DTIC ADA525985: Breast Cancer Detection Research

9 Pages·0.55 MB·English

by  Defense Technical Information Center

Dual Use ofIntelligence Technologies
Breast Cancer Detection Research
Sam GrantandPeter C. Oleson
On 4April 1995, the television net said, (cid:147)This creative, cooperative
works(cid:146) evening news programs workwill open the achievements of
carried an unusual item. TheActing the IC to the community at large.(cid:148)
Director ofCentral Intelligence, And he added, (cid:147)Our work will also
Adm. William 0. Studeman, held a benefit from the exchange ofideas.(cid:148)
conference the National Pho
press at
tographic Interpretation Center Senators Specter and Kerrey and Dr.
(NPIC) to address the topic ofbreast Blumenthal all added their endorse
cancer. Also participatingwere the ments to the unusual dual use
chairman andvice chairman ofthe intelligence initiative announced by
Senate Select Committee on Intelli Admiral Studeman. This effort, how
After lung breast gence (SSCI), SenatorArlen Specter ever, had started somewhat quietly
cancer, ofPennsylvania and Senator Robert almost a year before.
is the secondleading Kerrey ofNebraska, and Dr. Susan
cancer
Blumenthal, the deputy assistant sec
of death for
cause retary ofHealth and Human Services Breast Cancer NationalAction Plan
American One in and head ofthe Women(cid:146)s Health
women.
Office ofthe Public Health Service.
eight.American The National Cancer Institute
women
(NCI) reported that in 1995 the esti
can expect to be diagnosed Admiral Studeman announced that mated number ofnew cases ofbreast
with breast cancer during the Intelligence Community (IC) cancer in the United States for
had been involved in an effort to women and men would total
her lifetime. identify and promote technologies 182,000. It also estimated that
developed within the IC that had 46,000 Americans would die of
(cid:145)9
promise in the national fight against breast cancer in 1995. After lung
breast cancer. He said this effortwas cancer, breast cancer is the second
to continue with the CIA, National leading cause ofdeath forAmerican
Reconnaissance Office, and Commu women. One in eightAmerican
nity Management Staff(CMS) each women can expect to be diagnosed
contributing to the effort. A sum of with breast cancer during her life
$375,000 had been identified for time. An even higher percentage
this research. will have to undergo medical proce
dures(cid:151)at considerable emotional
Admiral Studeman commented, how and financial cost(cid:151)because ofthe
ever, that (cid:147)breast cancer research is limitations in cancer detection tech
only one area in which our resources nology available today.
can make a difference. We are break
ing down barriers that have kept the In 1994, President Clinton
Sam Grant is in the National Recon talents, expertise, and technical announced the NationalAction Plan
naissance Office. resources ofthe IC and our industry on Breast Cancer, a public/private
contractors from being applied to partnership addressing breast cancer
Peter C. Oleson is President ofPoto problems that lie beyond the scope etiology, treatment, and prevention.
mac Strategies &Analysis, Inc., a ofour traditional national security One goal ofthe plan is to harness the
consulting firm assisting the National mission.(cid:148) Illustrative ofthe changed resources needed to make a significant
Information Display Laboratory and security outlook since the end ofthe impact on breast cancer. Much ofthe
the National Media Laboratory. Cold War, Admiral Studeman also government(cid:146)s work is concentrated in
27
Report Documentation Page Form Approved
OMB No. 0704-0188
Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and
maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,
including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington
VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if it

does not display a currently valid OMB control number.
1. REPORT DATE 3. DATES COVERED
1997 2. REPORT TYPE 00-00-1997 to 00-00-1997
4. TITLE AND SUBTITLE 5a. CONTRACT NUMBER
Breast Cancer Detection Research
5b. GRANT NUMBER
5c. PROGRAM ELEMENT NUMBER
6. AUTHOR(S) 5d. PROJECT NUMBER
5e. TASK NUMBER
5f. WORK UNIT NUMBER
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION
Center for the Study of Intelligence,Central Intelligence REPORT NUMBER
Agency,Washington,DC,20505
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)
11. SPONSOR/MONITOR’S REPORT
NUMBER(S)
12. DISTRIBUTION/AVAILABILITY STATEMENT
Approved for public release; distribution unlimited
13. SUPPLEMENTARY NOTES
Studies in Intelligence, Volume 40, No. 5, Semiannual Edition, 1997, No. 1
14. ABSTRACT

15. SUBJECT TERMS
16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF 18. NUMBER 19a. NAME OF
ABSTRACT OF PAGES RESPONSIBLE PERSON
a. REPORT b. ABSTRACT c. THIS PAGE Same as 8
unclassified unclassified unclassified Report (SAR)
Standard Form 298 (Rev. 8-98)
Prescribed by ANSI Std Z39-18
CancerDetection
With technology,
current
20 percent to 30 percent of
detected
cancers are not
the Nd, the Centers for Disease Con with cancer cells.(cid:148) As a member ofthe
mammograms.
