them. All of these developments are detailed in several academic journals and the media. Application of these agents will have a profound effect on the Indian healthcare landscape, where a shortage of qualified specialists and diagnostic centres abound. While the AI systems may not be able to replicate all the capabilities of the medical specialists, it will in combination with telemedicine approaches be able to increase healthcare access for underserved communities and alleviate the burden of overstretched health services.
AI systems can also aid in the improvement of the quality of healthcare by reducing the variability of healthcare delivery and enabling evidence-based practice across the country. By incorporating government sanctioned and thoroughly evaluated AI applications in healthcare delivery, standardisation of healthcare delivery can be achieved. With inconsistency in healthcare delivery and non-evidence-based practices being common in India, roll out of authorised clinical decision support systems that run through machine learning processes will contribute to standardisation of healthcare delivery. Also, AI systems through ongoing analysis of ecological, biogeographical and public health data can alert authorities about outbreaks of infectious diseases and help contain the spread. For example, in recent years machine learning has been used to identify sources of outbreaks. During the Ebola outbreak in Africa, machine learning was used to analyse ecological data to determine the bat species harbouring Ebola virus and contain the spread of the disease. Thus, AI agents can also be used to strengthen India’s communicable disease surveillance infrastructure.
While the use of AI applications presents significant promise for the Indian healthcare system, one has also to be cognizant of the challenges in applying AI approaches. AI applications rely on a robust digital health foundation including ongoing access to electronic patient data and patient/population information management systems. With the Indian digital health infrastructure being nascent at best, widespread roll-out of AI applications can be a challenge. Also, with the low number of qualified health informaticians, machine learning trained data scientists and AI focused entities in India, there may be increased reliance on overseas companies to support the roll-out of AI applications. There are also issues like bias, lack of contextual reasoning and explainability problems that accompany AI applications. However, with advances in AI technology some these issues have now been addressed with number of solutions available.
To harness the benefit of AI approaches, the Indian government has to formulate a definitive AI strategy. A strategy that amongst many other things outlines the regulatory framework and implementation strategy for the roll-out of AI in India. The immense benefits that come through application of AI can be only be realised through the boldness and proactiveness of the Indian government. By pushing forward a national AI strategy and setting up an AI enabled healthcare delivery system, India can be a leading example for other countries as to how critical healthcare challenges can be addressed through AI approaches.
It has been roughly a year since Ashu and I started mfine. During the course of this journey, we studied the healthcare sector very closely and started interacting with people in the ecosystem. The more we learnt, the more we got passionate about healthcare and at times obsessed with it!Continue reading “mfine, the MVP of a “Cloud Clinic” by Prasad Kompalli, @pkompalli CEO & cofounder, mfine @mfinecare”
|Pic Credit: CreativeCommons.org|
Well, technology is here to rescue. It will be a grave injustice to talk about technology implementation in pharma R&D in just one article, and because AI and ML are the current buzzwords, I thought, maybe, we could specifically discuss the roleplays in streamlining and improving success rates of pharma R&D. Before we go any further, let us have a quick look at the drug discovery process and the typical timelines associated.
The article was first published on Ms. Manishree Bhattacharya’s LinkedIn Pulse Blog, here. Its been republished here with the authors’ permission.
[content title=”About Manishree Bhattacharya” icon=”fa-heart”]
[content title=”Latest Articles”]
AI & Machine Learning Terms
Artificial intelligence The development of computers capable of tasks that typically require human intelligence. A machine’s ability to make decisions and perform tasks that simulate human intelligence and behavior.
Machine learning Using example data or experience to refine how computers make predictions or perform a task. A facet of AI that focuses on algorithms, allowing machines to learn without being programmed and change when exposed to new data.
Deep learning A machine learning technique in which data is filtered through self-adjusting networks of math loosely inspired by neurons in the brain. The ability for machines to autonomously mimic human thought patterns through artificial neural networks composed of cascading layers of information.
