REST Slab / Presaturation Band

MRIninja Knowledge Base | MRI Parameter Deep Dive Version 1.0 — May 2026

MRIninja Knowledge Base | Parameter Child Page Parent page: MRI Parameters — Overview and Classification (9501) Related pages: FOV — Field of View · Phase Oversampling · 2D vs 3D Acquisition Version 1.0 — May 2026

1. Introduction and General Purpose

A REST slab (Regional Saturation Technique) — also called a presaturation band, saturation band, or spatial saturation pulse — is a slab-selective RF pulse applied to a defined region of space before the imaging excitation of each TR, with the purpose of nulling the MRI signal from that spatial region. It is the spatial equivalent of a spectral saturation pulse: just as a spectral saturation pulse nulls a specific resonance frequency (e.g., fat), a REST slab nulls all signals from a specific spatial volume.

The utility of spatial saturation is broad and addresses several distinct diagnostic problems:

Flow artefact suppression: blood flowing into the imaging volume from outside (inflowing unsaturated spins) produces bright signal in gradient echo sequences and pulsation ghosts in all sequences. Placing a saturation band superior and/or inferior to the imaging volume saturates this inflowing blood before it enters the imaged slices — eliminating the signal from moving blood, reducing both the flow-related enhancement of vessels in TOF-like sequences and the pulsatile ghosting artefacts in the phase-encoding direction.

Fold-over artefact suppression: anatomy outside the prescribed FOV in the phase direction can be saturated before the phase-encoding step, eliminating its contribution to the fold-over artefact. This is an alternative to phase oversampling (see the Phase Oversampling page) and is sometimes preferred when the phase-encoding direction or scan time constraints prevent adequate OS.

Fat suppression (regional): in regions where global fat suppression techniques (SPAIR, STIR) fail due to B0 inhomogeneity (near metallic implants, at body edges), a localised saturation band placed over a specific fat depot can provide targeted fat signal reduction.

Motion suppression from specific organs: a saturation band placed over the heart (for spine or liver imaging) reduces cardiac motion ghosting by saturating the myocardial and blood signal before each TR.

Spectroscopy outer volume suppression: in MR spectroscopy (MRSI), REST slabs are applied around the spectroscopic voxel to suppress lipid and water signals from the brain surface and scalp, which would otherwise contaminate the spectral baseline.

Unlike most MRI parameters that modify k-space acquisition, the REST slab is a real-time preparation pulse that modifies the tissue signal before the readout. Its effects are visible only in the saturation region and only for tissues that have not recovered their longitudinal magnetisation between the saturation pulse and the imaging excitation.

Historical context: spatial saturation pulses were introduced in the earliest days of clinical MRI (mid-1980s) as a solution to the flow artefact problem that dominated the early vascular imaging literature. The concept of presaturation to suppress inflowing blood signal was described in landmark papers on cardiac and vascular MRI from 1985 to 1990. REST slabs were subsequently adopted for fold-over suppression, spectroscopy outer-volume suppression, and regional fat/tissue saturation.

2. Physical Foundations

A REST slab is produced by applying a slab-selective RF pulse (typically 90° or large flip angle) to a defined spatial volume using a slice-selection gradient, immediately followed by a spoiler gradient that dephases the transverse magnetisation within that volume. The result:

• Longitudinal magnetisation (Mz) within the saturation volume is driven to zero (or close to zero) by the 90° pulse
• The spoiler gradient dephases any residual transverse magnetisation → no signal from this region during the subsequent readout
• The T1 recovery of the saturated tissue determines how much signal returns by the time of the imaging excitation

2.1 Mathematical Foundations

Signal recovery after saturation

After a 90° saturation pulse followed by a delay time TD (the time between saturation and imaging excitation):

Mz(TD) = M₀ × [1 − e^(−TD/T1)]

where:

• M₀ = equilibrium longitudinal magnetisation
• TD = saturation delay (time from saturation pulse to imaging excitation)
• T1 = longitudinal relaxation time of the tissue

Physical meaning: the saturated tissue recovers its Mz according to its T1. If TD ≪ T1: Mz ≈ 0 → complete saturation maintained. If TD ≈ T1: Mz ≈ 63% M₀ → partial saturation. If TD ≫ T1: Mz → M₀ → saturation ineffective.

