FFDM in the Field: Physicist's Role in the QC of Mammography Laser Printers May Carl R. Keener, Ph.D., DABMP, DABR

Transcription

1 FFDM in the Field: Physicist's Role in the QC of Mammography Laser Printers May 2010 Carl R. Keener, Ph.D., DABMP, DABR MARP Medical & Radiation Physics, Inc.

2 Physicist's Role in the QC of Third-Party Review Work Stations consulting physics group with variety of clients FFDM: laser printers: GE, Fischer, Hologic, Siemens Agfa, Konica, Fuji, Kodak how do we keep all of them in compliance? review of QC manuals development of laser printer evaluation survey incorporation of laser printer requirements into database

3 FFDM laser printers MQSA /FDA guidance FFDM manufacturer guidance printer tests / laser printer manufacturer guidance system tests medical physicist survey & documentation summary

4 we re digital why perform printer QC? MQSA requires it hard copy used for prior films archiving final interpretation occasionally ACR submission without QC, things degrade

5 FFDM laser printers MQSA /FDA guidance FFDM manufacturer guidance printer tests / laser printer manufacturer guidance system tests medical physicist survey & documentation summary

6 caveat This lecture involves MQSA. Use with care.

7 MQSA Guidance 21 CFR (e)(6) Quality control tests other modalities. For systems with image receptor modalities other than screen-film, the quality assurance program shall be substantially the same as the quality assurance program recommended by the image receptor manufacturer, except that the maximum allowable dose shall not exceed the maximum allowable dose for screen-film systems in paragraph (e)(5)(vi) of this section.

8 FDA Guidance allowed digital mammography in 2000 FFDM manufacturer supplied QC manual to FDA initially allowed final interpretation from laser film only later allowed soft copy mammography interpretation requires MP to follow QC program in FFDM manufacturer s QC manual varies with manual version & alternative standards

9 how does FDA address laser printers? To avoid unnecessary follow-ups by MQSA inspectors, we urge all medical physicists to review the FFDM manufacturer s QC manual and, where necessary, the laser printer QC or operator s manual to determine the appropriate testing. They then need to clearly document the testing of the laser printer in their reports. FDA Inspection News Update February 4, 2008

10 FFDM vendors Hologic GE Selenia Senographe 2000D Senographe DS Essential Siemens Novation Fischer Senoscan Fuji FCRm

11 laser printer vendors Agfa DryStar 4500M DryStar 5500, 5503 Konica-Minolta DryPro 793 Fuji DryPix 4000 DryPix 5000 / 7000 Kodak DryView 8610 Carestream DryView 6800

12 available tests FFDM & printer tests laser printer tests MAP phantom low-contrast resolution SMPTE pattern artifacts high-contrast spatial resolution MAP phantom gray-scale SMPTE pattern geometry artifacts mechanical QC

13 FFDM laser printers MQSA /FDA guidance FFDM manufacturer guidance printer tests / laser printer manufacturer guidance system tests medical physicist survey & documentation summary

14 Hologic printer tests MAN rev 006 artifact test tech (weekly) & MP digital flat field both film sizes (if supported) SMPTE pattern tech (weekly) small film LD (90%) +/ MD (40%) +/ DD (40%- 10%) +/ MAP phantom laser printer and RWS tech (weekly) & MP small film score 5 fibers / 4 specks / 4 masses 4.5 fibers / 4 specks / 3.5 masses if SNR >40 background OD> 1.20

15 Hologic printer tests MAN rev 006 artifact test tech (weekly) & MP digital flat field both film sizes (if supported) SMPTE pattern tech (weekly) small film LD (90%) +/ MD (40%) +/ DD (40%- 10%) +/ MAP phantom laser printer and RWS tech (weekly) & MP small film score 5 fibers / 4 specks / 4 masses 4.5 fibers / 4 specks / 3.5 masses if SNR >40 background OD> 1.20 MAN rev 008 MAN rev 001 MAN rev 003 artifact test tech (weekly) & MP digital flat field both film sizes (if supported) SMPTE pattern tech (weekly) small film LD (90%) +/ MD (40%) +/ DD (40%- 10%) +/ MAP phantom laser printer or RWS tech (weekly) & MP small film score 5 fibers / 4 specks / 4 masses 4.5 fibers / 4 specks / 3.5 masses if SNR >40 no OD requirements

16 other FFDM manufacturers defer to printer manufacturer GE Siemens Novation Senographe 2000D rev 0 SPB ver 05 GE Senographe DS EN follow printer manufacturer QC manual print MAP phantom follow printer manufacturer QC manual MAP phantom tech does not print MP prints if AWS fails tech (weekly) & MP score 4 fibers / 3 specks / 3 masses no disk no OD requirements printer check daily when printer used SMPTE pattern plot 11 ODs 3 LUTs Dmin = 0.20 Dmax = 3.50

