-
8/10/2019 0004-3702(78)90015-2
1/28
ARTIFICIAL INTELLIGENCE 4 5
A M o d e l B a s e d M e t h o d f o r C o m p u t e r A i d e d
M e d i c a l D e c i s i o n M a k i n g
S h o l o m M . W e i s s C a s i m i r A . K u l i k o w s k i a n d
S a u l A m a r e l
D e p a r t m e n t o f C o m p u t e r S c i e n c e , R u t a e r s - - T h e S t a t e U n i v e r s i t y ,
N e w B r u n s w i c k , N J 0 8 90 3 , U . S . A .
A r a n S a f i r
I n s t i t u t e o f C o m p u t e r S c i e n c e , M o u n t S i n a i S c h o o l o f M e d i c i n e ,
N e w Y o r k , N Y 1 0 0 2 9 , U . S . A .
R e c o m m e n d e d b y N . S . S r i d h a r a n
A B S T R A C T
A gene ral met hod o f computer-assiste d medical decision-making has been developed based on causal-
associational network CA SN ET ) models of disease. ,4 CA SN ET model consists of three main
components: observations of a patient, pathophysiological states, and disease classifications. As
observations are recorded, they are associated with the appropriate states. States are causally
related, for min g a networ k that summari zes the mechanisms o f disease. Patterns o f states in the
netw ork are linked to individual disease classifications. Recom menda tions for broad classes o f
trea tment are triggered by the appropriate diagnostic classes. Strate gies o f specific treatm ent
selection are gui ded by the individual patt ern o f observations and dia~qnostic conclusions. This ap-
proac h has been applied in a consultation prog ram for the diagnosis and treatment of the glaucomas.
n t r o d u c t i o n
I n t h e p r e s e n t p a p e r a g e n e r a l a p p r o a c h t o s t r u c t u r i n g m e d i c a l k n o w l e d g e f o r
c o m p u t e r - a i d e d d i a g n o s i s a n d t h e r a p y i s p r e s e n t e d . W e h a v e d e v e l o p e d a r e-
p r e s e n t a t i o n t h a t m o d e l s d i s e a s e p r o c e s s e s a s a c a u s a l - a s s o c i a t i o n a l n e t w o r k
( C A S N E T ) . T h i s m o d e l - b a s e d m e t h o d h a s b e e n u s e d s u c c e s s f u l l y i n d e s i g n i n g a
c o n s u l t a t i o n p r o g r a m f o r t h e d i a g n o s i s a n d l o n g - t e r m t r e a t m e n t o f t h e g l a u c o m a s .
T h e c o n s u l t a t i o n p r o g r a m u s e s a s e t o f g e n e r a l d e c i s i o n - m a k i n g s t r a t e g ie s in c o n -
j u n c t i o n w i t h a c l a s s o f c a u s a l - a s s o c i a t i o n a l m o d e l s [1 , 2 ] . I n t h i s p a p e r , e x a m p l e s
w i l l b e g i v e n f r o m a C A S N E T m o d e l o f g l a u c o m a . H o w e v e r , t h e m o d e l r e p r e s e n t a -
t i o n a n d d e c i s i o n - m a k i n g p r o c e d u r e s a r e g e n e r a l i z a b l e t o o t h e r m e d i c a l d o m a i n s .
D i a g n o s t i c p r o b l e m s h a v e o f t e n b e e n c a s t i n t o a p a t t e r n r e c o g n i t i o n o r s t a t i s t i c a l
d e c i s i o n t h e o r y f r a m e w o r k . C o m p u t e r r e p r e s e n t a t i o n i s n o t d i f f ic u l t , a n d a s a
r e s u l t m a n y w e l l - k n o w n m e t h o d s s u c h a s t h o s e b a s e d o n B a y e s t h e o r e m h a v e
rt i f icial intel l igence 11 1978), 145- 172
Copyright t~ 1978 by Nort h-Ho l lan d Publ ish ing Company
-
8/10/2019 0004-3702(78)90015-2
2/28
| 46 S .M . WEISS C . A . KULI KOW SKI A. SAFIR AND S. AMAREL
b e e n u s ed [ 3 - 5 ] . T h e d i ff ic u l t ie s w i t h a p p l y i n g t h e s e m e t h o d s s u c h a s s c a r c i t y o f
s ta t i s t ic s , a n d t h e u s e o f i n v a l id a p p r o x i m a t i o n s ) a r e a l s o s u f fi c ie n t ly p e r s i s te n t
t h a t a l t e rn a t i v e a p p r o a c h e s h a v e b e e n s o u g h t . I n m a n y m e d i c a l a r e a s , e x i s t in g
k n o w l e d g e c o u l d e n h a n c e t h e d e c i s i o n - m a k i n g c a p a b i l i t i e s o f a d i a g n o s t i c s y s t em .
T h e r e a r e m a n y u s e f u l d e c i s i o n - r u l e s s p e c i fi c t o a g i v e n m e d i c a l a p p l i c a t i o n t h a t
t h e p h y s i c i a n d i r e c t l y a p p l i e s i n h i s r e a s o n i n g .
I n t h e p a s t f e w y e a r s , t h e r e h a s b e e n i n c r e a s e d i n t e r e s t in th e a p p l i c a t i o n o f
a r ti fi c i a l i n te l l i g e n c e A I ) t e c h n i q u e s t o m e d i c a l d e c i s i o n - m a k i n g . A I t e c h n i q u e s
a t t e m p t to c a p t u r e d e c i s i o n - m a k i n g r u l e s e x p l ic i tl y , w h i l e s t a t i s t i c a l m e t h o d s m a y
e x t r a c t t h e m i m p l ic i t l y f r o m a c c u m u l a t e d s a m p l e e x p er i en c e . T h e A I a p p r o a c h e s
i n te n d t o o v e r c o m e s o m e o f t h e l i m i t a t io n s o f p u r e l y s t a t i s t ic a l m e t h o d s b y d e v e l o p -
i ng a m o r e s t r u c t u r e d r e p r e s e n t a t i o n o f t h e d i a g n o s t i c a n d t h e r a p y s e le c t i o n
p r o b l e m s . A p r o g r a m w h i c h u s e s d e c i s i o n s t r a te g i e s b a s e d o n e x p l ic i t r e p r e s e n t a -
t io n s o f m e d i c a l k n o w l e d g e c a n m o r e e a s i ly i n c o r p o r a t e e v o l v i n g c h a n g e s i n it s
k n o w l e d g e b a s e , i n d e p e n d e n t l y o f t h e r e a s o n i n g s t r a te g i e s . I t c a n a l s o i n c o r p o r a t e
t h e r e su l ts o f c l in i ca l e x p e r i e n c e b y m a t c h i n g t h e m o r e e x p l ic i t p a t t e r n s o f r e a s o n i n g
t o t h e d ec i s io n s a n d o p i n i o n s o f p h y s i c i a n s . S u c h s y s t e m s a r e m o r e l i k e ly t o b e
a c c e p te d b e c a u s e t h e y a r e e x p re s s e d i n a d e c i s i o n - m a k i n g c o n t e x t f a m i l ia r t o th e
c l i ni ci an . A s t r u c t u r e d r e p r e s e n t a t i o n c a n a l so p e r m i t t h e f o r m u l a t i o n o f c o m p l e x
h y p o t h e s e s t h a t e x p r e s s p r o g r e s s i o n a n d s e v e r i t y o f d i s e as e . S o m e r e s e a r c h e r s
h a v e a t t e m p t e d t o in c r e a se t h e s co p e , a c c u r a c y , a n d e x p l a n a t i o n c a p a b i l i t i e s o f
t h e i r s y s t e m s b y i n c r e a s i n g s t r u c t u r e , w h i l e s ti ll p r e s e r v i n g a s ta t i s t ic a l f r a m e -
w o r k [ 6] . O t h e r s h a v e r e l ie d o n l o g ic a l a n d s e m a n t i c e n c o d i n g s o f c o n t e x t u a l
k n o w l e d g e , w i t h i n a n a r ti f i c ia l i n te l l i g e n c e f r a m e w o r k [ 7 -1 0 ] .
S e v e ra l f u n d a m e n t a l A I i ss u es a r e ra i s ed b y m e d i c a l d e c i s i o n - m a k i n g p r o b l e m s .
O n e im p o r t a n t i s su e c o n c e r n s t h e d e v e l o p m e n t o f r e p r e s e n t a t i o n s t h a t a r e p o w e r f u l
e n o u g h t o c a p t u r e a c o m p l e x a n d c h a n g i n g k n o w l e d g e b a s e in a r e a l is t i c t a s k
d o m a i n . T h e r e h a s b e e n a n i n c r e a s e d i n t e re s t in r e c e n t y e a r s i n d e v e l o p i n g A I
s y s t e m s t h a t u s e e x p e r t k n o w l e d g e in a v a r i e t y o f a p p l i c a t i o n a r e a s [ l l - 1 4 ] .
