ScanMed PROCURE™ Coil – SNR Comparison to the GEMS 32 Channel Solution on the 750W

Project Goals Overall:

A primary purpose for developing the PROCURE™ coil was to provide a relatively low cost, diagnostic antenna for prostate cancer screening without the use of an endo-rectal coil. To be effective and put into wide-spread use, this antenna set must exhibit a measurable improvement of signal-to-noise ratio (SNR), and hence resolution and sensitivity improvement over the existing external MRI antenna solutions. It must be easy to use and conform to a wide variation in patient habitus such as thin patients and those with protruding stomachs.

Having ‘checked the boxes’ on the physical and ease-of-use characteristics, the goal below is to demonstrate the performance and clinical results of the PROCURE™ coil in comparison to the 32 Channel GEMS solution on the 750 and/or 750W platforms.

The examples below are provided for educational and informational use only. The interpretation of the referenced images are a result of consultation with participating radiologists and presented by Dr. Randall Jones, PhD, a MRI physicist and electrical engineer; hence, additional radiological specialists’ views are not only welcome, but encouraged.

Protocol Overview:

We recommend the following series using our coil: (Our apologies to the Siemens users as this is using the GE vernacular. We’ll edit with updated Siemens sequence names soon.)

1. 3 Plane Localizer using FSGRE or SSFSE or Siemens Fast Gradient sequence

2. Obl Sag T2 for organ localization with 1mm spacing between slices

3. Obl Ax T2 throughout the organ with zero skip or spacing between slices

4. Obl Cor T2 throughout the organ with zero skip or spacing between slices

5. Diffusion weighting (DWI – SE/EPI) or Siemens equivalent

6. ADC and/or exponential ADC mapping

7. Dynamic contrast scanning and/or 1H Spectroscopy – not done in this study

Volunteer Images for Review and Comparison:

One volunteer was scanned with the ScanMed PROCURE™ coil on a GE 750W system (Series 4-21), followed by repeating a number of the series (22-30) using identical parameters with the GEMS coil on the same patient. Initially, the PROCURE™ coil was scanned using the standard protocol developed for use with the 32 Channel GE Anterior Torso/ Posterior table coil combination. In several additional series, we varied resolution parameters to determine the “clinical limits” of the PROCURE. After completing the series of test scans, and pushing the protocol, we then repeated the base line and more pushed protocols with the GEMS solution.

PROCURE™ Scanning by Series Number:

1. 3PLoc w/SSFSE, 48FOV

2. Fast B1 Map (or CAL Scan) Not copied

3. FRFSE-XL 111 Obl Sag T2 FOV 24X24 1/1, TR 4000, TE 131, BW 31kHz, Th 3/1, 288X224, 3NEX, PURE, Options: FC, Sat SI, NP, ED, TRF, SPFIP, Accel=1.5, freq=A/P

4. FRRSE-XL 111 Obl Ax T2: ALL SAME AS ABOVE except TR 3000, 4 NEX and 320X320 Th 3/0.0, Options: FC, Sat RL, NP, ED, TRF, Accel=1.5, freq=A/P – 2:18.

5. Pushed Obl Ax T2 Same as no. 4 above but with 2NEX, th2/0.0 – time 1:15

6. Pushed Obl Ax T2 Same as no. 4 above but with 3NEX, 22X22 FOV, th2/0.0 – time 1:50

7. Pushed Obl Ax T2 Same as no. 4 above but with 2NEX, 24X24FOV, th3/0.0 – time 1:42

8. Turned off Acceleration

9. Pushed Obl Ax T2 Same as no. 4 above but with 2NEX, 24X24FOV, th3/0.0 – time 1:42

10. Turned off Acc and SATs

11. Ax DWI SE/EPI, b100/b1000 FOV 26X26, 0.5FOV, TR 3000, TE Min,Th 4/0.0, Freq R/L, No Internal Cor, Sampling 1, Recon 1, Diff= ALL, 160/80 6NEX @ 100, 16NEX @1000: 3:21

12. Same as above but with reduced slice thickness of 2.5mm: 3:21

18. 3P Loc repeated to turn on clinical mode and Bitmap (20)

21. Obl AX T2: Chose nicest presentation at highest resolution: TR 3135/129, 2/0.0, 24X24, 31kHz, 320X320, 4NEX – 2:24

Discussion:

We observed some line artifacts on some of the T2s, so we explored a number of variations to determine cause. For illustration, series 8-11 use various FOV, NEX, and SATs, as well as additional series not copied from scanning this volunteer. Even though the coil sensitivity can produce diagnostic images at 2 or 3 NEX, it appears as if any number of averages different then 1 or 4 leads to line artifacts. The averaging of 4 acquisitions seems to nearly eliminate them (ser 21).

Scanning with the 32 Ch GEMS Torso/Spine combination:

1. 3PLoc

2. Bitmap

3. SAME as Ser. 7: FRFSE-XL 111 Obl Ax T2 FOV 24X24 1/1, TR 3000, TE 131, BW 31kHz, Th 3/0, 320X320, 4NEX, PURE, Options: FC, Sat RL, NP, ED, TRF, Accel=1.5, Freq A/P – 2:18

4. SAME as Ser. 10: 2 NEX and 320X320 Th 3/0.0 – 1:42

5. Note same line artifacts as with PROCURE – Averaging artifact

6. Ignore. Mistaken phase direction

7. SAME as Ser. 12: Ax DWI Focus b100/1000 FOV 26X13, TR 3000/72.6, 167kHz, 4/0 Matrix 160/80, 6NEX @ 100, 16 NEX @ 1000

8. Ignore. Ran twice

9. SAME as Ser. 8: Pushed Obl Ax T2 with 2NEX, th2/0.0 – time 1:15

10. Note same line artifacts as with PROCURE – Averaging artifact

30. SAME as Ser. 13: AX DWI Focus b100/1000 SE/EPI FOV 26X13, TR 3000/73.1, 1/1, 10X80, 6X16 NEX, 250KHz, 2.5Th/0.0, 3:21

Discussion:

Provided are the comparison images from the same volunteer using the two coils and running identical protocols. Series 7 and 24 are Obl AX T2s at 4NEX and 24X24 FOV, the least SNR demanding of the protocols where little difference can be discerned.

Series 8 and 29 compare the “pushed T2s” using 2 NEX at 2mm slices and the noise begins to become more pronounced in the GEMS solution compared to the PROCURE. Line artifacts are consistent in both.

Finally, FOCUS DWIs from the two coils were compared using identical protocols; the first series pair (12 and 27) were at 4mm slice thickness and the second series pair (13 and 30) were at 2.5mm slice thickness. Although the SNR differences become readily apparent in these sequences, we took data to coarsely quantify. SNR data was taken from the image using the simple method of Signal mean divided by noise Standard Deviation (See table below). We used a consistently sized ROI that fit well within the center of the prostate for the signal value, and the same sized ROI within the femoral head as the noise SD. Please refer to the JPG images attached to verify the SNR ROI placement.

Conclusions:

1. We have demonstrated that one can see marked improvement in SNR using the ScanMed Procure coil over the OEM external solution; and hence, encourage its further exploration as to the clinical viability in lieu of using an endo-rectal probe.

2. In our next studies, we shall further explore the feasibility of increasing resolution at the expense of scanning time to demonstrate whether the PROCURE can yield diagnostic results in the signal starved DWI enhanced diffusion protocol.