Intelligent Algorithms & Healthcare


Healthcare industry is generating enormous amounts of structured and unstructured data. Finding innovative insights from data is of particular interest to Providers to incentivize predictive and preventive health management. This leads to healthcare industry focus on breaking data silos and leverage intelligent algorithms backed with advance analytics to create value. New age machine learning algorithms including natural language processing, pattern recognition, and deep learning are enabling better healthcare. I would like to bring up the following developments,

  • Algorithms for Cybersecurity:  Patient privacy and Cybersecurity are key focus areas for every healthcare provider. Sophisticated algorithms are aiding human skills with an ability to patrol security perimeters with more sensitivity and responsiveness. Algorithms identify patterns of normal usage and alert or flag events that are out of the ordinary by calculating a risk score for specific events as they happen based on the similarity or not to the normal behavior observed for the user performing the specific events. Supervise machine learning can cull things out that are less risky or classify them in terms of making multiple categories, like in terms of malware families. Getting Healthcare industry is in an early stages of implementing cognitive technologies ensuring security and reduce the rising threat of ransomware.
  • Algorithms for Unstructured Data (EHR to MRI Data): Extracting usable meaning from voice recordings of patient interactions, PDF images of faxed lab reports, and free-text HER are critical to an effective healthcare provision. This is where algorithms like natural language processing (NLP) can turn images of text into editable documents, extract semantic meaning from those documents, or process search queries written in plain text to return accurate results.
  • Algorithms for Clinical Support: Another important focus area is an ability extract meaning from large volumes of free text for better clinical decision support. Algorithms are helping ranging from precision medicine techniques to augment physician-guided diagnosis. Identifying and addressing risks quickly can significantly improve outcomes for patients with number of serious conditions, both clinical and behavioral. The silo nature of data organization and analysis limits the insights that doesn’t tell us a great deal about whether or not the patient has actually gotten better as a result of accessing that care. Semi-supervised and unsupervised machine learning algorithms like Clustering & Dimension Reduction can help improve the processes and dig deeper into that data and all the other variables that impact an individual’s life.
  • Algorithms for Pathology and Imaging: Healthcare organizations in improving patient outcomes is relying on Improved imaging analytics and pathology. Machine learning can supplement the skills of human radiologists by identifying subtler changes in imaging scans more quickly, potentially leading to earlier and more accurate diagnoses. Google research published during March 2017 – a new approach to imaging analytics driven by machine learning algorithms can identify metastasized breast cancer with sensitivity rates that exceed other automated methods and even rival human pathologists, is one of the best examples.

21st Century Healthcare industry has a tremendous opportunity to seize with smart usage of data science breakthroughs and evolution of intelligent machine learning algorithms.

More data on the way, preparing for “Decision-Easy Data Visualization”


“Seeing is believing.”; “A chart worth thousand numbers/words.”

I was thinking on the striking relevance of data visualization in today’s sphere of Big Data and Hyper-Connected world. As oceans of data available and zetta bytes of data coming in (44 zettabytes by 2020 i.e. 6 to 7 stacks from Earth to Moon), data scientists and analytics engines are doing more than satisfactory job in handling Volume, Velocity, Veracity and Variety of big data. While leapfrog advancement in analytics driving predictability and prescriptive insights, human touch is inevitable in making decisions. To get insights into action, humans have to view, interpret and validate insights to get into actions. Hence data visualization plays a significantly important role in enabling decision-easy representation of insights. In view of visuals and graphics that are processed 60,000 times faster by the brain than other means of data, here are few anecdotes (deviating purposefully from business context for ease of reading) that substantiate importance of data visualization.

  • Data visual on what the average American is doing with their time at any hour of the day, then question why so many people are still in bed at 8 a.m?
  • Bloomberg’s interactive index for ranking of the world’s richest people, which is a dynamic measure of the world’s wealthiest based on market changes
  • Showing how the number of disease cases have plummeted as a result of widespread vaccine usage
  • Seeing on US map how perceptions are changed on gay marriage. Is it really like as the country goes, so goes the Supreme Court?

A clear understanding of what any data set is representing is key to visualization. More the data more the dimensions to handle. Data is no longer just a numeric. Recent patterns of big data constitute text messages, pictures, videos, virtual copies that represent physical world – factory layouts, processes and supply chains, voice etc. Channeling any type of data and pattern to a visual is a scientific process intermingled with intellectual horsepower. The success of visualization is a direct measure of how fast the viewer connect with visual and move on to next step in decision making in non-ambiguous manner, which is what I am calling as “Decision-Easy” frame of mind. Choosing the right type of visualisation depends on what needs to be shown (comparison, distribution, composition, or relationship), how much detail the viewer needs, and what information the viewer needs in order to be successful. With this context, I have provided details with the help of the diagram above on depth and breadth of data visualization techniques and their applicability in real life scenarios. Leveraged Dr. Andrew Abela chart chooser.

Inventory Data Visualtion

Shown above is the summary of most of the available data visualisation techniques. Although the variety of data visualisation options may feel overwhelming, choosing a right representation that clearly comprehend data and insights will be a game changer.