trol and Prevention (CDC), and the Senate Science and Technology Com
Food and DrugAdministration 9~ mittee, Senator Mikulski often
(FDA) in the Department ofHealth addressed issues oftechnology and
and Human Services (HHS). women(cid:146)s health. Dr. Blumenthal,
when she joined the Clinton adminis
Without a reliable cure, early detec tration, took up the challenge.
tion ofmicrocalcifications within the Some ofthese non-normal determi
breast is the most important factor in nations are false positives. The NCI InJuly 1994, the National Informa
rtehdruocuignhgmdaeamtmhso.gTrhaimssis(maopsrtolcyedduornee epsetrifmoartmesedthbaatsetdheorenasruespfiivceiobuisopmsiaesm itniovnitDedisbpylaCyIAL(cid:146)saboOrfaftiocreyo(fNRIeDsLea)rcwhas
for actually
usingX-rays) and clinical breast mograms every cancer and Development (ORD) to prepare
found.
examinations. According to the a background paper for Dr. Blumen
National Institutes ofHealth (NIH) thal describing ofthe advanced
some
all women age 50 and over should While the cost ofa standard screen image-processing techniques that
have an annual mammogram, as this ing mammogram in the United might be applied to the field ofmam
is the population most at risk for States ranges from $80 to $120 with mography. NIDL is a laboratory
breast cancer. In recent years, the per an average of$90, experts at the sponsored by the IC to help leverage
centage ofwomen under 50 also NCI estimate that the (cid:147)real(cid:148) costs, advanced commercial and consumer
obtaining periodic mammograms has including the costs offollowup proce technologies into the government.
increased significantly. dures occasioned by a suspicious
mammogram, average about $120. It
X-ray mammography, however, has is estimated that more than $2 bil DIA digitized some sample mammo
several shortcomings. Its resolution is lion is spent each year on breast grams for NIDL experimentation.
limited. With current technologies, cancer screening in the United TinhteereesxtpienrgimtheanttsDrw.erBelusmufefnitciheanltlaynd
the smallest microcalcifications that States. Improvements in the technol
caapnprboexidmeatteecltyed1b0y0tmhiicsrmoentshtood1a5r0e othgeyreofformea,mhmaovegrbaeepnhyideanntdifiitesdcoassts, ae1xn9p9eH4r.HtsTSvhiecsiotimemmdaigNtetI-epDerLoocfiensmseSidenipgctatelemcbehr
mpriicnrtoends. Mafmilmmoagnrdatmhsenarreevtiyepwiceadllbyy AmcatjioorngPoallasn.under the National niques demonstrated to the group
on during that visit led her to invite
a radiologist on a light box, not dis NIDL to participate in a public sym
staibmliel.arAtwooamnanim(cid:146)sagelatiensttermparemtemro(cid:146)sglriaghmt Request forAssistance pthoesifoulmlotwoibneghmeolndtho.n FCaopritthoilsHsiyllmpo
is difficult for a radiologist to com sium, NIDL(cid:151)with the assistance of
pare to earlier mammograms. In early 1994, Dr. Blumenthal wrote CMS and ORD(cid:151)set up several dem
Differences in angle, resolution, and to other elements ofthe Federal Gov onstrations ofadvanced but
a number ofother factors complicate ernment inquiringwhether they had unclassified image-processing tech
the process and make change detec technologies that might contribute to nologies that were applicable to
tion in awoman(cid:146)s breast over time a the initiative against breast cancer. As medical procedures. DCI
tricky determination. the head ofwomen(cid:146)s health issues for
R. James Woolsey attended the sym
the government, she had called atten posium and spoke ofthe need to
With current technology, 20 percent tion to the need to devote more apply national resources to such
to 30 percent ofcancers are not resources on breast cancer. Members challenges.
