Courses and where to find them
“The capacity to learn is a gift; The ability to learn is a skill; The willingness to learn is a choice.” Brian Herbert — Dune Author
Courses, whether in the form of short-courses or as part of degrees, are one of many ways to up-skill and become competent in new areas.
Here we’ll focus on just a handful of the many courses and topics that may interest you in your career as a clinician in a health-tech field.
While there are excellent paid courses and formal degree programmes, free or less expensive courses offer exposure to a subject matter, often in-depth and comprehensive, without the high price-tag.
Courses- platforms
These are some popular platforms where you can explore a range of courses:
Business model:
Courses on these platforms are available to audit. Auditing a course means that course content is available for no fee but additional benefits: access to interactive quizzes, graded assignments, and certifications of completion must be paid for on a course-by-course basis, or monthly/annual membership subscriptions to the platform.
Business model:
A platform created for course creation: some courses are better than others as anyone can create a course. Reading reviews and getting referrals is a good way to establish the quality of a particular course.
Courses are priced at high prices but have regular sales where courses are sold between £10–15. Sales are so regular, in fact, that you should never spend more on a Udemy course.
Usually, you get bang for your buck with lifetime access, upgrades to course content, and certification on completion.
Business model:
Technical “nanodegrees” offered at much cheaper rates than competitors providing similar experiences.
Courses are priced much higher than Udemy but much lower than many other online paid courses and promise professional support and guidance.
Udacity regularly has sales so watch out for month-off and half-price specials before purchasing.
Udacity offers free courses so get a taste of the learning experience before committing. See the examples below.
Not a course but an excellent source of information where highly skilled people have taken the time to break down concepts and share insights.
Business model:
Medium offers 5 articles free per month
Offers a wide variety of courses. Excellent source for revisiting or learning maths and statistical concepts.
Business model:
Free with an option to donate
Courses — per topic
AI and Machine Learning
Why should someone from a healthcare background learn about machine learning (ML)?
Course 1:
Dr Christopher Lovejoy, one of our DDN members, is a data scientist and medical doctor based in London. He believes there are 2 main reasons:
- It will help you in your role as a medical professional:
An INR result on its own does not mean much. Just as you would interpret the result of investigation/s in the context of a patient’s presentation and other results, as ML models become integrated into the delivery of healthcare, there will be results that need to be interpreted in context. How do you apply the results of ML algorithms in a health care setting appropriately?
-For interested doctors, ML is an opportunity to get involved in an exciting area that could have a scalable impact on healthcare and expand one’s career opportunities
Christopher runs in-person machine learning for healthcare courses in London, both independently and for UCL. See the video series below:
“From mid-2018 until early 2020, I ran courses entitled ‘Machine Learning for Healthcare’ in London. Most resources for learning machine learning were aimed at people from maths or computer science backgrounds, so the course was designed to ‘bridge the gap’ — by providing a less-technical and more healthcare-tailored introduction. I no longer have time to run the courses, but have condensed the key points into this series of videos.”
Course 2:
Sandy Wright is a medical doctor and DDN member. He is a digital health fellow, a member of the Royal Society of Medicine’s Digital Health Council, and will be starting a masters at UCL in AI-enabled Healthcare. He reviews the above course (available on Coursera):
“What is it?
A Coursera course split into three parts:
- AI for Medical Diagnosis
- AI for Medical Prognosis
- AI for Medical Treatment
Brought to you by ‘deeplearning.ai’ (Andrew Ng) and was initially promoted by Eric Topol. All of the content is delivered by Pranav Rajpurkar who does the Doctor Penguin email newsletter in conjunction with Eric Topol
What can you expect?
It’s a mix of short video lectures, quizzes to test your learning, and structured coding exercises. You’ll build image classification and segmentation models, random forest predictors, and an NLP tool to extract information from radiology reports. There’s quite a bit of emphasis on the evaluation of models, and so you’ll create a lot of the statistical methods for comparing different models.
Why did I choose it?
I’m about to start an MRes programme at UCL in ‘AI enabled Healthcare’ so this seemed like a great primer for the course and a good way to get more familiar with terms and concepts I’m probably going to be coming across a lot.
Who is this course for?
I think it’s aimed at computer science people who have an interest in the healthcare space. That being said, I think it’s super relevant for doctors who are interested in getting more hands-on experience building ML/AI models.
Do I need to know how to code?
Some knowledge of Python is a prerequisite. I have beginner-level experience of writing Python and I struggled a lot with some of the coding challenges. That being said, there is an excellent Slack support group where people will help you work through any problems you might have.
Did enjoy it?
Very much so. Some of the coding challenges got…frustrating, but I liked the way the material was delivered in the short videos. Concepts like C-statistics are approached in a very accessible way and derived from first principles, so you feel like you’re developing a good understanding of the underlying maths.”
Other machine learning course and materials:
- Overview of Machine Learning Catalogue: A summary of machine learning to minimise and explain the jargon
- https://www.elementsofai.com/ A fun introductory course offered by the University of Helsinki created to expose as many people as possible to understand AI, what AI is (and what it is not).
- AI for everyone: This course is about making the most of AI in your organisation. I enjoyed the broader context this course gives to AI
Data, Mathematics, & Statistics
In risky business: doctors’ understanding of statistics, Christopher Martyn argues the importance of clinicians’ grasping statistical concepts to accurately assess the absolute risk of conditions in their patients (especially in the context of preventative and interventive care).
To have research results requiring clinical decision-making, you need research. Understanding statistics is important for interpreting research methods and creating defensible, reproducible research.
Data is being created at a rapid pace in vast abundance. To put that in perspective:
“Decoding the human genome involves analysing 3 billion base pairs — which took ten years the first time it was done, in 2003, but can now be achieved in one week.” Data, data everywhere,
The same article highlights the related difficulties of ensuring data security and privacy concerns- especially relevant when it comes to clinical data.
Additionally, if one wanted to dive into the deeper technicalities of machine learning, “probability theory”, “linear algebra”, and “calculus” may be areas of interest.
See below a small selection of courses related to statistics, mathematics, and health informatics:
Statistics and Mathematics
- Udacity: Intro to Descriptive Statistics
- Udacity: Intro to inferential Statistics
- BMJ statistics at Square One
- Statistical-analysis with for r for public health specialisation
- Khan academy
Health Informatics
Health informatics is ‘The knowledge, skills, and tools that enable information to be collected, managed, used, and shared to support the delivery of healthcare and to promote health and wellbeing’.
Resource: https://www.e-lfh.org.uk/programmes/health-informatics/
Alex Szolnoki is the clinical informatics lead at Babylon Health and is currently enrolled in MSt in Healthcare Data: Informatics, Innovation, and Commercialization. He is a DDN member and has offered to provide members with guidance in this area.
Here are a few links to get you thinking about health informatics:
- https://www.e-lfh.org.uk/programmes/health-informatics/ course available to NHS staff
- https://elearning.ihtsdotools.org/ 40 Hour certified SNOMEDCT e-learning course entirely free for member states
Other Topics
Many topics are missing:
- Product
- UX
- Business
- Safety
- Regulatory
- Health Management …
What do you want to learn about?
Please share courses you would recommend (or perhaps not recommend) to other members in the comments!
By Kim Gresak, Digital Doctors Network Co-Founder