17 other FFDM manufacturers defer to printer manufacturer Fischer Fuji FCRm Senoscan 3 rd edition Feb 2007 follow printer manufacturer QC manual no printed phantom requirement follow printer manufacturer QC manual if no printer manufacturer QC manual: set up weekly program like film/screen us TG18-QC or TG18-PQC weekly MAP phantom if printer used for final interpretation background OD > /-0.20 DD +/- 0.05

18 FFDM printer requirements Hologic GE Siemens Fischer Fuji CRm small large small large small large small large small large MAP phantom (score)?? MAP phantom (OD) SMPTE pattern artifact follow printer manu.????

19 FFDM laser printers MQSA /FDA guidance FFDM manufacturer guidance printer tests / laser printer manufacturer guidance system tests medical physicist survey & documentation summary

20 gray-scale display function 1 based on film/screen sensitometry base+fog lowest step (OD=~0.20) mid-density speed index step closest to (but greater than) 1.20 OD density-difference contrast index difference between step closest to 2.20 OD step closest to (but greater than) 0.45 OD top & bottom of linear portion of characteristic curve

21 film-screen processor QC action limits establish baselines with 5-day average MQSA limits: base+fog > 0.03 over baseline do not use processor mid-density ± 0.10 from baseline contact service ± 0.15 from baseline do not use processor density-difference ± 0.10 from baseline contact service ± 0.15 from baseline do not use processor developer temperature typically 95.0 F ± 0.5 F from baseline do not use processor if out of limits repeat if still out of limits do not use processor

22 laser printer sensitometry gray scale pattern generated internally in laser printer may use internal photometer to read and report ODs printers uses different calibration strip for self-calibration

23 gray-scale Agfa 5-step gray-scale Base + Fog, Low, Mid, High, Dmax establish aim values must meet Agfa operating levels B+F = 0.22 ± 0.03 MD = 1.40 ± 0.15 DD = HD-LD = 1.90 ± 0.15 Dmax = 3.70 ± 0.25 daily test image must meet aim values B+F = aim ± 0.03 MD = aim ± 0.15 DD = aim ± 0.15 Dmax = aim ± 0.25 uses internal photometer one film size only?

24 gray-scale Konica 6-step gray-scale 1, 2 (LD), 3, 4 (MD), 5 (HD), 6 establish benchmarks all 6 steps Contrast = HD LD (5 2) daily test image action / control limits 1 aim ± 0.03 / (LD) aim ± 0.10 / aim ± 0.10 / (MD) aim ± 0.10 / (HD) aim ± 0.10 / aim ± 0.20 / 0.20 Contrast aim ± 0.15 uses external photometer one film size only?

25 gray-scale Fuji 17-step gray-scale uses 4 steps only 1, 5, 10, 15 control limits set by Fuji weekly test images ± ± ± 0.15 uses external photometer all film sizes used for mammo

26 gray-scale Carestream 21-step gray-scale edge - automatic densitometer center - manual densitometer establish operating levels Dmin = step 1 LD = step closest & 0.45 MD = step closest & 1.20 HD = step closest & 2.20 weekly test image must meet aim values Dmin = operating level ± 0.03 LD = operating level ± 0.07 MD = operating level ± 0.15 HD = operating level ± 0.15 uses external photometer all film sizes used for mammo?

27 gray-scale response function - Carestream annual physicist test 21-step gray-scale Dmin = densities 2-20 target & tolerances based on Dmin Dmax = 3.60 ± 0.16

28 Agfa QC mammo test image weekly spatial resolution resolve each set of smallest dots magnifier & bright view box no artifacts low-contrast resolve 0%/5% & 95%/100% ovals 5500 & 5503 only annual geometric horizontal (A) & vertical (B) distances ± 1% of reference A/B ratio 0.01 one film size only?

29 Konica mammo QC image daily sharpness horizontal & vertical line pairs SMPTE pattern recommended at setup generated from FFDM 90 % 0.42 ± % 1.55 ± % 2.54 ± % 3.60 ± 0.15

30 Fuji QC images annual resolution spatial resolution test pattern 50 micron lp resolved in both directions mammo set to high resolution geometry geometry test pattern horizontal (H) & vertical (J) dimensions within 1% of standard all mammo film sizes artifacts mechanical no transport errors no abnormal noises

31 Carestream multipurpose QC test image semiannual (tech) annual (MP) low-contrast 6 fibers 5 speck groups 5 masses all must be resolved spatial resolution all line pairs visible all mammo films sizes?