detected with mammograms. ofCongress from both parties had
Approximately 11 percent ofmam become increasingly focused on the
classified problem. Senator Barbara Mikulski
mograms are as non-
normal, requiring some type offol ofMaryland, in introducing an initia Technologies ofInterest
lowup action by the patient, such as tive in 1991, stated, (cid:147)If we can build
a repeat mammogram, needle aspira Patriot missiles to intercept Scuds, we A number ofadvanced technologies
tion, biopsy, or other procedure. can do technology that will intercept and demonstrations were displayed
28
CancerDetection
The challenge is that
a
is large,
mainmogram a
high-resolution image, and
at the October symposium, includ the small that indicate effort, ifapplied to digital mammog
cues
ing searches ofhigh-resolution raphy, could revolutionize the
digital images using various high-res breastcancer are subtle and radiologist(cid:146)s ability to analyze large
olution displays. Specific be easily overlooked. mammogram images and improve
can
technologies covered included experi the ability to locate cancers at early
mental high-definition television and 9~ stages and to evaluate better the
head-mounted displays; searching results oftherapy. For example, with
high-resolution digital mammo a head-mounted display, the radiolo
grams; detection ofchanges in serial gist would be able to view a
mammograms; pattern recognition mammogram from any point ofview
tools; and 3-D alignments for detect at high resolution. By turning his
i(nMgRl)esbiroenasstinscmaangsn.etic resonance Atnhaelcyrsiitsicaolffmirastmmstoegprianmtsheiesaorfltyen hberaodwsientahnryoudigrhectthieonX,-rhaey.woBuyldmov
detection ofbreast With ing a (cid:147)spaceball,(cid:148) he could zoom in
esStdhtiixaoaevgspfsaeikltrtsos-iacischlchtootaasponwttoyeinfpbowcteenoht-rrnhcfetkiaofaostqpfsrtuyekiamecsmsttsayiim.ivgsonoetenAslte-snythme,esxpiN.epefmxlPruRacotfIeegiuposteCrtreaematforoieubrocdsjnhieoncng aorttrtlheehonlpaasdeyottaorylsotai(cid:146)sucr.ntamdintTitaoswhcenmoaicatsihmemfmnoiitaoblnygglrmpueoeritg,aicaeysamsat,mla.nlcgayadiTensrochtaacaehedcrneliu.adocrrslhsgmpoaeagr,lwlielilspshettuainhcbrggviuteheienls-sseisau waBagtbhnroneyaoatduttmuwlhsosemde,uieirntong,agvhnaeortidnrlendgldbsaiaeesnt(cid:147)fyrtalrieyolaurn(cid:148)ynnmdzaedeietcnnmanhtdstereamidwodavmileuefm(cid:147)glgrflohoeocfmrvgheteaaanrhsntec,amiga(cid:148)aXmenmt-cserehooareny..
search tasks. There are diverse opin and be easily ovecralnocoekredar.eThese One objective ofthe NIDL effort is
lhiiionukncmeslautanods-istnceogoawrmacvphhauyitolaienbmrlsaeoigfnfettie-ealrcndfoaaplocyyfesstviysisssedtuweeosma.snn,odt ctcehureesimscaoairsnnetimtcosrsuetcairladiliefsfttiicsumtlaetg(cid:151)wetsho.edInentneoctvthaeatticvaen olrtaoagdbyiioormtlaprotargoniorssvifteeessr.atfnordoompseprdeaeetdvieoulnpoalptmhueesnettebcyhnol
tools are needed to assist the radiolo
The National Exploitation Labora
gist in the early phases ofthe search
tory (NEL) in NPIC established a Change detection in serial
process.
testbed project to determine how mammograms. Comparison ofmam
best to perform search exploitation mograms taken over a period oftime
using soft-copy systems. One goal of New medical imaging equipment enables a physician to detect change
that produce high-resolution digi
this testbed project, conducted at the can that signals the early start ofcancer.
