**SAR** and **B1+rms** are two radio-frequency (RF) exposure metrics used in magnetic resonance imaging (MRI).

**SAR** – the **Specific Absorption Rate** – is the RF power transmitted (in watts – W) into the patient divided by the patient weight (in kilograms – kg). SAR has units of W/kg. The allowable SAR is limited to the Normal Mode with a whole body SAR of 2 W/kg, or the First Level Controlled Mode to 4 W/kg. These limits apply to the in-built body coil. Different limits apply for dedicated transmit head or local transmit coils.

You can use the **SAR and B1+rms** calculator spreadsheet to investigate the effect of changing patient weight and height, sequence type (Turbo spin echo – TSE, spin echo -SE and gradient echo- GRE), RF flip angle, RF pulse type, number of echoes and slices, TR, and even scanner magnetic field strength B0 on the whole body SAR. Hyperechoes only apply for TSE sequences.

You can read more about SAR and B1_rms in* Essentials of MRI Safety* chapters 5 and 11.

**B1+rms** is a measure of a time-weighted average RF magnetic field exposure, calculated over 10 second intervals, measured in micro-tesla (uT). It is independent of the patient characteristics, but depends upon flip angle, pulse type, number of echoes, slices, and TR. You can use the **SAR and B1+rms** calculator to investigate the influence of these parameters on B1+rms.

B1+rms is useful in determining compliance with implant scanning conditions. The **Fixed Parameter Option (basic)** – FPO:b – is a mode of scanning with RF limits given in terms of B1+rms and peak B1 value – these are shown on the calculator.

Also shown on the calculator are:

Specific Energy Dose (**SED**): the RF energy absorbed per kg. This is the SAR multiplied by scan time – also shown in minutes and seconds. The total absorbed energy in kilo-joules (kJ) is the SED times patient weight.

**Duty cycle **is the fraction of the the pulse sequence for which the RF pulse is on. It is used in the calculations.

**Notes on the use of the SAR and B1+rms spreadsheet.**

The small italic numbers indicate the maximum or minimum values permitted for various parameters. The chosen parameter values will not change automatically but must be entered manually.

The dropdown menus (for B0, sequence, RF type, matrix, NSA and hyperechoes) may not work on all devices. Values may be typed in. Allowable values are shown on the *menus* sheet. The sheet does not yet work on IOS devices. unless you have the *Zoho Sheet* app.

When selecting SE or GRE as sequence type, the number of echoes needs to be set manually to 1.

None of your data entered will be saved.

*Please remember, this is an educational tool only and is not to be used for choosing parameters for human scanning.*

Hi tanks

Does SAR values help at all in measuring the potential heating affects of a passive implant in a clinical setting? From my understanding RF induces e-fields get stopped up at different parts of the body or typically the ends of an implant causing burns. But SAR and induced RF are not essentially proportional, and the astm test doesn’t factor in flowing blood, also the SAR values reported at the scanner is inaccurate. So how does giving a passive stent example “life stent” a 1w/kg rating below umbilical centering even make sense?

Hi, you are correct. It’s a very indirect metric for potential heating. The test methodology indeed excludes perfusion cooling as you rightly say. What the conditions do give is a safe limit. It may be conservative but it poses no additional risk provided the conditions are met. Thanks for your comment.

Thank you so much for you speedy response.

Follow up question (s)

Active and Passive implants have the same testing standards (as far as I know) before they go to market. Do you think they should have different standards? Do you think we should replace SAR with B1 rms completely for implant testing?

B1rms is best for active implants.

Dr. McRobbie,

I have been reading your book Essentials of MRI Safety – it is very useful!

I do have a question.

In the formal definition of B1+rms above, you are saying that the average should be taken over T = 10 seconds. On the other hand, the formula for the rectangular train of pulses becomes (B1+)*sqrt(N*tp/TR). If T is a multiple of TR, this is completely understandable – the number of TR’s appearing in the numerator (integral) and the denominator (T) of the formal definition would cancel each other. But what if T is not a multiple of TR. In this case, I do not see them cancelling each other. Is it that, for 10 seconds, you will typically have multiple TRs and the very last incomplete TR cycle will not have too much of an effect? In this case, the TR-based formula is not exact, but rather is a very close approximation. Am I right?

Thanks.

Hello Sadik,

Thank you for your kind comment. Yes you are correct- it is an approximation.

Thank you very much.