🎥 Lecture 8 Video 4 — Cryo-EM Image Formation, Contrast & Fourier Space

This lecture explains how electron microscope images are formed, why contrast is difficult in cryo-EM, and how Fourier transforms help us reconstruct 3D structures.

⚡ 1. Electron Wave–Particle Duality & Image Formation

Electrons behave both like:

• 🔵 Particles → detected as individual hits on detectors
• 🌊 Waves → interfere and produce contrast patterns

An electron beam entering the microscope is almost like a plane wave that:

1. Interacts with atoms in the specimen
2. Gets scattered or transmitted
3. Recombines in the image plane
4. Produces an interference pattern = image

Important fact:

👉 Most electrons are NOT scattered — they just pass through. Only a small fraction interacts and contributes to image formation.

This is one major reason why cryo-EM images are noisy and low contrast.

🎨 2. What is Contrast in Electron Microscopy?

Contrast simply means:

Difference between bright and dark areas that allows us to distinguish structures.

In cryo-EM there are two main contrast mechanisms:

🔴 Amplitude Contrast (Particle view)

This occurs when:

• Some electrons are absorbed or scattered away
• Fewer electrons reach the detector behind dense objects

Result:

➡ Darker regions where electrons were removed.

Mechanism:

• High-angle scattering
• Absorption
• Electron-dense material blocks beam

You can imagine:

“Electron shadow” behind dense objects.

🟣 Phase Contrast (Wave view)

This occurs when:

• Electrons are not removed
• Their phase (timing of wave oscillation) changes.

Key idea:

➡ The wave shifts slightly in position → interference creates contrast.

In cryo-EM:

⭐ Phase contrast is the dominant mechanism.

This is because:

• Proteins and water have similar electron densities
• So amplitude contrast is weak
• Phase shifts carry most structural information.

💥 3. Types of Electron Scattering

When electrons hit atoms, three outcomes are possible:

✅ 1. No scattering (Transmission)

• Electron passes straight through
• No energy loss
• Does NOT contribute much to image contrast

⭐ 2. Elastic scattering (Very important!)

• Electron is deflected but keeps its energy
• Produces phase shifts
• Forms the useful structural signal

👉 Elastic scattering is what builds the image.

⚠️ 3. Inelastic scattering (Bad!)

• Electron loses energy
• Can eject secondary electrons
• Causes radiation damage
• Adds noise

Therefore:

Cryo-EM tries to maximize elastic scattering and minimize inelastic scattering.

🧪 4. Negative Stain vs Cryo-EM Contrast

This lecture gives a very important conceptual comparison.

🟤 Negative Stain EM

• Heavy metal stain surrounds protein
• Huge electron density difference
• Strong amplitude contrast

Result:

➡ Protein appears white in dark background

Why?

• Heavy stain scatters electrons strongly.

❄️ Cryo-EM

• Protein embedded in vitreous ice (water)
• Small density difference

Result:

➡ Protein appears dark on grey background

Main contrast source:

⭐ Phase contrast (not amplitude).

This explains:

Why cryo-EM images look noisy and faint.

🧠 5. Real Space vs Fourier Space

A fundamental concept in structural biology imaging.

🖼️ Real Space

• What we normally see
• The actual particle image (micrograph)

🌐 Fourier Space (Power Spectrum)

After Fourier transform:

• Image becomes concentric rings
• Represents spatial frequency information

Interpretation:

Low frequencies help:

✅ Particle detection ✅ Alignment

High frequencies help:

⭐ Atomic resolution reconstruction

But:

⚠️ Also contain more noise.

⚙️ 6. Why Use Fourier Transforms?

Because they make image processing:

🚀 Much faster computationally.

Key uses:

• Resolution estimation
• Contrast Transfer Function (CTF) fitting
• Defocus determination
• Filtering (noise removal)
• Particle alignment
• Correlation analysis
• 3D reconstruction

Fourier transforms allow:

Moving between real space ↔ frequency space to manipulate information efficiently.

🎛️ 7. Filtering in Fourier Space

Very powerful trick in cryo-EM.

Types:

🔵 Low-pass filter

• Removes high frequencies
• Keeps overall shape
• Improves contrast

Used for:

➡ Particle picking ➡ Alignment

🔴 High-pass filter

• Removes low frequencies
• Enhances edges / fine details

🟢 Band-pass filter

• Removes both very low and very high frequencies
• Keeps useful mid-range information

These filters help:

Reduce noise and make particles easier to see.

🦆 8. The Famous “Duck” Example — 3D Reconstruction Concept

Beautiful conceptual explanation.

Idea:

• A 3D object produces many 2D projections
• Each projection has its own 2D Fourier transform
• These are slices through a 3D Fourier transform

If we collect enough projections:

➡ We can reconstruct the full 3D Fourier space

Then:

➡ Fourier inversion → real space 3D structure.

This is the core mathematical principle of single-particle cryo-EM reconstruction.

🧊 9. Why Not All Electron Microscopes Can Do Cryo-EM

Important practical insight.

Cryo-EM requires special features:

• ❄️ Cryo holder (liquid nitrogen temperature)
• 🎯 High-contrast biological objective lens
• ⚡ Low-dose imaging system
• 📸 Direct electron detectors

Material science TEMs usually lack these.

So:

You cannot just use any TEM for biological cryo-EM work.

🧩 Final Big Picture

This lecture teaches the physics + math foundation of cryo-EM image processing:

Core ideas to remember:

⭐ Image contrast = amplitude + phase ⭐ Cryo-EM contrast is mainly phase contrast ⭐ Elastic scattering builds signal ⭐ Inelastic scattering builds noise + damage ⭐ Fourier space separates structural information by resolution ⭐ Filtering improves visibility ⭐ 3D reconstruction uses many 2D projections in Fourier space