Category Archives: Technology and Informatics

Posts related to technology, gadgets, cloud, informatics, or just about anything that is/can be/plugs into a computer. Relationship to radiology optional.

Your Radiology AI Briefing – May 12, 2018

In this briefing:

  • Research: Machine learning algorithm predicts wait time for outpatient imaging.
  • Commercial brain imaging AI receives FDA clearance
  • Paul Chang shares insight on the future of AI
  • Dreyer and Allen publish their views on the radiology AI ecosystem
  • CB Insights publishes market research on Google’s increasing involvement in healthcare AI.

Radiology AI Briefing logo graphic

Stay up to speed in 2 minutes. Radiology AI Briefing is a semi-regular series of blog posts featuring hand-picked news stories and summaries on machine learning and data science.


Continue reading

Your Radiology AI Briefing – May 3, 2018

In this briefing:

  • RSNA launches a new AI journal.
  • ACR makes several moves to advance the use of artificial intelligence in the future of medical imaging.
  • Researchers publish in high impact journal the successful use of AI to detect breast density.
  • ACR and MICCAI sign agreement to advance AI in medical imaging.
  • Geisinger declares successful implementation of algorithm to improve time-to-diagnosis of intracranial hemorrhages 20-fold.

Radiology AI Briefing logo graphic

Stay up to speed in 2 minutes. Radiology AI Briefing is a semi-regular series of blog posts featuring hand-picked news stories and summaries on machine learning and data science.


Continue reading

A Simple Tool to Brainstorm AI for Your Radiology Practice

Use the AI Canvas.

Source: A Simple Tool to Start Making Decisions with the Help of AI

In a recent Harvard Business Review article, Ajay Agrawal and coauthors shared a simple tool to think about how an AI tool may be deployed.  Although the tool is no more complex than 6 boxes of free text, it does follow a number of best practices when thinking about general data and machine learning:

  1. Always define an end-goal – what’s the desired outcome?
  2. You should always make a hypothesis of what may drive this desired outcome.
  3. You should determine how to present the ML prediction in a way that drives action, not just the data itself.
  4. Your data acquisition strategy should include a feedback mechanism.

For example, this is how one might fill out the AI Canvas tool in a radiology use case:

  • Prediction: Predict whether a brian MRI for a cancer patient contains increasing or new hydrocephalus
  • Action: Label the examination as critical, and denote that AI has determined a critical finding.  For example, create an “AI-STAT” category on worklist priority.
  • Judgment: Compare the cost of interpreting this brain MRI at its usual turnaround time, versus immediately.
  • Outcome: Observe whether the action taken in response to a study labeled AI-STAT is correct.
  • Input: New MRIs of the brain MRI performed, and their prior studies.
  • Training: Historical brain MRIs
  • Feedback: Identify false positives – perhaps the prior study was from 20 years ago, or there’s been surgical resection, so that ex vacuo dilation of ventricles is not hydrocephalus.  Perhaps there has been recent surgery Identify false negatives – subtle enlargement of the temporal horns missed by AI.  Use this information to improve the AI.

How might you use this worksheet to brainstorm AI for your radiology practice?

 

Your Origin Story for Data Science

There’s an origin story for every superhero; even those without superpowers (like Batman – that’s right) got started somewhere. What we sometimes forget is that there is also an origin story for every regular person, every profession, every hobby.

Source: Wikimedia Commons

 

If you’re a radiologist looking to learn a few things in radiology data science, a simple web search will reveal a seemingly overwhelming amount of material you might have to know.

Fortunately, only a very small subset is necessary to start being productive.  Here are a few resources I used to get started.

Continue reading

William Chen’s answer to What are the top 5 skills needed to become a data scientist? – Quora

A Quora answer/article about data science.

Incidentally, the same 5 skills are also highly relevant to be a physician-informatician, particularly in radiology.  Give it a read.

Source: William Chen’s answer to What are the top 5 skills needed to become a data scientist? – Quora

DICOM Processing and Segmentation in Python – Radiology Data Quest

There is something strangely satisfying about being able to take things apart and putting it back together.  Inspired by the popularity of Lego sets in our childhoods, Minecraft brought this sense of wonder to video games.

For those of us who are life-long tinkerers who happen to be radiologists, I published in Radiology Data Quest a DIY on how one take DICOM apart and manipulate it.  All in Python, no less.

 

DICOM is a pain in the neck.  It also happens to be very helpful.  As clinical radiologists, we expect post-processing, even taking them for granted. However, the magic that occurs behind the scene…

Source: DICOM Processing and Segmentation in Python – Radiology Data Quest

Informatics Sessions at RSNA 2016 You Don’t Want to Miss

The Radiology Society of North America (RSNA) Annual Meeting is a place to expand your knowledge base, both by taking a deeper dive into your core interest and by getting your feet wet a few new skills.

If informatics is something you’ve been interested in but need a good way to get started, then the RSNA offers some solid opportunities for beginners. Continue reading

When Gut Instinct and Logic Work Together

After the recent presidential election, you are probably either particularly alarmed or especially excited about the outcome.  Regardless of your particular political predilection, it is fair to say that this election puts data science on its head when so many got so wrong.

On an earlier issue of Harvard Business Review, the venerable magazine shared a piece of research from the University of Southern California on forecasting. When forecasting sales, the best estimators use a combination of intuition and logic – with both the logic-heavy and intuition-heavy forecasters performing less accurately.

In the age of artificial intelligence and big data, it can be sobering to realize that despite the staggering volume of data we are now collecting, ignoring your gut instincts can take a heavy toll on your decision-making abilities.

 

Source: “What type of forecaster are you?” Harvard Business Review (March): 26.

 

 

6 Elements of a Data-Driven Informatics Solution (1/3)

Big Data has become a radiology buzzword  (the others: machine learning, AI, and disruptive innovation are also up there).

However, there is a real problem with using the term Big Data – it isn’t just one set of data problems.  Big Data is a conglomerate of different data challenges: volume of data, heterogeneity of data, or the velocity of data are all important dimensions.  Machine learning and internet of things are others layers superimposed on the big data problem.

Sometimes it is helpful to step back and approach data problems with a common framework, a way to think about how and which facets of data science fit in a real-life workflow in the face of an actual problem.

Below is a 6-element framework that helps me think about data-driven informatics problems. They are generally in chronological order, but they are not “steps” because you frequently will find yourself going back and redefining many things.  However, the framework helps you maintain a big-picture outlook.  The reason any sufficiently complex data problem requires a team approach. Continue reading

8 Ways to Gauge The Quality of Your Data Team

At the turn of the century, Joel Spolsky came up with the idea of a “Joel Test” – a highly irresponsible, sloppy test to rate the quality of a software team.

Then, this group thought to come up with their own criteria to rate the quality of a data science team.  How do your analysts in the radiology department fare?

The “Joel Test” for Data Science

  1. Can new hires get set up in the environment to run analyses on their first day?
  2. Can data scientists utilize the latest tools/packages without help from IT?
  3. Can data scientists use on-demand and scalable compute resources without help from IT/dev ops?
  4. Can data scientists find and reproduce past experiments and results, using the original code, data, parameters, and software versions?
  5. Does collaboration happen through a system other than email?
  6. Can predictive models be deployed to production without custom engineering or infrastructure work?
  7. Is there a single place to search for past research and reusable data sets, code, etc?
  8. Do your data scientists use the best tools money can buy?