Clinical implication: saturation effectiveness depends on T1 and the delay TD between the saturation pulse and the imaging excitation. For blood (T1 ≈ 1200–1400 ms at 1.5T; 1600–1800 ms at 3T), saturation remains effective for TD up to 500–700 ms (approximately 50% signal suppression at TD=T1/2). For fat (T1 ≈ 260 ms at 1.5T), saturation recovers much faster — a fat saturation band is effective for only TD ≈ 100–150 ms.

Saturation efficiency vs flow velocity

For flowing blood, the effectiveness of a saturation band depends on how much of the saturated blood remains within the imaging slices when the excitation pulse fires. Blood that flows too slowly (laminar flow in small vessels) is still fully saturated at the time of excitation. Blood that flows very fast (cardiac output, large arteries) may already have been replaced by fresh unsaturated blood from beyond the saturation band by the time the excitation fires.

The critical velocity for saturation loss:

v_crit = d_band / TD

where d_band = thickness of the saturation band, TD = delay between saturation and excitation. For a 40 mm saturation band and TD = 100 ms: v_crit = 40/100 = 400 mm/s = 40 cm/s. Blood flowing at > 40 cm/s through the saturation band exits the band before the excitation fires → unsaturated → appears bright.

Practical implication: for arterial flow suppression in the aorta (peak velocity 80–120 cm/s), the saturation band must be thick (≥ 80–120 mm) or placed very close to the imaging volume to prevent fast-flowing blood from exiting the saturation region before excitation.

SAR contribution

Each REST slab adds an RF pulse to every TR interval. The SAR contribution:

SAR_REST = k × B₁² × t_pulse × (N_REST_slabs / TR)

At 3T, the higher B₁² × RF_power scaling means that REST slabs contribute substantially more SAR than at 1.5T for the same flip angle and pulse duration. Multiple REST slabs (e.g., 3–4 saturation bands for spectroscopy outer volume suppression) can generate SAR levels that limit the maximum flip angle or minimum TR of the imaging sequence.

3. Units, Terminology and Vendor Nomenclature

REST slabs are defined by their spatial position, thickness, and orientation — all expressed in mm and degrees.

Parameter specifications for a REST slab:

• Position: distance from isocentre in the selected orientation (mm); or defined by graphical placement on the planning image
• Thickness: extent of the saturation volume in the orientation direction (mm); typical range 30–100 mm
• Orientation: the saturation slab can be placed in any arbitrary orientation (axial, sagittal, coronal, or oblique)

4. Typical Value Ranges

4.1 REST Slab Applications and Typical Parameters

4.2 REST Slab SAR at Different Field Strengths

5. Parameter Interaction Ecosystem

5.1 Parameter Relationships Matrix

5.2 The TD (Saturation Delay) — the Critical Hidden Parameter

The saturation delay TD is not directly settable by the technologist on most platforms — it is determined implicitly by:

TD = time_from_REST_pulse_to_excitation = position_of_REST_pulse_in_TR × TR

In multi-slice 2D acquisitions with interleaved acquisition, the REST pulse fires once per TR before all slices are excited. The effective TD for each slice differs depending on its position in the interleaved sequence:

• The first-acquired slice sees TD ≈ T_REST-to-first-excitation (typically 50–100 ms)
• Later slices in the interleaved sequence see longer TD
• At TR = 2000 ms with N_slices = 20: the last slice sees TD ≈ 2000 − (20−1)×100 ms ≈ 100 ms before the next TR REST pulse

Practical implication: slices acquired late in the TR period have a longer TD from the REST pulse of the previous TR. For tissues with short T1 (fat, T1=260 ms), partial T1 recovery occurs → saturation is less effective for late-acquired slices → the fat signal in late-acquired slices appears brighter than in early-acquired slices → heterogeneous fat suppression across the stack.

6. Effects on Image Appearance

6.1 Correct REST Slab Placement (Outside FOV)

• Target region (REST band): signal is nulled or substantially reduced. This region appears dark or very low signal on all contrast-weighted images.
• Imaging FOV (untouched by REST band): no change in signal, contrast, or resolution. The imaging sequences behave exactly as without the REST slab.
• Artefact reduction: motion ghosts from the saturated region are eliminated or substantially reduced in the phase-encoding direction of the imaging FOV. Flow-related enhancement in vessel segments within the saturation zone is removed.

