Sorry, you need to enable JavaScript to visit this website.
Sheetal Sharma

Data science & AI for healthcare
Sheetal Sharma Group Manager, Digital & Analytics | Jul 09, 2020

The potential that data science and artificial intelligence (AI) hold to revolutionize the delivery of healthcare is undeniable. The healthcare sector is well positioned to take advantage of what these technologies have to offer since they store and manage tons of lifetime patient data. From diagnostics, interpretation of lab tests and scheduling appointments to personalizing care, finding cures to conditions, and creating new and innovative solutions to long-standing problems like COVID-19, AI and data science offer endless possibilities.

However, compared to other sectors, healthcare lags behind in technology adoption. We have seen how data-led technologies have helped transform the way other sectors deliver services to consumers, and there’s a need to learn from these examples and apply the lessons to the unique context of healthcare.

Healthcare requires below key attributes to be successful:

  1. Organization-wide data repositories
  2. Data governance and security
  3. Interoperability of data within and across health systems
  4. Data science capabilities
  5. Repeated reuse of data to improve decision-making and care

Also, in order to augment the potential of the data repositories and other data assets that health systems hold, we need a clear chit from governments and policy makers to deal with challenges in data governance. We also need to deal with restrictions on collection, curation, and storage of data, along with regulating the use of data assets.

Basic Understanding of AI and Data Science

Data science is an inter-disciplinary field that uses scientific methods, processes, algorithms, and systems to extract knowledge and insights from structural and unstructured data.

AI, a subset of data science, refers to computers that can learn from data and interact with the human world. The goal is to give machines human-like cognition, meaning that they can ‘think’, and recommend actions based on that thinking, and predict outcomes basis their learning.

AI can be characterized by four broad technologies

  • Natural language processing (NLP): Technology’s ability to analyze human language, extract meaning and sentiment, and reply intelligibly is transforming our communication with each other and with machines
  • Computer vision: At the heart of every self-driving car or number-plate recognition system, computer vision comprises the extraction of information from images
  • Machine learning: This consists of programs and tools that recognize patterns in data and make predictions based on those patterns – or that can learn from data.
  • Robotics: This covers machines’ physical navigation and interactions with the human world.

Key Factors– Now is the time for Potential Endowments in Data Science and AI

Below are the key shifts in the use of data:

  • In the digital economy, tons of structured and unstructured data are getting created daily, so more data repositories are available for analysis.
  • It’s now possible to store and process huge data sets economically– at around 10 percent of the cost of a decade ago.
  • Thanks to innovations in the design and use of processing power, complex machine learning can now interpret the data that is created, stored and processed.

It’s now possible to store and process huge data sets at only around 10 percent of the cost of a decade ago.

Such a transformation is now needed more than ever in health systems, which are facing increasingly complicated huddles. More patients are suffering from complex conditions with multiple causes. The world has changed with the occurrence of serious diseases like MERS, SARS, and now, COVID-19, where the diagnosis requires input from several specialties. These complexities cannot be easily modeled with existing tools. However, new data sets and processing technology offer new capabilities for understanding new patterns.

How Data Science and AI Can Help Remold the Healthcare Sector in a Better Way

All the healthcare systems are facing the same challenges: predicting and preventing the onset of avoidable diseases, determining the safest and most effective treatment option, and delivering cost-effective care. From risk detection, triage, and scheduling, through diagnosis, prescription, and treatment, to patient engagement and public health, data science can contribute to efforts to optimize resources, better engage patients and improve the quality and outcomes of care.

Key Healthcare use cases enabled by Data Science and AI

Enhancing diagnostic precision and efficiency- The diagnostic failure rates are still relatively high in spite of huge amounts of health data at hand. However, application of data science can increase the precision and efficiency of diagnostics. For e.g. - Deep learning algorithms have been used to read imaging data (such as x-rays, CT scans, etc.), and analyze the same, checking the given results against extensive database of clinical reports and laboratory studies.

Identify, track, and forecast viral outbreaks through AI- The better we can track the virus, the better we can fight it. By analyzing news reports, social media platforms, and government documents, AI can learn to detect an outbreak.

Revamping clinical performance through actionable insights- Data science and predictive analytics are a valuable tool which can help healthcare providers optimize the way hospital operations are managed. Data science and machine learning can be used by healthcare providers to optimize the clinic staff scheduling and reduce the wait times, manage supplies and accounting, and even build efficient action programs for epidemics, such as seasonal flu outbreaks.

Monitoring and avoiding health problems using wearable data- Due to advances in technology, we can now collect real-time data such as heart rate, sleep patterns, blood glucose, stress levels, and even brain activity. Equipped with such health data, scientists are pushing the boundaries in health monitoring. Machine learning algorithms can be used to detect and track more common conditions, like heart or respiratory diseases. By collecting and analyzing heart rate and breathing patterns, technology can detect the slightest changes in the patient’s health indicators and predict possible disorders.

Reducing the risk of irregular prescription medicine- Self-learning AI based systems can check all prescriptions against similar cases in the database and inform the doctor when the prescription contains any deviations from the typical treatment plan.

Helping cure serious diseases (i.e. cancer, ebola, etc.) by enhancing pharmaceutical research- The cancer medication market is effectively served by the use of data science. Machine learning algorithms have been leveraged to analyze biological samples of thousands of patients to help in the development of drugs and medication which can detect and trigger the natural death of cancer cells in patients.

Moving to a Data-enabled Healthcare Journey

From foreseeing treatment outcomes, to curing cancer, and optimizing patient care, data science healthcare in the healthcare sector has proven to be an invaluable contributor to the future of the industry. Following are a few key factors to drive a data science healthcare transformation and data enabled healthcare journey:

  • Using/reusing data by exploration/pursuance of models of patient consent and also supporting dedicated multidisciplinary environments and centers of excellence
  • Investing in applications of data science and machine learning along with AI skills
  • Communicating and building the value of data integration, analytics, and data governance
  • Driving balanced regulations to protect patients’ rights while allowing health systems to innovate and improve
  • Advocating or enforcing organizations to adopt electronic health records (EHRs) to prescribed specifications using a range of incentives
  • Encouraging common standards to support interoperability and open APIs

While data science provides tools and methods to extract real value from unstructured patient information, it also eventually contributes to making healthcare more efficient, accessible, and personalized. The number of healthcare institutions making data-driven decisions continues to gain credence with more prominent data governance practices and increase slowly but steadily on other fronts as well.