Supervised learning Showing software labeled example data, such as photographs, to teach a computer what to do. A type of machine learning in which output datasets train the machine to generate the desired algorithms, like a teacher supervising a student; more common than unsupervised learning.
Unsupervised learning Learning without annotated examples, just from experience of data or the world—trivial for humans but not generally practical for machines. Yet. A type of machine learning algorithm used to draw inferences from datasets consisting of input data without labeled responses. The most common unsupervised learning method is cluster analysis.
Reinforcement learning Software that experiments with different actions to figure out how to maximize a virtual reward, such as scoring points in a game.
Artificial general intelligence As yet nonexistent software that displays a humanlike ability to adapt to different environments and tasks, and transfer knowledge between them.
Large-scale Machine Learning Design of learning algorithms, as well as scaling existing algorithms, to work with extremely large data sets.
Deep Learning Model composed of inputs such as image or audio and several hidden layers of sub-models that serve as input for the next layer and ultimately an output or activation function.
Natural Language Processing (NLP) Algorithms that process human language input and convert it into understandable representations. The ability for a program to recognize human communication as it is meant to be understood.
Collaborative Systems Models and algorithms to help develop autonomous systems that can work collaboratively with other systems and with humans.
Computer Vision (Image Analytics) The process of pulling relevant information from an image or sets of images for advanced classification and analysis.
Algorithmic Game Theory and Computational Social Choice Systems that address the economic and social computing dimensions of AI, such as how systems can handle potentially misaligned incentives, including self-interested human participants or firms and the automated AI-based agents representing them.
Soft Robotics (Robotic Process Automation – RPA) Automation of repetitive tasks and common processes such as IT, customer servicing and sales without the need to transform existing IT system maps.
Algorithms: A set of rules or instructions given to an AI, neural network, or other machines to help it learn on its own; classification, clustering, recommendation, and regression are four of the most popular types.
Artificial neural network (ANN): A learning model created to act like a human brain that solves tasks that are too difficult for traditional computer systems to solve.
Autonomic computing: A system’s capacity for adaptive self-management of its own resources for high-level computing functions without user input.
Chatbots: A chat robot (chatbot for short) that is designed to simulate a conversation with human users by communicating through text chats, voice commands, or both. They are a commonly used interface for computer programs that include AI capabilities.
Classification: Classification algorithms let machines assign a category to a data point based on training data.
Cluster analysis: A type of unsupervised learning used for exploratory data analysis to find hidden patterns or grouping in data; clusters are modeled with a measure of similarity defined by metrics such as Euclidean or probabilistic distance.
Clustering: Clustering algorithms let machines group data points or items into groups with similar characteristics.
Cognitive computing: A computerized model that mimics the way the human brain thinks. It involves self-learning through the use of data mining, natural language processing, and pattern recognition.
Convolutional neural network (CNN): A type of neural networks that identifies and makes sense of images.
Data mining: The examination of data sets to discover and mine patterns from that data that can be of further use.
Data science: An interdisciplinary field that combines scientific methods, systems, and processes from statistics, information science, and computer science to provide insight into phenomenon via either structured or unstructured data.
Decision tree: A tree and branch-based model used to map decisions and their possible consequences, similar to a flow chart.
Fluent: A type of condition that can change over time.
Game AI: A form of AI specific to gaming that uses an algorithm to replace randomness. It is a computational behavior used in non-player characters to generate human-like intelligence and reaction-based actions taken by the player.
Genetic algorithm: An evolutionary algorithm based on principles of genetics and natural selection that is used to find optimal or near-optimal solutions to difficult problems that would otherwise take decades to solve.
Heuristic search techniques: Support that narrows down the search for optimal solutions for a problem by eliminating options that are incorrect.
Knowledge engineering: Focuses on building knowledge-based systems, including all of the scientific, technical, and social aspects of it.
Logic programming: A type of programming paradigm in which computation is carried out based on the knowledge repository of facts and rules; LISP and Prolog are two logic programming languages used for AI programming.