M e t h o d s o f a c q u i r i n g k n o w l e d g e f r o m e x p e r ts , c h o o s i n g a p p r o p r i a t e l e ve ls o f
a b s t r a c t i o n a n d r e s o l u t i o n f o r d e s c r i b i n g a g i v e n p r o b l e m , a n d t h e ch o i c e o f
c o m p u t e r r e p r e s e n t a t i o n o f t h e k n o w l e d g e b a s e a r e a l l p r o b l e m s t h a t im m e d i a t e l y
a r is e in d e v e I o p i n g s u c h s y s t e m s . T h e y a r e c l o s e ly l i n k e d t o t h e c o n t r o l s t r a te g i e s
o r m e t h o d s u s e d t o p r o d u c e i n t e r p r e t a t i o n s f o r i n d i v i d u a l c as e s. F u n d a m e n t a l t o
m o s t s u c h c o n t r o l s t r a te g i e s is t h e c a p a b i l i t y o f a p p r o x i m a t e r e a s o n i n g . T h i s is
n e e d e d to m a n a g e t h e m u l t ip l e h y p o t h e s e s t h a t c a n b e g e n e r a t e d f r o m a l a rg e
a n d c o m p l e x k n o w l e d g e b a s e , w h i c h i n c l u d e s s t a t e m e n t s a t d i f fe r e n t l ev e ls o f
u n c e r ta i n ty . O n c e d e c is io n s a r e r e a c h e d , p r o d u c i n g e x p l a n a t i o n s b e c o m e s a n
i m p o r t a n t t a s k i f t h e a c c e p t a b i l i t y o f t h e s y s t e m i s t o b e e n h a n c e d . P r a c t i c a l i s s u es
o f i m p l e m e n t a t i o n f o r t h es e l ar g e k n o w l e d g e - b a s e d s y s t em s i n c l u d e : e a s e o f
k n o w I e d g e m a n a g e m e n t u p d a t i n g ) , e ff ic ie n cy , c h o i ce o f l a n g u a g e s , a n d t r a n s f e r -
a b i l it y i n to p r a c t i c a l u s e b o t h in t h e o r i g in a l d o m a i n a n d o t h e r s i m i l a r o n e s .
-
8/10/2019 0004-3702(78)90015-2
3/28
M O D E L B A S E D M E D I C A L C O N S U L T A T I O N 4 7
T he p r e s e n t pa pe r de s c r i be s t he m e t ho ds o f r e p r e s e n t a t i on a nd i n t e r p r e t a t i on
de ve l ope d w h i l e bu i l d i ng a know l e dge - ba s e d s ys t e m fo r m e d i c a l c ons u l t a t i on . I n
t he c ou r s e o f de s c r i b i ng t he s e m e t h ods , s pe c if ic so l u t i ons t o s om e o f
t h e i s s u e s
r a i sed above a re of fe red .
C a u s a l- A s so c ia t io n a l N e t w o r k C A S N E T ) M o d e l s
A causa l -a s soc ia t i ona l ne tw ork is a pa r t i cu la r type of semant ic ne two rk [15]
des i gne d to :
( a) desc r ibe dynam ic processes in t e rms of ( loop- f ree ) causa l r e la t ionsh ips
amo ng a se t o f in te rna l va r iab le s ;
(b) r e la te th i s desc r ip t io n to ex te rna l va r iab le s tha t a re con s ide red to be mani -
f e s t a t i ons o f the i n t e r na | p r oc e ss e s ; a n d
(c) desc r ibe va r ious c la s s i f i ca t ions imp ose d on the dyn ami c processes .
C A SN E T m ode l s c a n be u s e d t o de s c ri be m a ny d i f fe r e n t c om pl e x p roc e ss e s,
a l t h ough w e ha ve de ve l o pe d t he m t o de s c r i be pa t hop hys i o l og i c a l p r oc es s e s o f
d i sease [2] . Knowledge , in our scheme , i s r epre sen ted by th ree types of da ta
e l e m e n t s , c o r r e s pond i ng t o t he t h r e e k i nds o f de s c r i p t i on ou t l i ne d a bove . T he s e
a r e : obs e r va t i o ns o f t he pa t i e n t ; pa t hophys i o l og i c a P ' s t at e s; a nd d i a gnos t ic ,
p rogn os t i c , a nd the r apeu t ic ~categor ie s . Obse rva t ions a re the d i rec t ev idence
ob t a i ne d a bou t a pa t ie n t . Pa t ho phys i o l o g i c a l st a te s a re i n t e r m e d i a t e c ons t r uc ts
t ha t de s c r i be i n t e r na l c ond i t i ons a s s um e d t o t a ke p l a c e in t he pa t i e n t ; t he y s um -
mar iz e r e su lt s f rom man y d i f fe ren t obse rva t ions . Ca tegor i e s of d i sease are con-
ceptua l ly a t the h ighes t l eve l o f abs t rac t ion , summar iz ing pa t t e rns of s t a te s and
obse rva t ions . In F ig. 1, we summ ar i ze th i s th ree l evel desc r i p t ion of d i sease
p r oc e s se s . C on s i de r a t i ons o f a ll t h r e e le vel s e n t e r i n t o t he r e c o m m e nda t i on o f
the ra py . Bon ne r [16] dev e lo ped a sing le l eve l mod e l w i th causa l an d a s soc ia t iona l
r e l a t i ons i n t e r m i ng l e d . W he n d i a gnos i s t o to be m ode l e d i n a dom a i n o f know l e dge
w he r e m e c ha n i s m s o f d i s e a s e a r e unde r s t ood , t he c a us e a nd e f f e c t m ode l c a n be
used to s ign i f ican t ly imp rov e the basi s on whic h dec i s ions a re made . Whe n,
however , l e s s in forma t ion i s ava i l ab le , a s soc ia t ions be tween f ind ings mus t be
re l i ed on to a grea te r ex ten t, and the goa l s o f r eac hing s t ruc tured , and we ll -
e xp a i ne d c onc l us i ons a nd r e c om m e nda t i ons m a y no t be f u ll y sa ti sf ie d.
C a u s a l n e t w o r k o f s t a te s
In our m ode l o f d isease , the pa thogen es i s and mec han isms of a d i sease process a re
desc r i bed in t e rms o f cause and e f fec t r e la t ion sh ips be twee n pa th ophy s io l og ic a l
s ta te s . S ta te s a re summ ary desc r ip t ions of event s whic h a re devia t ions f rom
norma l i ty . S t ri c t causa l i ty [17] is no t a s s ume d- - the re may be mul t ip le causes and
effec ts, and in a given pa t ient , a cause ma y be pr esen t wi th ou t an y o f i ts e f fec ts
occu r r ing a t the same t ime . Var ious e ffect s can fo l low f rom a g iven cause , each
p r odu c e d w i t h a d i f f er e n t s t r e ng t h o f c a us a t i on . E xa m pl e s o f s ta te s w ou l d be
-
8/10/2019 0004-3702(78)90015-2
4/28
148 S. M WEISS C. A. KUL1K OWSKI A. SAFIR AND S. AMAREL
D s e a s @
C a t e g o r @ s
CIoss fJcoDon _ nks
/
/ ~
/ / / / \
I
Po/ t
G /oh
. /nitS.
I I
/ /
I
Obs e r vo I / o /
F IG. 1. T h ree - l ev e l d e sc r i p t i o n o f a d i s ea se p ro ces s .
i n c r e a s e d i n t r a o c u l a r p r e s s u r e o r g l a u c o m a to u s v i s u a l f i el d l o s s . M a n y s u c h
s t at e s m a y o c c u r s im u l t a n e o u s l y i n a n y d i s e a s e p r o c e s s . A s t a t e t h u s d e f i n e d m a y
b e v i e w e d a s a s et o f v a lu e s o f a s t a t e v a r i a b l e a s u s e d i n c o n t r o l s y s t e m s t h e o r y .
I t d o e s n o t c o r r e s p o n d t o o n e o f t h e m u tu a l l y e x c lu s iv e s t a t e s t h a t c o u ld b e u s e d
-
8/10/2019 0004-3702(78)90015-2
5/28
M O D E L - B A S E D M E D I C A L C O N S U L T A T I O N
1 4 9
t o d e s c r i b e a p r o b a b i l i s t i c s y s t e m . T h i s d e f i n i t i o n w a s c h o s e n t o c o r r e s p o n d t o t h e
b a s i c e n t it i e s p h y s i c i a n s u s e w h e n t h e y d e s c r i b e d i s e a s e m e c h a n i s m s . A s o m e w h a t
s i m p l i f i e d g r a p h m o d e l o f g l a u c o m a is i l l u s tr a t e d in F ig . 2 w h e r e e a c h n o d e n~ i s
~ I J o i n
I I IMARY A,N~.LE
I
LOSU t |
M ECT~A N~ M ~ ~ C A T A R A C ~
_ _ ~ - ~
~ F k 5
[ S u O O E N E ) tNGLE
CLOSURE
d E X
TREMF..LY
J
DECREASEO i
~ U E O U S I
u '~ L w /
s P , c E @ l |
~ ~ r , 9 ~ J
PRESSURE 1
D ' F F C ~ N T I A L { ~ I
ANGLE
~ ~
ANGLE ~ P~LONGEO
C L O S U R E ~ G L
CLOSURE ~
I R I S
B O M E I F . '
SUDDEN, H IGHLY I
I N C R E A S E D [
I N T R A O CU L A R r ~ , 9
PRESSURE ~)J
t.8,, - - L
A N D O C U L A R I
2 - ' ~ E 1
2 '
1 ~ SYNECHI A
( T U M O R O R
C Y S T ) ( ~
P E R I P H E R A L A N T E R I O R
S Y N E C H J A S : I M P A I R [ O
: F U N C T IO N O F T H [
~TRABECULAR M~, ,~ ' ~ /OAK
I ~
S E G N 4 E N T
I N F L & I ~ W I A T ~O N
2
1 9 J LA
I o ~ 0 s o I
OF IR IS ( :
OF TRABIECLR.AR
I~ $ E H W t ~ K ( ~
D E C R E A S E D
O U T F L O W
F A C I L I T Y
T T ~ M .