NIDL, is to transfer the necessary talmammograms will enable In current practice, a physician views
technology and specifications to com radiologists to migrate from film- two mammograms side by side, men
mercial off-the-shelfvendors to make based mammography to a digital tally matching key features to align
search-capable soft-copy systems environment, allowing the applica and compare the mammograms.
commercially available and therefore tion ofpowerful image-processing Change detection by this method is a
less expensive for the government. tools to assist in analysis. Because challenging task because soft tissue
conventional monitors are low resolu takes on different shapes at different
Avariety oftechnologies are being tion, however, radiologists have to imaging sessions, resulting in mis
experimented with under the testbed choose between viewing a large area aligned mammograms. When cancer
program, including high-resolution ofthe X-ray at low resolution or a is first detectable by mammogram,
displays (5 megapixels), high-defini small area at high resolution. Neither important changes may be quite sub
tion television displays, head- choice is good. The use ofcurrent tle and difficult to detect.
mounted displays, speech-recogni displays to search large images has
tion systems, and tactile input been likened to viewing theworld Computer-assisted change detection
devices. For the medical community, through a soda straw. in mammograms is a dual-use appli
NIDL is investigating how the same cation ofchange-detection
technologies might be applied to digi The same new technologies being techniques used for interpretation of
tal mammography. developed in the NEL(cid:146)s testbed reconnaissance imagery. At the
29
CancerDetection
Mammogram 1993 Mammogram 1994
Difference Image Difference Image Using
Intelligence Community Technology
Figure 1
NIDL, techniques are being devel examine the resulting (cid:147)difference Pattern recognition. Besides change-
oped to align disparate aerial images image.(cid:148) detection approaches, pattern detec
for detection ofchanges such as new tion in imagery is potentially
roads and military sites. For surveil Figure 1 shows two time-separate significant for both intelligence and
lance purposes, it can be imprecise, mammograms and the corresponding medical purposes. Detectingsmall tar
costly, and inefficient to have a difference images before and after gets in reconnaissance imagery is a
human compare each image and alignment. After alignment, regions challenging problem for intelligence
search for change; the holds
same that have changed appearbright in the analysts because militarilysignificant
for evaluation.
true mammogram difference image, while regions that only small
Computer-assisted change detection targets may occupy avery
provide invaluable tool by have not changed appear dark. In the region in alarge image. A novel sys
csaunpplementianng and improving the difference images, the suspicious areas tem beingdeveloped at the NIDL
have been markedwith red circle.
physician(cid:146)s ability to detect change. The difference image befaore alignment attacks this problem by using target
is difficult to interpret because true context. The location ofmost targets
The first step in automating change change is obscured bydifferences due is not random; rather, their position is
detection is precise alignment of to misalignment. The difference image linked to the surroundingarea by
mammograms. Once aligned, two after alignmentshows onlythose dif some logic. For example, transporta
mammograms are digitallysub ferences due to change, and the tion requirements tend to place
tracted, and th~ physician can suspicious region is readilyidentified. buildings near roads. (See figure 2.)
30
CancerDetection
~targets
\
Co ntext
Targets and associated context information for both
a reconnaissance image and a mammogram
Figure 2
Because roads stretch over large screening. For example, microcalcifi Volumetric alignmentfor3-D
areas, they are more readily detect cation position often depends on images. Besides mammogram
able in imagery and can be used to mammalian duct and vasculature X-rays, physicians are increasingly
guide the search for associated small- location. Integration ofthis context using MR imaging for breast cancer
scale targets, in this case buildings. aids in detection ofcancerous diagnosis. MR techniques use con
tumors. An advantage ofthe neural trast agents that enhance the visibility
The automatic target detection tools network approach is that, after the ofcancer in the MRscans. Cancerous
being developed at the NIDL exploit system has learned the task, the net areas appear as bright regions in the
this type ofcontextual information, work can be analyzed to discover the (cid:147)post(cid:148)-MR scans acquired after
matciaocknuironaftgettaahrngeedtdseetfofeficctiiineotnentresfsoctr.hetThmheeedmeotreec stpeexctitfihcatniatturreecoogfnitzheeduansdebreliynigng con Tathduemmi(cid:147)ponoirsssttra(cid:148)-arMteRidoenteowcfitttehhdetbhcyeonc(cid:147)ptorrmaesp(cid:148)-taMraRgienngt.