Machine intelligence: An umbrella term that encompasses machine learning, deep learning, and classical learning algorithms.
Machine perception: The ability for a system to receive and interpret data from the outside world similarly to how humans use our senses. This is typically done with attached hardware, though software is also usable.
Recurrent neural network (RNN): A type of neural network that makes sense of sequential information and recognizes patterns, and creates outputs based on those calculations.
Swarm behavior: From the perspective of the mathematical modeler, it is an emergent behavior arising from simple rules that are followed by individuals and does not involve any central coordination.
|Automated communications||Also known as an interactive agent, or artificial conversational entity, these are computer programs which conduct a conversation via auditory or textual methods. For example, chatbots, mailbots.|
|Automated data analyst||AI solutions aimed at performing the job of data analysts and data scientists and bridging the gap between such roles and business imperatives. For example, these might include programs that are able to develop a deep understanding of customer preferences from data, identify high-risk customer groups and tailor interaction touch points in a manner personalised to such customers.|
|Automated operational and efficiency analyst||AI solutions targeted at increasing operational efficiency and reducing costs. These include AI programs and bots aimed at automating repetitive manual tasks such as identifying and correcting data and formatting mistakes, performing back office tasks and automating repetitive interactions with customers.|
|Automated research and information aggregation||Applications of AI that involve aggregating and processing large volumes of information on a topic so as to generate meaningful insights. For example, aggregating information from research papers or medical journals for diagnosis support, identifying online hoax, bad reporting and statistics, and identifying plagiarised publications.|
|Automated sales analyst||AI-powered digital analysts for sales and marketing decisions. These programs are able to test a range of scenarios using internal and external data to predict the impact of marketing strategies such as promotions and campaigns, simulate ‘what if’ scenarios against multiple hypotheses and perform root cause analyses against business results.|
|Business decision makers/influencers||A sub-set of participants in the survey who have identified themselves to be either in a decision making role or an influencing role in their current organisations. Some of the survey questions had been specifically targeted towards this group.|
|Decision support systems||Decision support systems (DSS) are a specific class of computerised information systems that support business and organisational decision-making activities.|
|Machine learning||Machine learning is concerned with computer programs that automatically improve their performance through experience.|
|Predictive analytics||Predictive analytics is an area of statistics that deals with extracting information from data and using it to predict trends and behaviour patterns–for example, sales forecasts, predicting customer churn and industrial|
|Robotics||Robotics deals with the design, construction, operation and use of robots, as well as computer systems for their control, sensory feedback and information processing. Environmental information such as imagery and sound are captured using a group of sensors and the same are processed using various computerised techniques for the robot to respond.|
|Virtual personal assistants||Virtual assistants use natural language processing (NLP) to match user text or voice input to executable commands. Many continually learn using AI techniques, including machine learning. For example, Apple’s Siri, Amazon’s Alexa, Google Now.|
|AI advisors||AI advisors are machines or systems that monitor employees’ progress and performance. They are responsible for the growth of the employee in the organisation and for the delivery of projects.|
|AI assistants||AI assistants are machines or systems or application programming interfaces ([APIs] a set of subroutine definitions, protocols and tools for building application software) that perform non-value adding services such as scheduling and email management.|
: pwc AI: https://www.pwc.com/ai
: AI: The Complete Guide:
The report, ‘Rewire for Growth,’ estimates that AI has the potential to increase India’s annual growth rate of gross value added (GVA) by 1.3 percentage points, lifting the country’s income by 15 percent in 2035. To avoid missing out on this opportunity, policy makers and business leaders must prepare for, and work toward, the AI revolution.
The era of AI has arrived. Established companies are moving far beyond experimentation. Money is flowing into AI technologies and applications at large companies. The number of patents filed on AI technologies in G20 countries has increased at a more than 26 percent compound annual growth rate since 2010. Funding for AI startups has been growing at a compound annual growth rate of almost 60 percent.