T 8
ELIr.VATE D
INTI~AOCULAf l
P R E S S ~
I PROL(~GED
LEV 'T[O
I N T R A O C U L A R
P R S S u R I r
( E ~ )
. . 9
I E L E V A T E D
P R E S S U R E
T R & N S M I T T E D
T O O P T I C
N [ RV E H E A D { ~
I ~ C R E A S E D
B L O O O
S ,,IIq~LY
T O O P T I C
N E R V E : H E A O 0
I r . 8
NN~URAL TISS UE [
L O S S A N D C U P P I N G I
O F N E RV E H E A D ~ 1
'1~.9
l G L A U C O M A T O U S I
V I S U A L F r E L D |
L O S ~ ]
INCd~OWT H
OVE R ". Id.
O 4 m E N ' iNGLE
M I [ C H A N I S N
* E s E c R E * I
S U S C E P T I B L E [
T O ~ M L I
F I G . 2 . P a r t i a l c a u s a l n e t w o r k f o r g l a u c o m a . S t a t e s w i t h n o a n t e c e d e n t c a u s e s a r e i n d i c a t e d b y a . .
T h e c i r c l e d n u m b e r s C c o r r e s p o n d t o t h e s t a t e l a b e l s tt u s e d i n e x a m p I e s i n t h e t e x t .
-
8/10/2019 0004-3702(78)90015-2
6/28
] 5 0 S . M . W E I S S C . A . K U L I K O W S K I A . S A F I R A N D S . A M A R E L
a pa t ho phys i o l og i c a l s t a t e a nd e a c h e dge is a c a us a l c onn e c t i on . D i s e a s e p r oc e s s e s
m a y b e c h a r a c t e r i z e d b y p a t h w a y s t h r o u g h t h e n e t w o r k . A c o m p l e t e p a t h w a y
f r om a s t a r t i ng t o a t e r mi na l no de u s ua l l y r e p r e s e n t s a c ompl e t e d i s e a s e p r oc e s s
w hi l e p a r ti a l p a t h w a y s , f r o m s t a r t i n g t o n o n t e r m i n a l n o d e s , r e p r e s e n t v a r i o u s
de g r ee s o f e vo l u t i on w i t h i n t he d i s e a s e p r oc e s s. P r o g r e s s i on a l ong a c a us a l pa t hw a y
is u s ua ll y a s s oc i a t e d w i t h i nc r e a s i ng s e r i ous ne s s o f t he d i s e a s e . Fo r e xa m pl e , i n
F ig . 2 a c omp l e t e pa t hw a y i s t r a ve r s e d f r om n35 ( p r i ma r y ope n a ng l e me c h a n i s m )
t
r/a1
( g l a uc oma t ous v i s ua l f i el d l o s s ) ; ( n ss nz s
n26 n27 n28 n29
nao
n31 .
A
pa r ti a l pa t h w a y is t r a ve r s e d f r om n3s ( p r i ma r y ope n a ng l e me c ha n i s m ) t o n26
( e l e va t e d i n t r a oc u l a r p r e s s u r e ) ; ( n3s n2s n26 ).
W he n a s et o f c a us e a nd e f fe c t r e l a t i ons h i ps be t w e e n s t a t e s is spe c i fi e d , t he
r e s u lt i ng s t r uc t u r e is a ne t w or k , o r d i r e c t e d a c yc l i c g r a ph o f s ta t es . T he s t a te
ne t wor k is de f ine d by (S , F , X , N) whe re S i s the se t o f s t a r t i ng s ta te s , those s ta te s
wi th no an t ece den t causes ; F i s the se t o f f ina l s t a te s, th ose s ta te s wi th n o e f fec t s ;
N is the to ta l s e t o f st a te s ; X is the se t o f map pin gs be t wee n s ta te s in d ic a t i ng causa l
re la t ionsh ips .
al j
The map pin gs a re of th e for m n i --* n i wh e re a o i s the s t r eng th of ca usa t io n
( i n t e r p r e t e d i n t e r ms o f f r e qu e nc y o f oc c u r r e n c e ) a n d n i a nd n j a r e s t a t es . T h i s r u l e
i s i n t e r p r e t e d a s : s t a te n i c a us e s s t a t e n j , i nde p e nde n t o f o t he r e ve n t s , w i t h f r e-
que nc y a u. S t a r t i ng s t a t es a r e a l s o a s s i gne d a f r e que nc y me a s u r e i nd i c a t i ng a
p r i o r o r s t a r ti n g f r e q u e n c y . T h e s t r e n g t h s o f c a u s a t i o n a r e r e p r e s e n t e d b y n u m e r i c a l
va lue s, f r a c t i ons be t w e e n z e r o a nd one , w h i c h c o r r e s p ond t o qua l i t a t i ve r a nge s
such a s : some t imes , o f ten , usua l ly , o r a lways .
S ta te s a re s u m m a r y s t a t e m e n t s . M a n y e v e n t s a n d m a n y c o m p l e x r e l a t i o n s h i p s
may be sum mar ize d by a s ing le s t a te. For exa mpl e , in F ig . 2 , ' neura l t i s sue tos s
a n d c u p p i n g o f t h e n e r v e h e a d is a s u m m a r y o f a m u c h m o r e c o m p l e x s i t u a t i o n .
I f a h i ghe r r e s o l u t i on de s c r i p t i on i s de s i r e d , s e ve r a l d i f f e r e n t t ype s o f ne r ve l o s s
a nd c upp i n g c ou l d be s pe ci f i ed . T he r e s o l u t i on o f s t a t es s hou l d be ma i n t a i ne d a t a
l evel c ons i s t e n t w i t h t he ob j e c t i ve o f e f fi c i e nt de c i s i on - ma k i n g . A s t a t e ne t w o r k
c a n be t hou gh t o f a s a s t r e a ml i ne d mo de l o f d i s e a s e t ha t un i f ie s se ve r a l i mpor t a n t
c onc e p t s a nd gu i de s u s i n ou r goa l o f d i a gnos i s . I t is no t me a n t t o be a c ompl e t e
mode l o f d i sease .
u l e s f o r a s s o c i a t i n g s t a t e s t o o b s e r v a t i o n s
O bs e r va t i ons ( te st s) t he h i s t o r y , s i gns , s ympt oms , a nd l a bo r a t o r y t e s t s - - a r e t h e
f o r m i n w h i c h i n f o r ma t i on a bou t a pa t i e n t i s p r e s e n t e d . T h e s e c l i n i c a l f e a t u r e s ,
h o w e v e r , m u s t b e u n i f i e d i n t o s o m e c o h e r e n t f r a m e w o r k f o r e x p l a n a t i o n a n d
d i a gnos i s . O bs e r va t i ons a bou t a pa t i e n t a r e u s e d t o c on f i r m o r de ny c e r t a i n s t a t e s
i n t he ne t w o r k w h i c h de s c r i b e t he d i s e a s e p roc e s s . A s i ng l e s t a t e ma y be a s s oc i a t e d
w i t h ma ny o bs e r va t i ons . T he s e s t a t es c a n t he n be r e l a t e d by c a us al p a t hw a ys t ha t
e xp l a in t he me c ha n i s ms o f d i s e a s e i n a pa t i e n t .
T he r e l a t i ons h i p be t w e e n t e st s a nd s t a te s is non - c a us a l ; i t is a s s oc i a t i o na l . Fo r
-
8/10/2019 0004-3702(78)90015-2
7/28
M O D E L - B A S E D M E D I C A L C O N S U L T A T I O N 1 5 1
a g i v e n obs e r va t i on , c on f i de nc e me a s u r e s a r e u s e d t o i nd i c a t e a de g r e e o f be l i e f
in the pre sence of spec i f i c s t a te s .