6.2 Misplaced REST Slab (Overlapping with Imaging FOV)

When the REST slab partially or fully overlaps the imaging FOV:

• Tissue within the overlap region is saturated → signal reduction or signal void
• This appears as a dark band across the image at the location of the REST slab → simulates a flow void, haemorrhage, or anatomical structure
• If the REST slab saturates part of a lesion, the lesion may appear smaller or signal-different from its true extent
• Enhancement post-contrast may be suppressed in the saturated region

This is one of the most clinically dangerous technologist errors in MRI — a misplaced REST slab can simulate pathology or mask it entirely.

7. Effects on Acquisition Time

7.1 Direct Effect

Each REST slab adds one RF pulse and one spoiler gradient pair per TR interval. The time required for the REST pulse (including the slice-selection gradient, the RF pulse itself, and the spoiler gradient) is typically 5–20 ms per REST slab.

For N_REST slabs per TR:

T_REST_total = N_REST × T_per_REST ≈ N_REST × 10–20 ms

At TR = 2000 ms: N_REST = 3 slabs × 15 ms each = 45 ms extra per TR → 2.25% increase in TR period.

This is negligible in absolute terms. The practical impact is:

• The TR period is longer by T_REST_total → marginally fewer slices can be acquired per TR (the maximum slice number for a given TR is reduced by T_REST_total)
• At 3T with SAR limitations: the extra RF energy from REST slabs may trigger automatic TR extension (→ indirect time cost)

7.2 SAR-Driven TR Extension

This is the dominant time effect at 3T. Each REST slab adds substantial SAR (B₁² × duration × 1/TR). If the scanner's SAR limit is already near-exceeded by the imaging sequence:

• Adding REST slabs pushes SAR over the limit
• Scanner automatically extends TR to reduce SAR per time → TR increases
• Longer TR → scan time increases

At 1.5T, REST slabs rarely cause SAR-driven TR extensions. At 3T, 3–4 REST slabs with a long-ETL TSE sequence may require TR extension of 10–30%.

8. Effects on SNR and CNR

8.1 SNR in the Saturation Zone

Within the REST slab region: SNR → near zero. This is the intended effect for the specific tissue being suppressed (flowing blood, motion-producing anatomy, fold-over tissue).

8.2 SNR in the Imaging FOV

No direct SNR effect within the imaging FOV. The saturation pulse does not alter the magnetisation of tissues within the imaging volume (provided the REST slab does not overlap the imaging FOV).

Indirect benefit: eliminating ghost artefacts from the saturation zone removes a source of structured noise (periodic ghosting in the phase direction) that competes with lesion signal. Removing this structured artefact improves the effective CNR for the target anatomy — a lesion that was previously obscured by a pulsatile ghost is now clearly visible.

8.3 Field-Strength Dependency

At 3T: the higher T1 values at 3T mean that blood T1 ≈ 1700 ms. A REST slab for blood suppression needs TD ≪ 1700 ms to maintain effectiveness. At short TR (GRE sequences with TR ≈ 5 ms), TD is essentially the pulse duration → excellent saturation. At longer TR (TSE with TR ≈ 3000 ms) and with multiple slices: later slices see TD approaching T1_blood → partial saturation recovery → residual blood signal.

At 1.5T: blood T1 ≈ 1200 ms. Saturation is slightly more effective at equivalent TD than at 3T because T1 is shorter → less recovery between REST pulse and excitation.

9. Artefacts Associated with REST Slabs

10. Behaviour Across Sequence Families

Spin Echo and Turbo Spin Echo

REST slabs are applied once per TR before the excitation pulse. In multi-slice 2D TSE, the REST pulse fires at the beginning of each TR period, and all slices within that TR share the same REST saturation. The TD for each slice increases with its position in the interleaved order. Typically 1–2 REST slabs are used for spine and brain SE/TSE protocols (superior for arterial suppression, inferior for venous suppression in carotid imaging).

Gradient Echo (GRE/FLASH)

At very short TR (5–10 ms), the TD between REST pulse and imaging excitation is minimal → excellent saturation maintained. REST slabs are highly effective in GRE sequences for flow suppression because the short TR means very little T1 recovery between saturation and excitation. In DCE (body, breast), a REST slab placed over the heart effectively suppresses cardiac ghosting in the phase direction.