technique uses an advanced neural important. The system therefore can scan to look for differences. This pro
network(cid:151)a computational processor contribute in several ways to current cess can be facilitated and the
inspired by biology(cid:151)to learn con state-of-the-art mammography, act accuracy increased ifthe two scans are
text automatically through training ing as both an automatic screening subtracted to highlight change. Move
on a large set ofexamples. system and a tool physicians can use ment ofthe patient(cid:146)s breasts,
to uncover subtleties in the breast however, causes the (cid:147)post(cid:148)- and
This detection approach can poten image that may not be easily detect (cid:147)pre(cid:148)- MRscans to be misaligned.
tially be used for mammogram able via traditional methods. False bright areas that appear in the
31
CancerDetection
View 1 View 2
View 3 View 4
Volume Rendering of Difference Volume
After Alignment of Pre and Post Breast MRI Data
Figure3
difference image may cause the physi localization ofa tumor by the doctor, computer-based techniques because
cian to label improperlyareas affected it is necessary to align the (cid:147)pre(cid:148)- and ofthe nonrigid, flexible nature of
by the motion as cancerous. There (cid:147)post(cid:148)- MRdatasets. Alignment of breast tissue.
fore, for precise detection and the scans is a challenging task for
32
CancerDetection
Techniques have been developed in motivated by future commercial mar products driven by the commercial
the IC to align multisensor, two- ket opportunities also have joined in marketplace.
dimensional images. Theyhave been the effort, including Silicon Graph
eovmeprlotyimeedbtaosdeedteocnticmhaagnegsetiankaenscene NicsU,MSROIN,In-CViSsiCono,rpaonrdation. 01W The gfounvdeirnngmefnorttphreovNiIdDesL.limCiotreed
from different viewpoints. These and the NEL have overseen the core
funding provides for the essential
two-dimensional techniques are effort on behalfofthe government.
infrastructure oithe laboratory and
being extended to estimate the non
allows (cid:147)seed money(cid:148) for exploration
rigid three-dimensional deformation
ofthe (cid:147)pre(cid:148)- MRbreast scan to align NIDL Background opfrosbolluetmisonasndtotuhseerasp(cid:146)ploipcearbaitliiotnyaolf
with the (cid:147)post(cid:148)- MRscan. In figure
newly emerging technologies. Other
3(cid:147)p,ret(cid:148)-he taonpdr(cid:147)poowsts(cid:148)-howMsRa sblrieceasftrsocmana.- In establishing the NIDL in 1990, commercial and government part
The bottom shows the differ the government sought to leverage ners also provide funds and
row
slices before and after alignment. the resources ofthe world(cid:146)s commer investment for specific projects appli
ence
cial and university leaders in crucial cable to their needs, for
information technologies where commercialization ofdeveloped tech
Before alignment, the difference
developments and advances increas nologies, and for research and
image (bottom left) falsely indicates
ingly were outpacing the development ofcommercially attrac
the presence ofa change in the upper government(cid:146)s efforts. Recognizing tive technologies.
left corner ofthe breast. This area is the dynamic developments in the
(correctly) not highlighted in the dif commercial marketplace, the objec
ference image after alignment The NIDL is hosted by the David
tive was to take advantage of
(bottom right). SarnoffResearch Center in Prince
commercial developments and
replace the old acquisition paradigm ton, NewJersey, known for its world-
leading developments in high-defini
that slow, increasingly behind
was
tion digital TV, advanced displays,
Partnerships the state ofthe art, commercially
and computing and soft-copy tools.
incompatible, and comparatively
Since the inception ofthe NIDL, expensive. NIDL is a distributed laboratory,
encompassing many industrial and
partnering has been a key factor in academic partners that are leaders in
many ofits programs in support of The NIDL brings together in part their respective fields. The goal is to
government users. This is also the nership commercial and academic obtain the best solution for govern
case with the support being provided leaders in advanced display hard ment needs regardless ofcompany or
to the Office ofWomen(cid:146)s Health. ware, soft-copy information
entity.