AI is a new factor of production that can augment labor productivity and innovation while driving growth in at least three important ways:
Mobilize Intelligent Automation
Automate complex, physicalworld tasks that require adaptability and agility.
Empower Existing Workforces
Complement and enhance the skills and abilities of workforces.
Let AI be a catalyst for broad structural transformation of the economy. Do things differently, do different things.
The report points out AI is expected to raise India’s annual growth rate by 1.3 percentage points—in a scenario of intelligent machines and humans working together to solve the country’s most difficult problems in 2035
AI TENDING TO INDIA’S HEALTH
India’s healthcare providers have embraced artificial intelligence, recognizing its significant value in better diagnostics with data intelligence and in improving patient experience with AI-powered solutions.
Take Manipal Hospitals, headquartered in Bengaluru, which is using IBM Watson for Oncology, a cognitive-computing platform, to help physicians identify personalized cancer care options across the country.
In cardiac care, Columbia Asia Hospitals in Bengaluru is using startup Cardiotrack’s AI algorithms to predict and diagnose cardiac diseases, disorders, and ailments.
And in eye care, Aravind Eye Hospital is working with Google to use AI in ophthalmology for diabetic retinopathy screening. Also, the government of Telangana is planning to use Microsoft Intelligent Network for Eyecare (MINE), an AI platform, to reduce avoidable blindness, which would make it the first state in India to deploy AI for eye care screening as part of the Rashtriya Bal Swasthya Karyakram program under the National Health Mission.
Accenture, for its part, has developed an AI-powered smartphone solution to help the visually impaired improve the way they experience the world around them and enhance their productivity in the workplace. The solution, called
Drishti, was initially developed and tested through a collaboration with the National Association for the Blind in India.
AI has the potential to have a broad-based disruptive impact on society, creating a variety of economic benefits. While some of these benefits can be measured, others, such as consumer convenience and time savings, are far more intangible in nature. Our analysis focuses on measuring the GVA impact of AI.
Read the press release here >>
Read the complete report here >>
House MD vs Doctor AI- Who will turn out to be the better diagnostician?
Can we put our data to better use – to improve healthcare on a scale unimaginable a decade or two ago?
Artificially intelligent systems in healthcare
AI systems feed on big data. Big data is nothing but a massive amount of data, the sheer size of which makes data analysis a challenge. AI systems are being developed to analyze and recognize meaningful patterns out of this complex data. Of particular significance is deep learning, a branch of AI which attempts to mimic the thinking part of the human brain. Several startups are now attempting to put AI, particularly deep learning to meaningful use in healthcare.
Diagnosing diseases: Clinical diagnosis is essentially a data problem, says founder of Enlitic, a machine learning startup focusing on data-driven medicine. They aim to revolutionize clinical diagnosis by helping physicians to automatically screen for specific diseases using their proprietary technology. Another startup, Deep Genomics, is approaching diagnosis from a genomic perspective. They capitalize on machine learning technology to link genetic variations to diseases. As a cherry on top, a study published as recent as last week reported that AI-enabled automated cancer detection perform as well as approaches that require costly clinician input.
“We think that its no longer necessary for humans to spend time reviewing text reports to determine if cancer is present or not”- author of the study, Shaun Grannis M.D., M.S.
Educating Patients: IBM is teaming up with the American Cancer Society to create an adviser for cancer patients, powered by IBM’s Watson health – health division of its AI brain. It would be designed to provide cancer patients or their caregivers with personalized guidance based on the patient’s particular disease stage and treatment. Watson would sift through countless digital sources like health websites and draw relevant and trustworthy information catering to individual needs.
Improving clinical trials: Many clinical trials fail because patients fail to take their medications. AiCure is addressing this issue using AI to monitor medication adherence thereby increasing trial success. They directly monitor patients using artificial intelligence on mobile devices via an app which collects real-time dosing data on a centralized and cloud-based platform. IBM Watson is trying to solve another issue facing clinical trials – enrolling patients. Watson would sift through clinical trial data at Mayo Clinic and in public databases, such as ClinicalTrials.gov. and match patients more accurately and consistently to clinical trial options.