T he r u l e s f or a s s oc i a t i ng t es t s w i t h s t a t e s a r e r e p r e s e n t e d a s :
Q U
t i --+ nj
w he r e t~ is t he i t h obs e r va t i on o r B oo l e a n c omb i na t i on o f obs e r va t i ons ) , nj i s t he
. i th no de , an d Qo - 1 ~< Qo
-
8/10/2019 0004-3702(78)90015-2
8/28
1 5 2 s . M . W E I S S C . A . K U L I K O W S K I A . S A F I R A N D S . A M A R E L
o f 0 .5 . A h i g h a p p l a n a t i o n t e n s i o n r e a d i n g , w h i c h i s u s u a l l y m o r e r e l i a b le , m a y
b e a s s ig n e d a h i g h e r c o n f i d e n c e , s u c h a s 0 .7 . I f , b y o p h t h a l m o s c o p y i t i s f u r t h e r
d e m o n s t r a t e d t h a t t h e a p p e a r a n c e o f t h e o p t i c d is c i n d i c a t e s d a m a g e t o t h e o p t ic
n e r v e ( w i t h a c o n f i d e n c e o f 0 . 3 ) , t h e s e r e s u l t s m a y b e c o m b i n e d a n d a s s i g n e d a
c o n f i d e n c e o f 0 . 8 t h a t t h e p r e s s u r e h a s b e e n a n d i s t r u l y e l e v a t e d . F i g . 3 i l l u s t r a t e s
t h es e r e l a t io n s h i p s , w i t h c i r c u l a r n o d e s s t a n d i n g f o r s t a te s , a n d s q u a r e n o d e s f o r
o b s e r v a t i o n s . T h e n u m b e r o n t h e l in k t h a t c o n n e c t s t e s t s t o s t a t e s i s t h e c o n f i d e n c e
w i t h w h i c h a t e s t s u p p o r t s a s ta t e .
R u l e s f o r a s m ~ e i a t i ~ d i s e a s e e n t e g o r i e s t o s t a t e s
D i a g n o s t i c a n d p r o g n o s t i c c a t e g o r i e s o f d i s e a s e a r e d e f i n e d i n t e r m s o f o r d e r e d
p a t t e r n s o f r u l e s , w h i c h w e r e f e r to a s c l a s s i f ic a t i o n t a b l e s .
T h e t a b l e s c o n t a i n r u le s o f t h e f o r m :
n l A ~ n 2 ~ O 1
n l A n 2 A - l n a --* D 2
n i A n 2 A . . . - 1 1 7
i ~ D i 1
n l A
n 2 A . . , n~ --* D ~
w h e r e D , i s t h e i th d i a g n o s t i c a n d p r o g n o s t i c c a t e g o r y , w h i c h i s i m p l i e d b y t h e
g i v en o r d e r e d p a t t e r n o f s t a te s , n t A n 2 . . . A n~ . F o r th i s c h o s e n f o r m a n d
o r d e r i n g o f r u le s , t h e ta b l e s c a n b e r e f e r r e d t o b y u s in g a n a b b r e v i a t e d n o t a t i o n o f
o r d e r e d p a i r s :
( r tl , D I ) (1 12 , D 2 ) , . . . ( n f _ l , D i - 1 ) , ( n i , D i ) .
T h e c l a s s i f i c a t io n t a b l e s c a n b e a u g m e n t e d t o in c l u d e t h e r a p y r e c o m m e n d a t i o n s .
T h e s e t a b l e s a r e o r d e r e d t r i p l e s o f t h e f o r m :
( h i , D 1 , T I ) , ( n 2 , D 2 , T 2 ) , . . ( n f _ l , D i _ t , T i _ l ) , ( h i , O , , T , )
w h e r e T~ a r e t r e a t m e n t s ( o r t r e a t m e n t p l a n s ) f o r p a t i e n t s f a l l i n g i n to p a r t i c u l a r
d i a g n o s t i c c a t e g o r i e s .
I n t h e f o l l o w i n g s e c t i o n s , c l i n ic a l d e c i s i o n - m a k i n g w i ll b c c o n s i d e r e d a s a p r o b -
l e m o f u s i n g a C A S N E T m o d e l f o r :
( a) s e le c ti n g a n d i n t e r p r e t i n g o b s e r v a t i o n s ,
( b ) a n a l y z i n g a n d r e s o l v i n g co n f l i c ts a n d c o n t r a d i c t i o n s i n t h e o b s e r v a t i o n s ,
( c) s e le c t i n g d i a g n o s t i c a n d p r o g n o s t i c c a t e g o r i e s ,
( d ) r e c o m m e n d i n g t r e a tm e n t s .
T e s t R e s u l t n t e r p r e t a t io n
A t e s t r e s u l t h a s t h e f o l l o w i n g f o r m : o b s e r v a t i o n t~ i s t r u e , f a l s e , o r u n c e r t a i n .
B a s e d o n a g i v e n r e s u l t t~, a m e a s u r e o f c o n f i d e n c e , Q ~j, m a y b e a s s i g n e d t o s t a t e
n~. M o r e t h a n o n e t e s t m a y c o n f i r m o r d e n y a s in g l e s t a t e w i t h v a r y i n g d e g r e e s o f
-
8/10/2019 0004-3702(78)90015-2
9/28
-
8/10/2019 0004-3702(78)90015-2
10/28
54 s .M . W E IS S C . A . KUL IKOW S KI A . S AF IR AND S. AM AR E L
i s gu ided by the resu lt s o f p rev iously asked ques t i ons c an usual ly reduce the
n u mb er o f q u es t io ns th a t mu s t b e asked , o f t en e l imin a t in g i r r e l ev an t an d r ed u n d a n t
ques t ions . Ask ing the r igh t ques t ions an d in an in te l l igen t o rder i s an i mpor ta n t
aspect of the d iagno st ic process.
The s t rategies fo r tes t se lec t ion that have been developed fo r C ASNE T type
models can be ca tegor ized as those emphasiz ing :
a) local log ical const r a in ts am ong ques t ions ,
b) ca tegor ies o f causal pathways , and
c) l ike l ihood measu res over the states.
These s t ra teg ies are no t mutual ly exclus ive , and a l l th ree may be combined in to
a single overal l s trategy. The simplest s trategy, yet perhaps the most effect ive for
a wel l -c i rcumscribed dom ain o f app l ica t ion , i s the s t ra tegy that emphasizes local
log ical const ra in ts am ong the ques t ions . For th is s t ra tegy , ques t ions on re la ted
topics are organiz ed in to small local t ree structures. Eac h gro up of quest i ons is
asked only whe n a f ixed set of logical condit ion s is sat isf ied. The second strategy
depends on i so la t ing the causal pa thwa ys whic h po ten t ia l l y exp la in the observa-
t ions which have a l ready been recorded . The s t ra tegy would then pursue those
observat ions which are re la ted to the s ta tes foun d on these pathways . The iden t i fi -
ca t ion o f pathwa ys which ma y exp la in the curren t observat ions and re la ted p ro-
cesses of disease is discussed in the next section.
A l ikel ihood s t ra tegy fo r the CASN ET model i s based on the ass i gnment o f
weigh ts to each o f the nodes in the s ta te networ k . Tes ts whic h can p rod uce resu lt s
hav ing g reater measures o f conf idence tha n curren t ly hel d fo r the s ta tes , a re con-
s idered poss ib le cand idate s fo r fu r ther tes ting. Of these tests, the one th at re la tes
to the highest weighted node is selected.
A numb er o f character i s t ics o f the s ta te netwo rk are i mpor ta n t fo r the speci fi -
cat ions of inference strategies:
a) No loops may exis t in the ne twor k because a ll t rans i t ions between node
are u n idi rec t io n a l u n d er th e assu m p t io n o f cau sal p ro d u c t io n .
13) Start i ng node s have no ante cede nt causes or predecess ors in the net wor k)
and represen t even ts taken as the s tar t ing even ts in the causal chains . These nodes
are assigned prior) weights, ai , based on the ir relat ive frequenc y of occurrence .
c) Each t ra ns i t i on weigh t has a maxi mum value o f un i ty . The s um of t rans i t ion
weights leaving nod e n~ is not necessari ly unity , becaus e the successors of n i are
no t necessar i ly mutual ly exclus ive . In add i t ion , the model incorporates on ly
consequences o f even ts tha t are o f in teres t to the p rocess being descr ibed , leav ing
unspeci f ied any o ther poss ib le ou tcomes.
i j
d) The transition weight a~j, in a link n~ -~ hi, is assigned on the assumption
that n i is caused by ni with freque ncy ai~, inde pen den tl y of the way in whic h n~
was en tered f rom o ther nodes o f the network . For a given model , a consis ten t
in terp re t a t ion must be given to the t rans i t ion weigh ts th r ough out the network .
-
8/10/2019 0004-3702(78)90015-2
11/28
-
8/10/2019 0004-3702(78)90015-2
12/28
156 s . M . W E I S S ~ C . A . K U L I K O W S K I A . S A F I R A N D S . A M A R E L
Ex mpl e 2 (see Fig . 2) . Ass ume tha t n is confirmed, n17 and n~o are denied, and
the o ther nodes are o f und ete rmi ned s ta tus . The fo rwa rd weigh t o f nx9 =
(0.5)(0.8) + (0.01)(0.20) ----- 0.402. The weight is ca lc ula te d fr om pa th wa ys be gi nni ng
at n~ ( the neares t conf i rmed node) and a t n14 ( the on ly unden ied s tar t ing node
tha t leads to n~ 9)-
The ra t ionale fo r choosing the p roduct o f success ive t rans i t ion weigh ts in a
pathwa y lies in the assum pt ion that each t r ans i t ion weigh t a o is indepe nden t o f a l l
prece ding tra nsi tio ns. If nj_l is a c onf ir me d node , t he w eig ht wF(J --1 I i) -~ 1, so
a ll p revio u s t r an s i t io n we ig ht s w i th i n th e p a t h way n eed n o t b e co mp u ted . Hen ce
the weigh t o f the end node o f a pathw ay is ca lcu la ted on ly f ro m the neares t con-
firmed or star t ing node.
