* Total Points: 0
Back to Lessons The AI Landscape for Clinicians 0 pts Module 1 · Lesson 1
Introduction

The AI Landscape for Clinicians

What it actually is, stripped of the hype

The Problem

You are reading a vascular surgery abstract. It says the authors "used artificial intelligence to predict AAA rupture risk." That sentence tells you almost nothing.

Did they use a logistic regression with a fancy label? A random forest? A deep neural network trained on CT images? The words "artificial intelligence" have become so overloaded that they obscure more than they reveal.

AI is not a method. It is a category -- a big one. Saying "we used AI" is like saying "we used statistics." The first question is always: what kind?

Two Maps You Need

To make sense of any AI paper, you need two things: what kind of tool they used, and how that tool learned. Think of it like surgery -- you need to know both the procedure (bypass vs. stent) and the approach (open vs. endovascular).

Map 1: The Family Tree of AI

This is not a straight line -- it is a tree with branches. AI is the trunk. It splits into different approaches, and those split again. Tap any box to learn more.

Artificial Intelligence (AI)
The umbrella
Any system that does something normally requiring human intelligence. Everything below is AI.

An EMR drug-interaction alert? AI. A self-driving car? AI. ChatGPT? AI. The term alone tells you nothing useful.

Analogy: "AI" is like saying "surgery." It could mean a skin biopsy or a thoracoabdominal aortic repair.
Rule-Based Systems
Old school
Humans write the rules by hand. No learning from data.

Most clinical decision support in your EMR. Technically AI, but a human programmed every rule.

Example: Best Practice Alerts in Epic.
Machine Learning (ML)
Most papers
The computer figures out the rules by looking at data. Most surgical research sits here.

You give the computer variables and outcomes; it finds the best way to connect them. You gain prediction accuracy but lose interpretability.

Example: Predicting 30-day mortality after EVAR using 25 preop variables from NSQIP.
Traditional ML
Spreadsheets
Works on structured data -- REDCap exports, registry tables. Faster, needs less data.

The workhorse of clinical ML. Spreadsheet of patient variables + outcome column = these tools.

Example: XGBoost predicting amputation risk from VQI data.
Random Forest XGBoost SVM LASSO / Ridge k-NN Naive Bayes
Deep Learning
Images & text
Neural networks with many layers. Raw images, text, or waveforms -- not just spreadsheets.

Can take a raw CT scan or op note and learn features itself. Tradeoff: needs much more data, harder to explain.

Example: A neural network reading CTA and measuring AAA diameter.
CNNs
Images
Visual patterns -- CT, ultrasound, wound photos.

DL on CT scans, ultrasound, or wound photos = almost certainly a CNN.

Example: Carotid plaque classification from duplex.
LLMs
Text
GPT-4, Claude. Read and write text.

Extract info from notes, summarize literature. They do not run statistics.

Example: Pulling graft type from 5,000 op notes.
Others
RNNs, GANs
Time-series, synthetic data.

You don't need to memorize these. Deep learning is a whole family.

Example: RNN predicting instability from ICU waveforms.
!

The Key Takeaway

AI is a tree, not a ladder. When someone says "we used machine learning," there is still a branch to identify (traditional ML or deep learning?). When they say "deep learning," there is still a branch (images? text? time-series?). Each branch has different data requirements, different strengths, and different failure modes.

Map 2: How the Machine Learns

This is a separate question from the tool hierarchy above. Any ML or deep learning model uses one of these three approaches. You will see these terms in every ML paper.

Most common in clinical research

Supervised Learning

You give the computer both the inputs (patient data) and the correct answers (outcomes). It learns to predict the answer for new patients.

Analogy: Teaching a resident by showing them 1,000 cases where you already know the diagnosis. "Here are the features, here is what happened -- now predict the next one."

Less common, but you will see it

Unsupervised Learning

You give the computer the inputs but no answers. It finds patterns and groupings on its own. Useful for discovering patient subtypes or phenotypes you did not know existed.

Analogy: Handing a resident 1,000 patient charts with no diagnoses and asking, "Do you see any natural groupings?" That is clustering.

Rare in surgical research (for now)

Reinforcement Learning

The computer learns by trial and error, getting rewarded for good decisions and penalized for bad ones. Shows up in treatment optimization and dosing algorithms.

Analogy: A resident learning to titrate a heparin drip -- not from a textbook, but by adjusting the dose, checking the result, and adjusting again over hundreds of patients.