32 Carestream QC images gray QC test images light gray & dark annual artifact annual geometry horizontal & vertical ± 1% of reference straight ± 2% all mammo films sizes? phantom image must pass FFDM requirements GE, Fischer, Fuji 4 fibers 3 speck groups 3 masses Hologic, Siemens 5 fibers 4 speck groups 4 masses Hologic if SNR > fibers 4 speck groups 3.5 masses

33 laser printer manufacturer QC requirements Carestream 6800 Agfa 4500/5500 Fuji Konica 793 Hologic small large small large small large small large small large gray scale?? LCD? HCR?? geometry? artifacts? mechanical SMPTE pattern? MAP phantom?

34 laser printer QC manuals Agfa DryStar 4500M 2811B EN DryStar 5500, D EN Konica-Minolta DryPro 793 Version 1.1 Feb 06 Fuji DryPix th Edition Oct 2008 DryPix 5000 / st Edition Mar rd Edition Dec th Edition Oct 2008 Kodak DryView 8610 V 1.0 (5/01) Carestream DryView 6800 QC Manual 2007

35 FFDM laser printers MQSA /FDA guidance FFDM manufacturer guidance printer tests / laser printer manufacturer guidance system tests medical physicist survey & documentation summary

36 printing films multiple FFDM sources AWS(s) RWS(s) PACS multiple DICOM sources CT, MR, US, CR, etc. different film sizes different LUTs different Dmax different resolution printing file vs. image default vs. WL

37 system tests most tests only test the printer gray-scale printer test patterns resolution low-contrast geometry OD artifacts very few system tests Carestream requires MAP phantom score only, no ODs Fuji FCRm requires MAP phantom minimum OD Hologic requires uniform image artifacts only Hologic requires SMPTE constancy only, no aim values Konica DryPro recommends SMPTE OD aim values

38 SMPTE pattern suggestions generate SMPTE pattern at FFDM console origin, WL, high-resolution caveats send to printer (all film sizes) observe LCD squares: 0% / 5% & 95% / 100 observe gray levels use Konica DryPro OD recommendations: 90 % 0.42 ± % 1.55 ± % 2.54 ± % 3.60 ± 0.15 note: Dmax > 3.45

39 MAP phantom suggestions image MAP phantom on both receptor sizes send to printer (all film sizes) does it print? correct size? ACR recommends true size phantom ± 25% score? Dmax > 3.45? Agfa 3.70 ± 0.25 Carestream 3.60 ± 0.16 Konica (SMPTE) 3.60 ± 0.15 Siemens 3.50 ± 0.10 phantom background OD ? ACR recommendation artifacts? note: sending file vs. sending image

40 mammogram suggestion observe printed mammograms contrast & densities correct? Dmax > 3.45? Dmin> 1.0? old film/screen recommendation true size? ACR digital mammo standard recommends true size artifacts?

41 uniform image suggestion Hologic artifact test other FFDM may need to be sent from PACS TG18 UN 10 & UN80

42 test multiple sources send phantoms, SMPTE patterns and test patterns from various AWS, RWS and PACS used to print digital mammograms check Dmax check Dmin check MD

43 FFDM laser printers MQSA /FDA guidance FFDM manufacturer guidance printer tests / laser printer manufacturer guidance system tests medical physicist survey & documentation summary

44 survey scenarios perform tests required by FFDM QC manual perform all available tests acceptable for Hologic including: what if other FFDM shares the printer? perform tests required by printer QC manual what if Hologic FFDM shares the printer? all available printer QC test patterns all available film sizes artifacts SMPTE pattern from FFDM MAP phantom look at printed mammogram perform tests as deemed appropriate by medical physicist for each laser printer used for mammography

45 MARP laser printer evaluation sample

46 printer info

47 gray scale

48 SMPTE pattern

49 artifacts

50 low contrast detectability

51 high contrast resolution

52 geometric accuracy

53 mechanical checks

54 MAP phantom

55 technologist QC

56 laser printer report

57 FFDM laser printers MQSA /FDA guidance FFDM manufacturer guidance printer tests / laser printer manufacturer guidance system tests medical physicist survey & documentation summary

58 summary MQSA requires that physicist survey laser printers used for FFDM required tests vary based on FFDM manufacturer & printer manufacturer required tests may miss QC problems, especially incorrect film densities site may be better served by performing all available tests instead of minimum required system tests (sent from FFDM to printer) are particularly useful

59 FFDM in the Field: Physicist's Role in the QC of Mammography Laser Printers May 2010 Carl R. Keener, Ph.D., DABMP, DABR MARP Medical & Radiation Physics, Inc.