Using Watson’s cognitive computing capabilities, Mayo Clinic can consistently offer more cutting-edge medical options to patients and conclude trials faster” – Mike Rhodin, senior vice president, IBM Watson Group.
Accelerating drug discovery: Drug discovery is a lengthy, complex, and costly process, rooted with a high degree of uncertainty that a drug will actually succeed. Several machine learning startups are trying to solve this long-standing predicament in pharmaceutical drug development. Berg Health combines systems biology with its proprietary artificial intelligence machine learning analytics program to bring down the time for cancer drug development to almost half. Atomwise’s drug discovery AI platform “learns like a human chemist” using deep learning algorithms and supercomputers.
“Simulating billions of virtual medicines to find potential commercial candidates in weeks? That’s what we call truly transformative.”- Matt Ocko, Managing Partner of Data Collective, a venture firm backing Atomwise
Reducing hospital costs: Readmissions are costly for hospitals as payers are reluctant to reimburse preventable readmissions. Hindsait, an artificial intelligence technology provider, helps identify patients who are nearing the end of their stay and suggests whether keeping them might in fact be more cost effective than risking readmission. It does so by applying AI based data analysis to large health datasets.
These and many more upcoming developments in cognitive computing in healthcare leaves us with the thought – will AI systems perform better diagnosis than out best diagnosticians. Google’s AI program, AlphaGo beating Go world champion and IBM’s Watson beating humans in Jeopardy seems to suggest so.
Note: Interestingly, there were many recent publications that influenced the direction of this article. Application of AI technologies in healthcare is growing with new collaborations being made at this very moment. Interesting times indeed for all those who influence and follow these developments.
A typical workflow for booking an appointment can go like this:
1. Patient calls (or visits) the hospital, and speaks to the person at the reception, at a specific department
2. The person looks up the available time slots, that a doctor is free and available in the clinic
3. Consults with the Patient on the best time possible for her appointment and then schedules the appointment
Now this three step process can either happen on a call, at the hospital reception or via a website provided by the hospital. But in real life, the appointment booking process for a patient might not be so straight forward. Here are some of the different scenarios that might occur:
1. Doctor is not available, asks her medical assistant to cancel all her appointments. Existing appointments need to be shifted to other doctors or rescheduled based on patient priority.
2. Patient calls at the last moment and asks for her appointments to be re-scheduled or cancelled
3. Patient does not show up for the appointment, and asks for a new appointment
4. During a clinic day, multiple new and urgent cases need to be seen by the physician, which delay the subsequent appointments
5. Scheduling of renal therapy patients or cancer therapy patients also needs supervised scheduling that is closely related to the patients’ care protocols and care plans
6. Scheduling based on urgency and emergency situations also changes the “scheduled” visits of a doctor
Considering these challenges in the daily working environment of a hospital, an AI-based scheduling solution can help the hospitals in providing an optimal use of resources. For instance a research from Indiana University  found using Artificial Intelligence in patient care can be cost effective and improve patient outcomes.
Consider for instance the following statistics of a Government Hospital in Rajasthan, India :
- Nearly 1.27 crore patients were registered at OPD in medical centres affiliated to medical colleges and
- 9.27 crore in state medical institutions in the year 2014-2015
- in the year 2015 around 35,000 patients per day were registered at the OPD at medical college-affiliated centres
The High number of patients (the 35,000 per day patients registered at the OPD at medical college-affiliated centres) and the resource scheduling scenarios, presents an apt usecase to implement an AI based Appointment Scheduling system.
While it not only present a challenge to manage the care of all the visiting patients, it also allows for the administration to ask; How many doctors, nurses and medical assistants should be scheduled to manage the care planning & scheduling requirements of each of these patients, visiting one or many departments of the hospital.