When you read an AI paper, ask two questions: Where does it sit in the hierarchy? (basic ML on a spreadsheet, or deep learning on images/text?) and How did it learn? (supervised with labeled outcomes, or unsupervised looking for hidden patterns?). Most vascular surgery papers: basic ML, supervised, structured registry data.

The Translation Problem

The same technique gets called different things depending on who is talking. Here is the Rosetta Stone.

What authors write vs. what they mean

"AI-based prediction" Could be anything from logistic regression to a neural network. Meaningless without specifics.
"Machine learning model" Usually random forest, gradient boosting (XGBoost), or similar. Sometimes just penalized logistic regression.
"Deep learning approach" A neural network, usually processing images or unstructured text. Requires substantially more data.
"Novel algorithm" Often a standard method applied to a new dataset. True novelty is rare.
"Automated pipeline" Code that runs preprocessing, modeling, and evaluation in sequence. Not a method itself.

Red Flag: When a paper says "we used AI" or "we used machine learning" without specifying the exact algorithm, that is a reporting failure. You would not accept "we used statistics" without knowing if they ran a t-test or a Cox model. Apply the same standard.

1

The One Question That Matters

When you read an abstract or hear a presentation that mentions AI or ML, the first question to ask is: "What specific algorithm did you use, and what type of data did it process?"

The answer tells you whether you are dealing with a random forest on structured data (familiar territory, just an extension of what you know), or a deep learning model on images (different validation requirements, different failure modes).

AI in Vascular Surgery: What Actually Exists

A realistic landscape of where AI is being used in your specialty, not where someone hopes it will be in 10 years.

Most Common

Risk Prediction from Structured Data

ML models (usually random forest, XGBoost) trained on registry data (NSQIP, VQI) to predict outcomes: mortality, amputation, graft failure, readmission.

Typical AI layer: Machine Learning. Data: structured variables from EHR/registries.

Growing

Image Analysis

Deep learning on CT angiography (AAA sizing, plaque characterization), duplex ultrasound (stenosis grading), wound photographs (classification, area measurement).

Typical AI layer: Deep Learning (CNNs). Data: medical images.

Emerging

Natural Language Processing

Extracting structured data from operative notes, radiology reports, clinic notes. Identifying complications, procedures, and anatomy from free text.

Typical AI layer: Deep Learning (transformers) or traditional NLP. Data: clinical text.

Mostly Hype (for now)

Autonomous Decision Support

"AI that tells you what to do." Real-time surgical guidance, automated treatment planning, LLM-driven clinical reasoning.

Almost entirely aspirational. Regulatory, liability, and validation barriers are enormous. Do not confuse conference demos with clinical deployment.

If someone presents an "AI tool" for clinical use and cannot tell you which layer it sits in, what data it was trained on, and how it was validated -- that is not ready for your patients.

Before you use any of these tools with patient data: Feeding clinical data into AI models -- including commercial LLMs like ChatGPT or Claude -- requires a HIPAA-compliant environment and IRB approval. Most commercial AI tools are not HIPAA-compliant by default. De-identified data still requires IRB review. Your institution's data governance and compliance office should be involved before any patient data touches any AI system. This is not optional.

Exercise: Pick the Right Tool

You have a clinical question. Which type of AI would you reach for to answer it?

Scenario 1 of 8

Lesson Complete!

0
Total Points Earned
Scenarios (0/8 correct) +0 pts
Lesson Completed +100 pts

The AI Landscape for Clinicians

Module 1 - Lesson 1 complete

Key Takeaways

  • AI is a family tree: AI branches into rule-based systems and ML. ML branches into traditional ML and deep learning. Deep learning branches into CNNs, LLMs, and others.
  • Most surgical AI is ML: Random forests and gradient boosting on structured registry data. Not neural networks, not ChatGPT.
  • "We used AI" is meaningless: Always ask what specific algorithm was used and what data it processed.
  • Deep learning needs images or text: If the data is a spreadsheet of variables, deep learning is usually overkill.
  • LLMs are text tools: They extract and generate language. They do not analyze your dataset or test your hypothesis.
  • Autonomous clinical AI is mostly aspirational: Do not confuse conference demos with validated, deployable tools.

The next time you have a clinical question involving data, ask: what type of data do I have, and which branch of the tree fits? That single step separates you from 90% of the audience at any AI talk.