Inversion Recovery (STIR, FLAIR, MPRAGE)

REST slabs in IR sequences require care: the REST pulse fires before the inversion pulse in each TR. The saturation effect is further modified by the subsequent inversion pulse — the saturated tissue's Mz that has partially recovered is then inverted. The net effect on the IR signal is complex and tissue-T1-dependent. For STIR and FLAIR, REST slabs for flow/fold-over suppression are generally applied as standard (their effect on IR contrast within the imaging FOV is negligible if the band is correctly placed outside the FOV). For MPRAGE, REST slabs are not typically used.

EPI (DWI, DSC)

In EPI, the REST slab fires once before the RF excitation (for 2D multi-slice EPI, once per volume or per slice depending on the implementation). The saturation of inflowing blood in body DWI is important to suppress the high T2 signal from vascular blood (the IVIM pseudo-diffusion component). An inferior REST slab in abdominal DWI suppresses aortic and portal venous signal. For brain DWI, REST slabs are less commonly needed (the brain is not moving relative to the coil, and vascular ghosting is less severe).

DSC Perfusion

REST slabs are generally not used in DSC perfusion because the sequence is designed to detect the passage of contrast through blood vessels. Saturating blood flow would defeat the purpose. The pulsation artefacts in DSC are managed by the short TR (1.5 s per volume) and the EPI-based rapid readout.

DCE (Dynamic Contrast Enhanced)

In breast and body DCE (3D GRE), REST slabs are used to:

1. Suppress cardiac motion ghosting (for left breast/liver DCE): a REST slab placed obliquely over the heart reduces the bright blood signal from cardiac chambers that produces ghosting in the phase direction.
2. Suppress bowel/anterior abdominal wall fold-over in body DCE.

REST slabs in DCE must be planned to not cover the target anatomy. The SAR from REST slabs in 3D DCE at 3T must be accounted for in the total SAR budget.

ASL (Arterial Spin Labelling)

REST slabs have a special role and a special danger in ASL:

• Role: a REST slab placed between the labelling plane (in the neck) and the imaging volume (in the brain) suppresses venous signal that might contaminate the arterial label. Also used to suppress cardiac motion at the imaging volume edge.
• Danger: a REST slab placed at the labelling position would saturate the labelled blood before it reaches the brain → complete loss of ASL signal. REST slabs must never be placed in the ASL labelling or post-labelling delay zones.

bSSFP (TrueFISP/FIESTA/CISS)

REST slabs are rarely used in bSSFP sequences because the steady-state condition of bSSFP is sensitive to any magnetisation preparation that disrupts the established equilibrium. A REST slab applied at the edge of the imaging volume may partially disrupt the bSSFP steady state in adjacent slices, producing transient signal oscillation. For cardiac cine bSSFP, REST slabs are sometimes applied over the chest walls to reduce respiratory ghosting — placed sufficiently far from the cardiac slices to avoid steady-state disruption.

TOF MRA (Time of Flight)

REST slabs are central to the clinical utility of TOF MRA. In 3D TOF brain MRA (imaging the intracranial arteries):

• A superior REST slab (superior to the imaging volume) saturates blood entering from above (venous drainage) → veins appear dark → only arterial signal visible
• A travelling REST slab (moving with the imaging volume as it progresses through the stack) maintains venous suppression throughout the acquisition as the venous blood would otherwise re-enter the imaging volume from the superior direction

The travelling REST slab (venous saturation band, TONE — Tilted Optimised Non-saturating Excitation, or equivalent) is the defining technical element of high-quality TOF brain MRA. Its correct positioning and gap from the imaging volume determines the quality of venous suppression and the residual arterial signal.

Spectroscopy (MRS/MRSI)

Outer volume suppression (OVS) in brain MR spectroscopy uses multiple REST slabs (typically 6–8 slabs in different orientations) to suppress the scalp lipid signal around the spectroscopic voxel. Lipid from the scalp has a broad spectral peak that overlaps the metabolite peaks of interest (especially the choline and creatine peaks at 3.2 and 3.0 ppm). Each OVS slab fires per TR, adding substantial SAR at 3T. The flip angles for OVS slabs in spectroscopy may be individually optimised (130–150° each) to achieve better saturation than a simple 90° pulse. This is the most technically demanding application of REST slabs.