When there i s on ly one admiss ib le path way lead ing to a node, ignor i ng the
possib i l ity o f over lapp i ng causal even ts in t roduces no er ro r in the com put at ion
of weights . I f the n etw ork i s def ined wi th mu tual exclus iv i ty between a ll successors
of any g iven node, the p r ob l em can be ope ra t i onal l y t rea te d as a Marko v chain
calcu la tion . Tha t is , a l l conf i rme d nodes do in rea l ity lie on a un ique pathwa y . In
our represen ta t ion , when over lap between pathways does occur , some nodes in
the networ k may be g iven a g reater weigh t by the above co mput at i ons tha n would
resu l t f rom exact f reque ncy ass ignments over d is jo in t ly def ined pathways . Yet ,
because a pa th wa y to a node, n j , represents the ma nn er in which nj is prod uce d,
th is g reater weigh t ing i s accep tab le and even help fu l . The tendency towards over-
weigh t ing i s re la ted to the numb er o f pat hway s tha t lead to n j and the s t reng t h o f
t rans i t ions between the nodes in these pathways . But i t i s p recise ly those nodes
which have many poss ib le ways o f occurr ing and whi ch have s t r ong causal and
frequency conne ct i ons that are the mos t l ike ly to occur fo r the pat ien t . S ince a
p ro d u c t o f f r ac t ion s n o t g rea te r th an u n i ty is emp lo y ed fo r th e co mp u ta t i o n o f
weigh ts , weights co mput ed on the basis o f few observ at ions wi l l resul t in re la t ively
smal l weigh t ass ignments to the nodes . For some nodes , th is weigh t ass ignment
may be an accurate measur e o f f requency . Even wi tho u t exact f requenc y ass ign-
ments , the manner in which conf i rmat ion o r den ia l o f nodes i s inc luded in the
weight cal cul at i on can be quite effect ive in guidi ng the select ion of tests. Th e
conf i rma t ion o f a g iven node, n i , wi ll usual ly g reat ly increase the weigh t o f a l l o f
i ts ef fects. Ma ny fewer f rac t i onal mul t ip l i ca t ions wi l l be used in com put ing the
necessary path way s fr om nl, s ince for the successors o f nl the weight of n~ can be
assum ed to be unity . Simila rly , the weights resu lt ing fr om a denie d node, n~, will
decrease, s ince n i ca nno t l ie on a ny admi ssible p at hwa y to an ot he r node . This is
p recisely the response needed to gu ide the search fo r i n fo rma t io n to top ics sug-
ges ted by the accumulated ev idence.
In genera l , the overa l l e ffect o f fo rward weigh t ca lc u la t ion is to increase the
weigh ts o f those nodes resu l t ing f rom conf i rmed nodes whi le decreas ing those f r om
denied nodes.
-
8/10/2019 0004-3702(78)90015-2
13/28
M O D E L B A SE D M E D I C A L C O N SU L T A T I O N
157
n v e r s e w e i g h t s f o r t e s t s e l e c t i o n
T he f o r w a r d w e i gh t o f a node , h i , s umma r i z e s t h e w e i gh t o f e v i de nc e c a r r i e d f r om
t h e
c a u s e s o f n v
T he w e i gh t o f n~ c a n a l s o t a ke i n t o
a c c o u n t t h e
c o n f i r m e d n o d e s
t ha t a r e e f f ec t s o f nv Fo r t h i s w e mus t de f i ne s ome i nve r s e w e i gh t o f c on f i r me d
e f fect ny on t he cau se n v In ana log y to Bayes for mul a for in ve r se probabi l i ty , an
i nve r s e w e i gh t c a n be de f i ne d a s :
w , ( i l j ) = [ w F ( j I i ) w ~ . ( i )] / W F ( j) ,
w he r e c on f i r me d nod e s a r e i gno r e d i n t he pa t hw a ys ( a nd f o r w a r d w e i gh t s a re ,
t h e r e f o r e , c o m p u t e d f r o m s t a r t i n g st at es ). B e c a u s e a n a d m i s s i b l e p a t h w a y c a n n o t
c o n t a i n
a de n i e d n ode , a n i nve r s e w e i gh t is p r opo r t i on a l t o t he w e i gh t o f pa t hw a y s
p a s s i n g
t h r ou gh n j t ha t a l s o pa s s t h r ou gh n~ d i v i d e d by t he w e i gh t o f a ll c u r r en t l y
pos s i b l e pa t h w a ys t o
n j .
E xa mp l e 3 ( se e F ig . 2). A s s ume t ha t a l l pa t h w a ys a r e de n i e d , e xc e p t t hos e
be g i n n i ng w i t h ha s a nd n3~. L e t na t be c on f i r me d a nd t he r e ma i n i ng n ode s un -
de t e r mi ne d . T he i nve r s e w e i gh t s fo r na s a nd na t a r e c a l c u l a t e d a s f o l l ow s :
wF(35 ) = 0.30, wF(37) 0. 01 (s ta r t in g wei ght s)
wF(31) = (0.3)(0.9 )(0.8)(0 .5)(0.9)( 0.9)(0.8 )(0.9) + (0.01)(0.5)(0 .8)(0.9)
= 0.067
wF(31135) = ( 0.9)(0.8 )(0.5)(0 ,9)(0.9 )(0.8)(0 .9) = 0.210
we(31137) = (0.5)(0.8)( 0.9) = 0.36 0
wr(35131) -- wp(35) wr(3I I35) /wF(3 1)
= (0.30)(0.210) / 0.067 = 0.940
w1(37131) = wF(37) - wF(3 1137 )/we (31)
= (0.01)(0.36) /0.067 = 0.054
Since seve ra l e f fec t s may fo l l ow f rom a s in g le cause i t is des i ra b le to cho ose
s ome f unc t i on o f al l t he i nve r s e w e i gh t s t o r e p r e s e n t t he ove r a l l i nve r s e w e i gh t o f
a node n t. A rea son abl e ch oice i s the ma xim um of the inve r se we ight s for each n~:
w , ( i ) - ~
m a x
{ w ~ ( i [ j ) } .
T hi s f unc t i on w a s s e l e c t e d be c a u s e w e ar e s e a r c h i ng t he ne t w or k f o r s t r ong e v i de nc e
t ha t n~ is p r e s e n t . Fo r t he i mpo r t a n t s i t ua t i ons w he r e n ode s l i e on a s i ngl e pa t hw a y
t o a c on f i r me d node , o r node ~ l ie on e ve r y pa t hw a y t o a c on f i r me d no de , t he i nve r s e
w e i gh t f o r t hos e node s w i ll be c o r r e c t l y a s s i gne d a s un i t y . W he r e t he r e a r e t w o o r
mor e mu t ua l l y e xc l us i ve pa t hw a ys t o a c on f i r me d node , t he i nve rs e w e i gh t r e ma i ns
a re l a t i ve l y a c c u r a t e f r e qu e nc y me a s u r e . H ow e ve r , t he pa t hw a ys t o a c on f i r me d
node ne e d no t be mu t ua l l y e xc l us i ve . T he r e f o r e , t he ma x i mu m o f t he inve r s e
w e i gh t s is u s e d a s a n ove r a l l me a s u r e o f i nve r s e w e i gh t. T he i nve r s e w e i gh t f o r a
c on f i r me d node a s si gns w e i gh t on a f r a c t i ona l ba s i s t o i ts po t e n t i a l c a use s , e ve n
w h e n m o r e t h a n o n e c a u s e is s t r o n g l y i n d i c a t e d. T h e m a x i m u m w e i g h t c o m p e n -
s a te s f o r t he l a c k o f mu t ua l e xc l us i vi t y by c ons i d e r i ng e v i de nc e o t he r t ha n a s i ngl e
11
-
8/10/2019 0004-3702(78)90015-2
14/28
158 S .M . WEISS C. A. KULI KOWS KI A. SAHR AND S. AMAREL
c o n f i r m e d n o d e . S i n c e s e v e ra l c o n f i r m e d n o d e s m a y i n f a c t b e u n r e l a t e d , a n
a v e r a g e o r s u m w o u l d a p p e a r t o b e l es s e f fe c t i ve . T h e m a x i m u m i s e f f e c t iv e
b e c a u s e i t p r e s e r v e s t h e w e i g h t o f a s t r o n g p i e c e o f c o n f i r m a t o r y e v i d e n c e w i t h o u t
d i l u t i o n f r o m o t h e r n o n - e x c l u s i v e c a u s e s ; i t r e c o g n i z e s t h e p o s s i b l e m u l t i p l i c i t y
o f c o n f i r m e d e f fe c ts f r o m a s in g l e c a u s e ; a n d i t g e n e r a l l y p r o v i d e s a r e a s o n a b l e
b a s i s o f c o m p a r i s o n w i t h t h e f o r w a r d w e i g h ts .