Inasmuch as one ofthe purposes of processing tools, computing,
the government(cid:146)s effort is to pro distributed collaboration and com
mote the use ofadvanced munications techniques, and other With today(cid:146)s complex information-
technologies for improved cancer information technologies. The labo technology systems, no one company
detection and treatment, NIDL has ratory focuses on government users(cid:146) can satisfy government program
partnered with several leading medi needs that often are several years in offices(cid:146) range ofrequirements that
cal centers, foundations, and experts, advance ofcommercial markets. One tend to evolve rapidlywith advances
including Dr. Daniel B. Kopans of goal ofthe NIDL is to foster research in information technologies. The
Harvard Medical School, a leading in advanced capabilities in a manner NIDLwith its partnership approach
figure in radiology; the Murray Foun that provides incentives for commer seeks the best solution for the spe
dation; the University ofChicago; cialization. When successful, this cific users with specific needs. The
and the RobertWoodJohnson Medi approach benefits government users laboratory also serves as an agent for
cal School. Commercial partners in future years with commercially advanced research within the aca
interested in the technologies and available technology and low-cost demic community.
33
CancerDetection
The NIDL has a sister lab, the equated to the size ofa conventional Two weeks later, in a speech from
National Media Laboratory (NML), chestX-ray film. the floor ofthe Senate, Senator Ker
which focuses on the critical technol rey challenged the IC to complete
oagricehsivoifngm.asNsMsLtoriasgheoastnedddbaytathe 3M Imagery analysts reacted positively rmaepdiidclaylitcsoemfmfourntittoytruasnesffuelrttoooltshethat
Company in St. Paul, Minnesota. when presented with the option ofa would enhance the latter(cid:146)s ability to
slightly larger display; their initial
NnuImDbLeranodfiNnMfoLrmacotliloanbotreacthenoolnogay- 14-inch-square requirement had fdiegahdtlicnaencoefr.1S2etnoat1o8rmKoenrrtehysswaoidulad
been derived from their experience,
related programs. and the larger format deemed be appropriate. Consequently, the
was NIDL effort went into high gear. In
attractive. The medical-imaging mar lateApril 1995, an HHS working
The N1DL alreadyhad considerable ket dwarfs the government image- formed the
group was to oversee
dual-use experiencewith the medical interpretation market. NIDL ana NIDL-managed work and to help
communitywhen Dr. Blumenthal(cid:146)s lysts were able to converge the plan for clinical trials ofthe technolo
request to the IC was received. Work requirements ofthe two communi gies transferred. FDA help has been
ingwith the NEL, engineers at the ties. This means that the potential requested to help guide the way
NIDL had begun work on a high- market for the high-brightness, high- through the FDA-managed medical
brightness, high-resolution mono resolution monitor for the eventual device approval process. The FDA by
chromatic display that would manufacturer ofthe units will num law oversees mammography facilities
improve image analysts(cid:146) abilities to ber in the hundreds ofthousands. in the United States.
workwith digital reconnaissance Estimates are that the monitors will
imagery. One immediate challenge cost between $5,000 and $10,000,
order-of-magnitude decrease.
was how to make such an advanced an Conclusion
display affordable. At a time when
the intelligence budgetwas facing
downsizing, the development ofa IC Interest Whilework continues at the NIDL
$100,000-plus imagery workstation on the medical technologies, the IC
was deemed too costly. At most, the is mining contractors for more rele
IC might want to procure several Women(cid:146)s health issues, including the vant technologies. Given the more
thousand units ofsuch a display. fight against breast cancer, have liberal security environment ofthe
strongconstituencies on Capitol Hill post-ColdWar period, efforts are
in both parties. The intelligence over being made to capture clinicians(cid:146)
NIDL personnel reasoned that the sight committees, however, have requirements, to communicate these
medical-imaging market, which was never really been involved in such to IC vendors, to form additional
just at the onset ofusing digital tech topics. Interest in the potential appli partnerships, and to obtain funding
nolbgy, might be good leverage for cation to breast cancer detection of for promising technologies. From
the high-brightness, high-resolution technologies developed for intelli this initial case study in dual-use of
display. Resolution and size require gence purposes peaked in the spring intelligence technologies for medical
ments for monitors were discussed of 1995, when Dr. Barry Horton, purposes, government managers are
with radiologists. Theywere the principal deputy assistant secre planning to establish and institute a
unhappywith the 14-inch-square tary ofdefense (communications, process to identify and transfer IC
monitor identified by the reconnais command, control, and intelligence), technologies relevant to other sectors
sance imagery analysts, preferring a invited Senator Kerrey to visit the ofthe government and the nation
14-inch x 17-inch monitor that NIDL and review the research plan. that can use them.
34