11. Field Strength Behaviour

At 3T — the key clinical constraint: the B₁² SAR scaling means that REST slabs at 3T consume approximately 4× the SAR per pulse compared with 1.5T. For protocols with multiple REST slabs (spectroscopy OVS, or combined superior + inferior + anterior REST slabs in spine imaging), the scanner will typically extend TR by 20–50% to maintain SAR compliance. This TR extension directly changes the T1 contrast and scan time. The technologist must monitor TR after adding REST slabs and inform the radiologist if TR has been automatically extended.

At 7T: the extreme SAR at 7T makes REST slabs effectively impractical for most applications. The few available RF power budget must be used for the imaging sequence itself. Alternative flow suppression strategies (phase contrast, background suppression, black blood techniques) replace REST slabs at 7T.

12. Vendor-Specific Implementation

Siemens

REST slabs are added in the "Contrast" or "Sequence" tab as "REST" elements. Up to 6 REST slabs can be defined per sequence. Each REST slab has individually adjustable position, thickness, and orientation. The Siemens planning interface allows graphical placement of REST slabs on the localiser image — the most intuitive implementation. The "saturation efficiency" can be estimated from the vendor's SAR monitor: as REST slabs are added, the remaining SAR headroom decreases; when the limit is reached, TR extends automatically.

Siemens TOF sequences include a dedicated "Venous Saturation" tab that controls the travelling saturation band (separate from the standard REST slabs). The gap between the travelling slab and the imaging slab and the slab thickness are independently settable.

GE

SAT bands are added through the "Options" or "Saturation" control in the sequence setup. GE distinguishes between in-plane saturation bands (placed within the FOV to suppress specific anatomy) and out-of-plane bands (placed above/below the FOV for flow suppression). Up to 6 SAT bands are supported. GE applies the SAT pulse in a fixed temporal position within the TR sequence — the technologist cannot adjust when in the TR the SAT fires.

Philips

REST slabs are called "REST" in Philips and are placed graphically on the scan geometry. The Philips system provides explicit SAR feedback in real-time as REST slabs are added. The "REST effectiveness" is partially displayed as a note when the REST band overlaps with the imaging FOV boundary — a useful safety check.

Canon

SAT bands positioned graphically. Canon's system provides SAR monitoring per sequence element. REST slabs in Canon protocols are defined identically to slice positioning — thickness, gap, and orientation.

United Imaging

SAT slabs with graphical positioning. The UIH system includes an automatic warning when a REST slab is positioned within or immediately adjacent to the imaging FOV — reducing the most critical positioning error.

Hidden coupling — all vendors: on all platforms, when REST slabs are added to a sequence and the SAR limit is exceeded, the TR is automatically extended. This TR extension is not always prominently displayed — the technologist may not notice that TR has changed from 2500 ms to 3200 ms after adding two REST slabs. Always verify TR after adding REST slabs.

13. Practical Optimisation Strategies

13.1 Clinical Optimisation Recipes

13.2 REST Slab vs Phase Oversampling — Decision Framework

Both REST slabs and phase oversampling address fold-over artefact. The choice depends on:

General rule: phase oversampling is the more reliable and universal solution to fold-over. REST slabs are preferred when: the fold-over anatomy is a specific identifiable structure (aorta, heart) that can be precisely targeted; the tissue has long T1 (effective saturation); and the REST slab placement does not increase SAR beyond the protocol limit.

14. Parameter Extremes

14.1 No REST Slab

Standard for sequences where flow and motion artefacts are not the primary image quality problem, or where the imaging anatomy fully occupies the FOV without fold-over risk (e.g., brain axial at large FOV, extremity imaging with small FOV entirely within the anatomy). Most sequences are routinely acquired without REST slabs. The absence of REST slabs reduces SAR and maximises the number of slices achievable per TR.