T h e c a l c u l a t i o n o f i n v e r s e w e i g h t s i s s t r o n g l y i n f l u e n c e d b y e v i d e n c e f o r t h e
c o n f i r m a t i o n o r d e n i a l o f n o d e s . T h e w e i g h t o f a n o d e m a y b e i n c r e a s e d w h e n i ts
e f f ec ts a r e c o n f i r m e d . I n i t i a l l y , a p a t h w a y m a y b e a n u n l i k e l y a l t e r n a t i v e , b u t a f t e r
s o m e t e s t i n g it m a y b e c o m e t h e o n l y fe a s i b l e p a t h w a y t o a p a r t i c u l a r c o n f i r m e d
n o d e . T h i s re s u l t s i n i n c r e a s e d w e i g h t a s s i g n m e n t s t o t h e r e m a i n i n g c a u s e s o f t h e
c o n f i r m e d n o d e .
v e r a l l w e i g h t f o r t e s t s e l e c t i o n
I n o r d e r t o c h o o s e a n o d e f o r t e s t i n g , a s i n g l e f u n c t i o n o f t h e f o r w a r d a n d i n v e r s e
w e i g h ts o f t h e n o d e i s n e e d e d a s a n o v e r a l l m e a s u r e . T h e m a x i m u m o f t h e s e t w o
w e i g h t s h a s b e e n c h o s e n :
m s = m a x
{ w r i ) ,
w , i ) } .
T h i s c h o i c e r e f l e c t s t h e n e e d t o h a v e a m e a s u r e o f s t r o n g c o n f i r m a t o r y e v i d e n c e
f o r t h e p o t e n t i a l p r e s e n c e o f a n o d e h i . E v i d e n c e o f th e d e n i a l o f n i s i n c l u d e d i n
b o t h w F 0 ) a n d w l / ) . T h e s e f o r w a r d a n d i n v e r s e w e i g h t s r e p r e s e n t t h e c o n t r i b u t i o n
f r o m d i f fe r e n t p a r t s o f t h e n e t w o r k t o w a r d s t h e l i k e l i h o o d o f c o n f i r m a t i o n o f n i.
T h e m a x i m u m i s t h u s a m e a s u r e o f s t r o n g c o n f i r m a t o r y e v i d e n c e t o w a r d s n ,
t h r o u g h o u t t h e n e t w o r k . I t s h o u l d u s u a l l y p r o v i d e g o o d t e s t i n g c a n d i d a t e s .
R e l a t i v e l y e f f ic i e n t a l g o r i t h m s c a n b e s p e c if i e d f o r t h e c o m p u t a t i o n o f w e i g h t s [2 ].
T h e s e a l g o r i t h m s t a k e a d v a n t a g e o f t h e a c y c l i c n a t u r e o f t h e s t a t e n e t w o r k s o t h a t
t h e s t a t e s m a y b e t o p o l o g i c a l l y s o r t e d .
T e s t s e le c t io n w i t h c o s t a s s i g n m e n t
A w e i g h t i s a m e a s u r e o f l i k e l i h o o d , b a s e d o n t h e e v i d e n c e g a t h e r e d f o r t h e p o s s i b l e
c a us e s o f a n o d e . T h e w e i g h t d o e s n o t t a k e i n t o a c c o u n t t h e c o s t o f p e r f o r m i n g a
t e s t t h a t m a y c o n f i r m o r d e n y t h e n o d e . L e t t~ b e a te s t f o r n o d e n j , a n d C i b e t h e
c o s t o f t v W j i s t h e c u r r e n t l y a s s i g n e d n o n - z e r o ) w e i g h t o f n o d e n j.
T w o c o s t s t r a t e g ie s t h a t h a v e b e e n u s e d f o r te s t s e I e c t io n a r e :
a ) M a x i m u m w e i g h t - t o - c o s t r a t i o :
s e l e c t t i , s u c h t h a t W f l C i = m a x W , , , / C , , ) .
Wt n
b ) M a x i m u m w e i g h t w i t h i n a c e r t a i n r a n g e o f c o s t s :
s e l e c t t~ , s u c h t h a t W ~ ----- m a x 1 4r= ) f o r a l l t n w i t h C n < C .
m
A s t r a t e g y o f m a x i m u m w e i g h t s e I e c t i o n is a s p e c i a l c a s e o f a ) w h e n t h e c o s t s a r e
e q u a l o r a r e i g n o r e d . A m i n i m u m c o s t s t r a t e g y is a s p e c i a l c a s e o f b ) w h e r e C n i s
-
8/10/2019 0004-3702(78)90015-2
15/28
M O D E L B A S E D M E D I C A L C O N S U L T A T I O N 1 5 9
t a k e n a s t h e m i n i m u m c o s t ( f o r t h e r e m a i n i n g t e s t s ) a n d W j i s a n y n o n - z e r o
w e i g h t .
T h e s t o p p i n g r u l e f o r t h e l i k e l i h o o d s t r a t e g y c o n s i s t s o f t e r m i n a t i n g t e s t s e l e c t i o n
w h e n n o w e i g h t e x c e e d s a f i x e d t h r e s h o l d . F o r a n i n - d e p t h c o n s u l t a t i o n i n a n
a p p l i c a t i o n s u c h a s g l a u c o m a , w h e r e a l l t o p i c s m u s t b e c o v e r e d t h o r o u g h l y , a l l
q u e s t i o n s a r e a s k e d t h a t h a v e n o t b e e n l o g i c a l l y e x c l u d e d b y p r i o r r e s p o n s e s .
T h i s c o r r e s p o n d s t o s e t t i n g t h e t h r e s h o l d t o z e r o .
A f o r m o f h y p o t h e s i s - d r i v e n t e s t s e l e c t i o n h a s a l s o b e e n f o r m u l a t e d . A h y p o -
t h e s i s c o r r e s p o n d s t o a c l a s s o f l i k e ly c a u s a l p a t h w a y s t h a t e x p l a i n t h e p a t i e n t s
o b s e r v a t i o n s a n d r e l a t e d b u t a s y e t u n k n o w n f i n d i n g s . T h e s t r a t e g y t h e n s e l e c t s
t e s t s t h a t s u p p o r t t h e s e h y p o t h e s e s . T h e f o l l o w i n g s e c t i o n d e s c r i b e s m e t h o d s o f
i d e n t i f y i n g t h e l i k e l y h y p o t h e s e s ( c la s s e s o f p a t h w a y s ) f o r a p a t i e n t .
I n t e rp r e t a ti o n o f i s e a s e P r o c e s s e s w i th i n t h e S t a t e N e t w o r k
T h e s t a t e n e t w o r k i s a g e n e r a l s t r u c t u r e t h a t i m p l i c i t l y c o n t a i n s l a r g e n u m b e r s o f
b o t h c o m p l e t e a n d p a r t i a l c a u s a l s t a t e p a t h w a y s , r e p r e s e n t i n g p r o c e s s e s o f d i s e a s e .
S e v e r a l g e n e r a l c l a s s e s o f p a t h w a y s c a n b e d e s c r i b e d w h i c h a r e u s e f u l f o r d e c i s i o n -
m a k i n g a n d e x p l a n a t i o n . T h e s e c l a s s e s o f p a t h w a y s a r e c h a r a c t e r i z e d b y :
( a ) t h e i r s t a r t i n g n o d e s a n d ,
( b) t h e i r t e r m i n a l n o d e s .
S t a r t i n g n o d e s o r s t a t e s a r e t h o s e s t a t e s i n t h e n e t w o r k f o r w h i c h n o c a u s e s
h a v e b e e n d e f in e d . T h e s t a r t i n g n o d e s a r e e x p l i c i t l y d e t e r m i n e d b y t h e s t r u c t u r e
o f t h e s t a te n e t w o r k ; t h e c o m p l e t e s e t o f p o s s i b l e s t a r t i n g s t a t e s is i n d e p e n -
d e n t o f a n y c o n f i g u r a t i o n o f c o n f i r m e d s t a te s . I n F i g . 2 , h i 4 i s a s t a r t i n g s t a t e
a n d n 1 9 is n o t ; h i 9 w i l l n e v e r b e a s t a r t i n g s t a te , e v e n w h e n a l l o f i ts c a u s e s
n18, n17, n14, n 2 0) a r e d e n i e d . W i t h i n t h e m o d e l , a s t a r t i n g s t a t e i s t h e m o s t
a n t e c e d e n t c a u s e o f f u r t h e r p r o g r e s s i o n o f d i s e a s e in a p a t i e n t . I t r e p r e s e n t s a
b a s i c c a u s a l m e c h a n i s m w h i c h c h a r a c t e r i z e s a d i s e a s e p r o c e s s . A n y c a u s a l p a t h w a y
t h a t e x p l a i n s t h e d i s e a s e p r o c e s s i n v o l v e d i n a p a r t i c u l a r p a t i e n t c a n b e c h a r a c -
t e r i z e d b y i ts s t a r t i n g s t a te . W h e n a n o n s t a r t i n g s t a t e h a s a l l o f i t s a n t e c e d e n t
c a u s e s d e n i e d , t h i s s t a t e w i l l n o t a p p e a r o n a n y p a t h w a y t h a t a t t e m p t s t o e x p l a i n
t h e m a n i f e s t a t i o n o f d i s e a s e i n a p a t i e n t . T h e n o n s t a r t i n g s t a t e s r e p r e s e n t e v e n t s
w h i c h s h o u l d b e e x p l a i n e d b y t h e e v e n t s w h i c h c a u s e t h e m .