In addition to Patient Scheduling, AI based algorithms can be deployed in such settings  to help the hospital administration in optimising the time of their most important resources: Physicians, nurses and medical assistants.
|Handbook of Healthcare System Scheduling – Reference |
Additional Scenarios where the AI based resource scheduling systems in Healthcare  can be deployed are:
- Operating Theatre + Operating Team Scheduling
- Renal Dialysis Centers
- Radiology Diagnostic Facilities
- Medication Reminders Apps
- Acuity-based nurse assignment and patient scheduling in oncology clinics
- Care Plans based activity & event scheduling
- Procedure Scheduling
- Health Checkups Packages
Once an AI based solution has been implemented, the scheduling, rescheduling, planning, allocating and many other scenarios are handled by an AI based Scheduling Agent allowing for hospital administrators and physician scheduling managers to focus on treating the patients.
And Scheduling a patient appointment becomes an autonomous process:
A. Jane emails Dr. John to schedule an appointment for a followup visit. Jane receives a confirmation email regarding the appointment with Dr. John from his assistant Amy. A reminder is set in her calendar.
B. Jane, on the day of the appointment is unable to make it to the hospital and sends an email requesting for rescheduling her appointment to the next wednesday. Amy reviews, Dr. Johns schedule and responds to Jane with a confirmation of her re-scheduled appointment.
In the above example Amy is an AI assistant to the Physician, nurse or medical health professional. Or in fact it could be an assistant (Siri, cortana, or amy from x.ai etc) to the Patient.
What do you think, do share your thoughts?
Sundar Pichai, CEO, Google says we are moving from a mobile first world to an AI first world, quite fast.
- How to use AI to automatically schedule your appointments with x.ai – TechRepublic http://ow.ly/lSdJ300yH9w
- [1206.1678] A Distributed Optimized Patient Scheduling using Partial Information http://ow.ly/u3A0300yHu3
- Artificial Intelligence in Healthcare: A Smart Decision? | Health Standards http://ow.ly/IR3N300yIep
- Can computers save health care? IU research shows lower costs, better outcomes: IU News Room: Indiana University http://ow.ly/bPWs300yIs6
- Association for the Advancement of Artificial Intelligence http://ow.ly/4aoc300yIxY
- E-registration Facility Soon At SMS HospitaleHEALTH | EHEALTH http://ow.ly/njMx300yJgz
- Handbook of Healthcare System Scheduling – http://ow.ly/cvUn300yLql
- From Scheduling Meetings To Shopping Deals: 14 Early-Stage AI Assistants To Watch http://ow.ly/R9b7301lqjK
- Who will turn out to be the better diagnostician? #digitalhealth #ArtificialIntelligence https://t.co/TmzInbDlg5
- Robot Takes On Role Of Hospital Scheduling Nurse | Digital Trends http://ow.ly/QTAW100eEgR
- This is how the future of hospital operations resembles air traffic control – MedCity NewsMedCity News http://ow.ly/BJh1100eIdv
- Can Artificial Intelligence Help The Mentally Ill? https://t.co/e5NEnYOpAL #mentalhealth #AI
- On-line Appointment Sequencing and Scheduling – Brian Denton et al, http://ow.ly/RXXm300yLHX
- Artificial Intelligence Can Improve Healthcare | EMR and EHR http://ow.ly/MlBy302ur9Q
[content title=”About Manish Sharma” icon=”fa-heart”]
[content title=”Latest Articles”]
- TRIVENI: A remote patient monitoring solution via @msharmas – Part 1
- Health ID as Patient IDs unifier in India by Manish Sharma
- Top Healthcare & Digital Health Predictions for 2016
- Zen Clinicals: An Activity & Workflow based solution (1 of 3)
- 5D Delivery Framework
- RFID in Healthcare: Usecases from Hospitals
- 10 Solutions for the Healthcare IT Fringes