14.2 Maximum REST Slabs (5–6 per TR, Spectroscopy OVS)

MR spectroscopy outer volume suppression represents the most demanding REST slab application. Six or more OVS slabs per TR — each at a different orientation surrounding the spectroscopic voxel — collectively suppress all scalp lipid from all sides. At 3T, each OVS slab at 90° flip angle contributes approximately 15–20 mJ/TR to the SAR budget. Six slabs × 20 mJ = 120 mJ per TR. For TR=2000 ms: average SAR contribution ≈ 60 mW/kg → potentially exceeding the 3.2 W/kg IEC whole-body SAR limit without reducing flip angles. In practice:

• OVS slab flip angles are typically reduced to 130–150° (partial saturation acceptable to manage SAR)
• TR is extended to 2500–3000 ms (acceptable for spectroscopy contrast)
• The number of OVS slabs may be reduced to 4 (inferior, superior, bilateral) when the spectroscopic voxel is well-centred

15. Common Optimisation Errors

16. MRI Technologist Pearls

Always verify REST slab placement on three planes: the localiser provides axial, sagittal, and coronal views. A REST slab that appears correctly placed in the sagittal view may overlap the imaging FOV in the axial or coronal view. Check all three planes before starting. Any overlap between the REST slab and the first or last imaging slice is a critical error.

The 10 mm safety gap rule: place the edge of the REST slab at least 10 mm from the first imaging slice in all directions. The REST slab's RF profile has a transition zone of 5–10 mm where saturation is partial — this transition zone must not overlap the imaging slices.

Monitor TR after adding REST slabs at 3T: at 3T, the single most important technologist check after adding REST slabs is verifying that TR has not been automatically extended. Compare the TR in the sequence protocol with the TR after the REST slabs are added. If TR has increased by > 10%, consider reducing the number or flip angle of REST slabs.

REST slabs for fold-over vs phase oversampling: REST slabs suppress fold-over from specific anatomy but may be incomplete if the tissue has short T1 or if the anatomy moves. Phase oversampling (see Phase Oversampling child page) is more reliable for complete fold-over elimination. Use REST slabs when: the fold-over source is identifiable and stationary (aorta, liver edge); the tissue has long T1 (blood, solid organs). Use phase oversampling when complete fold-over suppression is required regardless of tissue type.

Travelling saturation band for TOF MRA — gap is critical: the gap between the travelling REST slab and the imaging slab in TOF MRA must be large enough to allow blood to partially recover (so that arteries are not fully saturated) but small enough to maintain adequate venous suppression. The optimal gap is approximately 10–20 mm. A gap < 5 mm may partially saturate slow arterial flow in small vessels; a gap > 30 mm allows fast venous blood to partly recover from saturation before entering the imaging volume.

OVS for MRS — reduce flip angle before reducing slab number: if SAR is limiting the number of OVS slabs in brain MRS at 3T, reducing the OVS flip angle from 90° to 70° reduces SAR by (sin70°/sin90°)² ≈ 0.88 factor per slab, allowing more slabs with the same SAR. The saturation at 70° is slightly less complete, but the trade-off is usually preferable to removing an OVS slab entirely.

17. Real Clinical Examples

Example 1: Cervical Spine MRI — Carotid Artery Ghost Elimination

Clinical scenario: cervical spine T2 TSE sagittal; A-P phase encoding direction; at TR=3500 ms, strong pulsatile ghost from the carotid artery traversing anteroposteriorly through the imaging FOV, obscuring C3–C5 disc levels in the mid-scan.

Problem: the common carotid artery runs anterior to the cervical spine. Its pulsatile flow produces a ghost in the A-P phase direction that crosses the vertebral bodies at multiple levels. No REST slab in the original protocol.

Solution: superior REST slab (axial orientation), 70 mm thick, placed immediately superior to the C1 imaging slice. Inferior REST slab (axial), 60 mm, immediately inferior to the T1 imaging slice.

• Superior slab: saturates descending carotid blood before it enters the FOV from above
• Inferior slab: saturates ascending signal from below

Result: carotid ghost reduced by > 80%. The disc herniations at C4–C5 and C5–C6 are now clearly visible without ghosting artefact.

Time cost: TR = 3500 ms; 2 REST slabs × 15 ms = 30 ms extra per TR → < 1% time increase. At 1.5T: no SAR extension. Clinically imperceptible.

Example 2: TOF Brain MRA — Incorrect Venous Saturation Band Position

Clinical scenario: 3D TOF brain MRA for aneurysm screening. Standard protocol with a superior REST slab for venous suppression. The technologist placed the slab 5 mm superior to the imaging slab (too close; gap insufficient).

Problem: the slab's RF transition zone partially overlaps the superior margin of the imaging volume. Slow arterial blood (in distal branches of the posterior cerebral artery and superior cerebellar artery) is partially saturated → appears as signal voids in the posterior superior circulation → simulating occlusion of distal posterior circulation branches.