T h e c l i n i c i a n is u s u a l l y m o s t c o n c e r n e d w i t h t h e m o s t l i k e l y c a u s e s o f d i s e a s e
f o u n d i n a p a t i e n t . T h e m o s t l i k e l y s t a r t i n g n o d e i s t a k e n a s t h e n o d e w h i c h
e x p l a i n s t h e g r e a t e s t n u m b e r o f s t a t e s o f d i s e a s e . T h i s is t h e s t a r t i n g s t a t e f r o m
w h i c h p a t h w a y s ( c o n t a i n i n g n o d e n i e d n o d e s ) a r e g e n e r a t e d t h a t t r a v e r s e th e
g r e a t e s t n u m b e r o f c o n f i r m e d n o d e s . I f t w o o r m o r e s u c h s t a r t i n g s t a te s a r e f o u n d ,
a l i k e l i h o o d m e a s u r e i s c o m p u t e d f o r t h e s ta t e s , a n d t h e s t a r t i n g n o d e w i t h t h e
g r e a t e s t w e i g h t i s s e l e c t e d . I f a s i n g l e s t a r t i n g s t a t e d o e s n o t e x p l a i n a l l t h e c o n -
f i r m e d n o d e s , t h e n a n o t h e r s t a r t i n g s t a t e i s f o u n d t h a t e x p l a i n s t h e g r e a t e s t r m m b e r
-
8/10/2019 0004-3702(78)90015-2
16/28
1 6 0 S. M . W E I S S C . A . K U L I K O W S K I A . S A F I R A N D S. A M A R E L
o f r e m a i n i n g c o n f i r m e d s t a t e s . T h e p r o c e d u r e i s c o n t i n u e d u n t i l a l l t h e c o n f i r m e d
n o d e s a r e e x p l a i n e d , a n d t h e c o m p l e t e s e t o f m o s t l i k e ly s t a r t i n g n o d e s i s i d e n t i f i e d .
T h e p a t h w a y s g e n e r a t e d f r o m t h e s e n o d e s r e p r e s e n t t h e m o s t l ik e ly e x p l a n a t i o n s
o f t h e d i s e a s e p r o c e s s e s m a n i f e s t e d i n t h e p a t i e n t .
T h e p h y s i c i a n m a y a l s o w i s h t o d i s c o v e r a l t e r n a t i v e t h o u g h t e ss l i k el y c a u s e s t h a t
p o t e n t i a l l y e x p l a in t h e d i s e a s e m e c h a n i s m s p r e s e n t i n a p a t i e n t . P o t e n t i a l e x p l a n a -
t io n s o f t h e d i se a s e p r o c e s s e s f o r a p a t i e n t c a n b e f o u n d b y g e n e r a t i n g a l l p a t h w a y s
t h a t r e a c h c o n f i r m e d s t a t e s , w i t h o u t t r a v e r s i n g a n y d e n i e d s t a t e s . I n a d d i t i o n ,
s in c e a s ta t e n e t w o r k i s u s u a l l y d e s i g n e d f o r a r e s t r i c te d d o m a i n o f d i s e a se s , t h e
c l in i c i a n m a y w i s h t o d e t e r m i n e t h o s e c a u s e s o f d i s e a s e t h a t h a v e n o t y e t b e e n
e l i m i n a t e d . T h e s e m a y b e o b s e r v e d b y g e n e r a t i n g a l l u n d e n i e d p a t h w a y s i n t h e
s t a t e n e t w o r k .
O b s e r v a t i o n s o f a p a t i e n t a r e o f t e n g a t h e r e d s e q u e n t i a l ly . H i s t o r y q u e s t i o n s a r e
s k e d
b e f o r e t h e p h y s ic a l e x a m i n a t i o n , w h i c h p r e c e d e s t h e l a b o r a t o r y t e s t s . F o r a
g i v e n c o n f i g u r a t i o n o f t h e s ta t e n e t w o r k , p a t h w a y s m a y b e g e n e r a t e d w h i c h , b y
n e c e ss i t y , a r e b a s e d o n a n i n c o m p l e t e s e t o f o b s e r v a t i o n s . F o r a s p e c i fi c p a t i e n t ,
t h e p h y s i c i a n i s o f t e n i n t e r e s te d i n d e t e r m i n i n g t h o s e d i s e a s e p r o c e s s e s w h i c h
h a v e n o t y e t b e e n r u l e d - o u t a n d m a y b e u n c o v e r e d b y a d d i t i o n a l o b s e r v a t i o n s .
P a t h w a y s w h i c h e x p l a i n d i s e a s e p r o c e s s e s f o r a s p e c i f i c p a t i e n t , a r e u s u a l l y t e r -
m i n a t e d a t a c o n f i r m e d n o d e . T h i s p r o v i d e s t h e d i r e c t e x p l a n a t i o n o f t h e e v e n ts
w h i c h h a v e b e e n o b s e r v e d . B y c o n t i n u i n g c a u s a l p a t h w a y s b e y o n d t h i s u s u a l
t e r m i n a t i o n p o i n t o f a c o n f i r m e d n o d e , a n d e x t e n d i n g t h e m t o i n c l u d e a l l n o d e s
w i t h a n u n d e t e r m i n e d s t a t u s , t h o s e a s p e c t s o f t h e d i s e a s e p r o c e s s w h i c h r e m a i n
p o s s i b l e c a n b e i n d i c a t e d .
M a n y d i s e a s e s a r e i r re v e r s ib l y ) p r o g r e s s i v e . O n c e t h e p a r t i c u l a r p r o c e s s e s a r e
d e t e r m i n e d , t h e p h y s i c i a n i s c o n c e r n e d w i t h i d e n t i f y i n g t h e s t a g e s o f t h e d i s e a s e
t o w h i c h t h e p a t i e n t m a y s u b s e q u e n t l y p r o c e e d . T h e p a r t i c u l a r p a t h w a y s w h i c h
h a v e b e e n g e n e r a t e d t o e x p l a i n o b s e r v a t i o n s f o r a p a t i e n t , m a y b e c o n t i n u e d t o t h e
t e r m i n a l n o d e s o f t h e s t a t e n e t w o r k , e v e n i f t h e y t r a v e r s e c u r r e n t l y d e n i e d n o d e s .
T h e s e p a t h w a y s w il l g i v e a n i n d i c a t i o n o f p o s s i b l e f u t u r e e v e n t s , a n d p r o v i d e t h e
b a s is f o r p r o g n o s t i c a s s e s s m e n t .
E x a m p l e 4 s e e F i g . 2 ) . A s s u m e t h a t n o d e s n x5 a n d n 1 9 a r e c o n f i r m e d a n d t h e
r e m a i n i n g n o d e s a r e u n d e t e r m i n e d ; n 14 w i ll b e s e le c t e d a s t h e m o s t l i k e ly m e c h a n -
i s m , b e c a u s e i t e x p l a i n s b o t h n ~ s a n d n 19 . T h e p a t h w a y s e m a n a t i n g f r o m n ~4 ,
n 1 4 n~ 5 n~ e n 4 n5 n 6 n~ n l s n l 9) a n d n 1 4 n 1 9 ) d i r e c t l y e x p l a i n t h e c u r r e n t o b s e r v a -
t io n s o f t h e p a t i e n t . H o w e v e r , f o r f u t u r e e x a m i n a t i o n s , m o r e o b s e r v a t i o n s m a y b e
r e c o r d e d , a n d o n e w il l p r o b a b l y b e i n te r e s te d i n c o n t i n u i n g t h e p a t h w a y s , t o
c h e c k f o r e l e v a t io n s o f i n t r a o c u l a r p r e s s u r e n ~o o r n2 6) . T h e r e a r e a l s o o t h e r
m e c h a n i s m s , w h i c h a r e le ss l ik e l y , b u t m a y p o t e n t i a l l y e x p l a i n n l 9 e .g . , n 2 0 ).
-
8/10/2019 0004-3702(78)90015-2
17/28
M O D E L B A S E D M E D I C A L C O N S U L T A T I O N
161
o n f l i c t s a n d o n t r a d i c t i o n s
T h e d i a g n o s t i c i a n i s s o m e t im e s f a c e d w i th t h e t a s k o f i n t e r p r e t i n g te s t r e s u l ts t h a t
a r e s e e m in g l y c o n f l i c t i n g a n d i n s o m e c a s e s c o n t r a d i c to r y . I t i s p o s s ib l e t o r e c o g n i z e
a n d r e s o lv e m a n y c o n f l i c t s a n d c o n t r a d i c t i o n s b e c a u s e t h e t es t r e s u l t s f o r a p a t i e n t
a r e i n t e r p r e t e d t h r o u g h a m o d e l o f d i s e a s e t h a t e x p r e s s e s t h e m e a n in g o f t h e s e
o b s e r v a t i o n s . T h e m o d e l m a y b e v i e w e d a s c o n t a in in g a n im p l i c i t s et o f c o n s i s t e n c y
c o n d i t i o n s w h ic h m u s t b e s at i sf i ed f o r e a c h p a t i e n t .