Radiologist report: "Possible occlusion of right superior cerebellar artery — clinical correlation required." The patient underwent CTA, which demonstrated a normal superior cerebellar artery.

Correction: increase gap between REST slab and imaging volume from 5 mm to 15–20 mm. The REST slab no longer overlaps the imaging volume's transition zone. Repeat TOF: normal posterior circulation.

Lesson: a gap of < 10 mm between the REST band and the imaging volume produces partial saturation of slow flow at the slab edge — a common cause of false-positive vessel occlusion on TOF MRA.

Example 3: Liver DCE at 3T — Cardiac Ghost Suppression

Clinical scenario: liver dynamic contrast-enhanced MRI (3D VIBE) at 3T; phase direction A-P; TR=4 ms; the left lobe of the liver shows pulsatile cardiac ghosting (bright bands in A-P direction from left ventricular wall motion).

Problem: the heart produces high-signal blood pool pulsation that ghosts in the A-P direction, overlying the left lobe of the liver and creating artefact bands that simulate hypervascular lesions on the arterial phase.

Solution: REST slab, oblique axial, 50 mm thick, placed posterior-left, overlying the left ventricular chamber. The slab is positioned to fully cover the LV while avoiding the posterior liver.

SAR consideration: at 3T, the REST slab adds SAR to a DCE protocol that already has a short TR (4 ms). The scanner calculates: adding the REST slab increases the average SAR. In this case, SAR is within limit because the base TR is very short (low base SAR per TR). No TR extension triggered.

Result: cardiac ghost reduced by 70%. The left liver lobe is now assessable without the confounding artefact bands.

Caveat: the REST slab must be placed carefully to avoid overlapping the posterior liver or diaphragm. Verify on the coronal and sagittal localiser that the liver parenchyma is entirely outside the REST slab.

Example 4: Brain MRS — OVS for Spectroscopy

Clinical scenario: single-voxel MRS of the left temporal lobe for suspected glioma (NAA/Cr ratio assessment). Without OVS: the baseline spectrum shows massive lipid contamination from the scalp temporal fat at 0.9 and 1.3 ppm, completely masking the metabolite peaks at 2.0–3.5 ppm.

OVS strategy at 3T:

• 6 REST slabs surrounding the spectroscopic voxel: bilateral (2 slabs), superior, inferior, anterior, posterior
• Each slab: 30–40 mm thick; flip angle optimised to 130° (instead of 90°) to reduce SAR by ~18% per slab
• Result: SAR within IEC limit at TR=2000 ms

Result: with OVS active, the lipid peaks are reduced by > 95%. The NAA (2.0 ppm), Cr (3.0 ppm), and Cho (3.2 ppm) peaks are clearly resolved. NAA/Cr ratio = 0.8 (reduced; consistent with the glioma diagnosis confirmed at biopsy).

Without OVS: the lipid contamination made the spectrum uninterpretable — would have resulted in a non-diagnostic examination.

Lesson: OVS (outer volume suppression with multiple REST slabs) is mandatory for brain MRS at any field strength. At 3T, SAR management (flip angle reduction, adequate TR) is required to implement 6 OVS slabs effectively.

18. Visual Educational Material

18.1 REST Slab Positioning Diagram (Cervical Spine)

SAGITTAL VIEW — CERVICAL SPINE T2 TSE

Superior REST slab (70 mm):
████████████████████████████████  ← 70 mm slab
░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░  ← 10 mm safety gap
─────────────────────────────────  C1 (first imaging slice)
─────────────────────────────────  C2
─────────────────────────────────  C3
─────────────────────────────────  C4   [IMAGING FOV]
─────────────────────────────────  C5
─────────────────────────────────  C6
─────────────────────────────────  C7
─────────────────────────────────  T1 (last imaging slice)
░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░  ← 10 mm safety gap
████████████████████████████████  ← Inferior REST slab (60 mm)

→ Carotid/vertebral artery signal saturated above and below
→ No ghosting in A-P phase direction within the imaging FOV
→ Safety gap prevents REST band RF profile from affecting C1 and T1 slices

18.2 Saturation Recovery Curve — T1-Dependence

Mz RECOVERY AFTER 90° REST PULSE:

Mz (% of M₀)
100%|                                 Blood (T1=1200ms)
    |                           /─────────────────
 80%|                      /───
    |                 /───          Fat (T1=260ms)
 60%|           /────
    |       /───────────────────────────────────────
 40%|   /───
    |/──
  0%|────────────────────────────────────────────
    0    200   400   600   800  1000  1200  TD (ms)

→ At TD=200ms: Blood Mz = 15% (well-saturated); Fat Mz = 53% (poor saturation)
→ At TD=600ms: Blood Mz = 39%; Fat Mz = 90% (saturation essentially gone)

LESSON: REST slabs are effective for blood suppression (long T1)
        REST slabs are unreliable for fat suppression (short T1)

18.3 REST Slab Error — Overlapping the Imaging Volume

INCORRECT (REST band overlapping FOV):

████████████████████  ← REST band (misplaced)
─────────────────────  Slice 1 (partially saturated → dark band)
─────────────────────  Slice 2 (partially saturated → dark band)
─────────────────────  Slice 3 (normal)
─────────────────────  Slice 4 (normal)
                       ...

→ Slices 1 and 2 show signal void from REST band RF transition
→ May simulate haematoma, mass, or calcification

CORRECT (REST band with safety gap):

████████████████████  ← REST band
░░░░░░░░░░░░░░░░░░░░  ← 10 mm safety gap (essential)
─────────────────────  Slice 1 (normal signal)
─────────────────────  Slice 2 (normal signal)
─────────────────────  Slice 3 (normal signal)
                       ...

19. Evidence Gaps and Ongoing Debate

Optimal REST slab thickness for flow suppression: the rest slab thickness required to achieve reliable blood signal suppression depends on blood velocity, TR, and field strength. No formal prospective study has defined the optimal thickness for each clinical application. The 60–80 mm range for arterial suppression is based on expert practice and early experimental data, not on rigorous flow-velocity-specific optimisation.

REST slab vs cardiac gating for aortic ghost suppression in spine imaging: both REST slab (anterior, over the aorta) and cardiac triggering are used to reduce aortic pulsation artefacts in lumbar spine MRI. No prospective comparative trial has definitively established the superiority of one approach over the other for ghost suppression in routine clinical practice. Most departments use REST slabs as a simpler, faster alternative to cardiac gating.

REST slab flip angle optimisation at 3T: while a 90° flip angle is the standard for REST slabs (maximum saturation), reduced flip angles (70–80°) are sometimes used at 3T to reduce SAR. The optimal flip angle for achieving adequate saturation while minimising SAR contribution has not been systematically studied across clinical applications. Empirical practice varies between centres.

AI-guided REST slab placement: automated planning tools (anatomy-aware AI) could in principle detect the anatomical position of aorta, heart, and fold-over anatomy from the localiser and automatically propose REST slab positions. No commercially validated AI tool for REST slab placement exists at the time of writing; REST slab placement remains an entirely operator-dependent skill.

20. Miscellaneous and Future Directions

Historical milestone: the concept of presaturation in MRI was first systematically described by Edelman et al. in 1987 [1] in the context of flow suppression for cardiac and vascular imaging. The original application — saturating blood entering the imaging FOV — remains the most important clinical use today. The technique was subsequently extended to fold-over suppression, spectroscopy OVS, and spatial fat saturation over the following decade.

Magnetisation transfer contrast (MTC) and REST slabs: REST slabs produce a small magnetisation transfer effect on tissues adjacent to the saturation zone. The RF pulse used for the REST slab is not perfectly spectrally selective and produces secondary excitation of bound water protons in macromolecular structures (myelin, collagen). This MT effect subtly reduces the T1 of adjacent tissues — an effect relevant to quantitative MRI but negligible in clinical routine interpretation.

REST slab integration with AI reconstruction: future AI-based k-space reconstruction algorithms may partially compensate for motion and flow artefacts that REST slabs are currently used to suppress. If AI can reconstruct clean images despite cardiac or respiratory ghosting, the need for REST slabs (and their associated SAR cost) would be reduced. This is an active area of development but not yet clinically implemented.

7T and beyond: at 7T, the severe SAR constraint effectively precludes conventional REST slab use. Alternative approaches — background suppression using inversion recovery schemes, motion compensation via trajectory corrections, and AI-based denoising — are being developed to replace the REST slab function at ultra-high field.