T h e p r o c e d u r e s f o r i n t e r p r e t i n g t e s t r e s u l t s h a v e b e e n d e s i g n e d t o r e s o lv e
exp l ic i t conf l i c t s in the se r e su l t s . A s de s c r i bed ea r l i e r , th e t e s t r e su l t whic h i s he ld
w i th g r e a t e s t c o n f id e n c e i s t a k e n a s t h e a c c e p t e d r e s u l t. I f c o n f l i c t i n g r e su l t s a r e
r e c e iv e d w i th e q u a l c o n f id e n c e , t h e n t h e c o n f l i c t i s n o t e d , a n d t h e s t a t u s o f t h e
s t a t e o f d i s e a s e r e m a in s u n d e t e r m in e d u n t i l a d d i t i o n a l r e s u l ts , w i th g r e a t e r c o n -
f idence , r e so lve the conf l ic t .
A t y p i c a l c o n t r a d i c to r y s i t u a t i o n o c c u r s w h e n a st a t e is c o n f i r m e d , y e t al l o f i ts
p o t e n t i a l c a u s e s i n t h e n e tw o r k a r e d e n i e d . F o r e x a m p le , i n F ig . 2, a c o n t r a d i c t i o n
wou ld re su l t in n19 i s con f i r med , and n la , n17 , n~a, /72o a re a l l den ie d . The r e i s no
a d m is s ib l e p a th w a y t o c o n f i r m e d n o d e n l 9, b e c a u s e a l l o f t h e s e p a th w a y s c o n t a i n
a d e n i e d n o d e . O n e p o t e n t i a l e x p l a n a t i o n f o r t h i s d i f fi c ul t y i s t h a t t h e m o d e l o f
d i s e a s e m a y b e i n c o m p l e t e aaad s o m e c a u s e s o f c o n f i r m e d n o d e n 1 9) a r e m i s s in g
f r o m th e n e tw o r k . F o r e x a m p le , a l t h o u g h i t is n o t i n d i c a t e d i n F ig. 2, n an o c u l a r
t r a u m a ) m a y i n f a c t c a u s e n l 9 p e r i p h e r a l a n t e r i o r s y n e c h i a s ) . T h e m o d e l d e s ig n e r
m a y i n t e n t i o n a l l y n o t s p e ci f y a ll t h e p o t e n t i a l c a u s e s ; i n s t e a d , h e m a y i n d i c a t e
t h a t f o r s o m e n o d e s n o c o n t r a d i c t i o n s h o u l d b e a s s u m e d b e c a u s e t h e m o d e l o f
c a u s e s f or t h e s e n o d e s i s i n c o m p le t e . E i t h e r t h e m o d e l is i n c o m p l e t e o r a c o n t r a -
d i c t i o n h a s b e e n f o u n d .
B a s e d o n a c o n f ig u r a t i o n o f c o n f i r m e d a n d d e n i e d n o d e s i n t h e s t a t e n e tw o r k ,
p a th w a y s o f d i s e a s e a r e g e n e r a t e d t o e x p l a in t h e p r o c e s s e s o f d i se a s e f o u n d i n a
g iv e n p a t i e n t . S o m e o f t h e n o d e s i n t h e s e p a th w a y s m a y h a v e a n u n d e t e r m in e d
s ta tus , w i th ICf n i) l < 0. When the Cf o f a node g ene r a ted in a pa t hway is un-
d e t e r m i n e d b u t i n t h e d i r e c t i o n o f d e n i a l , i. e. , - - 0 < C f n / ) < 0 , t h e n t h e e x p l a n a -
t i o n o f d i s e a s e is i n c o n s i s t e n t . T h e e x p l a n a t i o n p lo v i d e d b y t h e m o d e l m a y b e
v a l i d , b u t i t i n d i c a t e s t h a t f u r t h e r , m o r e c o n c lu s iv e e v id e n c e is n e e d e d . I f a n y
in c o n s i s t e n c i e s a r e f o u n d i n t h e s e p a th w a y s , i t is im p o r t a n t t o c h e c k f o r a n y
a l t e r n a t e e x p l a n a t i o n w h ic h , w h i l e n o t t h e m o s t l i k ely , is e n t i r e l y c o n s i s t e n t w i th
th e s t a t e s t h a t a r e e x p l a in e d . T h i s c a n b e a c c o m p l i s h e d b y c h a n g in g t h e t h r e s h o ld 0
to z e r o a n d t h e n f i n d i n g t h e m o s t l i k e ly s t a r t i n g n o d e . N o w a ll n o d e s w h ic h h a v e
b e e n t e s t e d w i th a n y d e g r e e o f c o n f id e n c e w i ll b e a s s u m e d e i t h e r c o n f i r m e d o r
d e n i e d . E i t h e r t h e s a m e m o s t l i k e ly s t a r t i n g n o d e s w i l l b e s e le c t ed , o r a l t e r n a t i v e
m e c h a n i s m s w i ll b e f o u n d .
Exam ple 5 see F ig . 2 ). A ssu me tha t n3s i s the mos t l ike ly s ta r t i ng node an d the
p a t h w a y
r / 35 n 2 5
n26 / /27 F/28
n 2 9 r / 3 o H a l i s
g e n e r a t e d . I f h o w e v e r , t h e s t a t u s o f n es
-
8/10/2019 0004-3702(78)90015-2
18/28
62
s . M . W E I S S , C . A . K U L I K O W S K I , A . S A F I R A N D S. A M A R E L
i s undetermined in the d i rec t ion o f den ia l , then th is ind icates an inconsis tency .
I f a search is made fo r a l ternat ive bu t consis ten t exp la nat ions , and
n s
i s assum ed
denied, then n36 is selected as the mos t l ikely star t ing node , and the co nsist ent
pathway 11136 11126 / /2 7 11128 29 / / 3 0 / / 3 1 i s genera ted .
l a s s if i ca t i o n o f D i s e a s e s
Recogni t ion o f the bas ic mecha n ism s of d isease fo r a p at ien t i s o f ten insuff icien t
fo r d iagnost ic class i fica t ion . An evalua t ion o f the s ta tus o f a pa t ien t must a lso
determine the degree o f p rogres s ion and severi ty o f d isease . Pat i en ts wi th the same
disease may exh ib i t d i f feren t degrees o f dysfunct ion , For example , g l auc oma may
lead to to tal b l indness, but many cases wil l be encountered with l i t t le or no loss of
vision, and these cases must be treated quite d ifferently .
The CASNE T sys tem d i f feren t ia tes between two i mpor tan t ca tegor ies o f c lass i-
ficat ion :
a) the mec han is m of d isease,
b) the severi ty and the degree of progres sion of d isease.
The cause or me cha ni sm o f d isease is described in. terms of the state n et wor k by
the star t ing nodes. For a g iven pat ient , a set of most l ikely star t ing nodes will be
found that iden t i f ies the under ly ing causal mechan ism of d isease . Impl ic i t in the
most l ikely path ways tha t fo l low f rom these s tar t ing nodes is a descr ip t ion o f the
progress ion o f the d isease. S ta teme nts are needed to summar ize s ign if ican t f ind ings
which take in to ac count such fac to rs as the curren t severi ty o f disease and the
prognosis fo r the pat ien t . Addi t i onal l y , specif ic and wel l -es tab l i shed d isease labels
often exist to g ive diagnost ic descript ions. While each name may direct ly cor-
respond to a speci f ic mec han i sm of d isease, f requen t ly severa l me chan i sms of
disease are summarized by a single name.
The classif icat ion tables, n t , DI) , n2,
D 2 . . .
h i ,
Di ,
enable us t p ro d u ce
such descr ip t ions o f the s ta tus o f a pat ien t . These tab les c on ta in o rde red sets o f
d iagnost ic s ta tements in terp re t ing the s ign i f icance o f the var ious f ind ings and
pathways o f disease. W hen the p rocesses o f d isease found in the pat ien t are known ,
as disp layed by the most l ikely pathwa ys genera ted fo r the pat i en t , c lass i fica t ion
tab les wi l l be searched to determine the appropr ia te s ta tements .
Each s tar t ing s ta te has po in t ers to the par t i cu lar c lass i f ica t ion tab les that con ta in
d iagnost ic s ta tements which evaluate th is d isease mechan ism. Severa l s tar t ing
states may refer to the same table, s ince several causal mechanisms may be included
in the same diagnost ic category. For a g iven pat ie nt , the m ost l ikely star t ing states
po in t to the app ropr ia t e tab les . The c lass i f ica t ion tab les con ta i n a ser ies o f rules
ordered by ser iousness o f d isease . The a ppro pr i a te d iagnost ic s ta tement cor-
responds to the single rule that is sat isf ied in the table. This ru le wil l correspond
to the deepest confi rmed state in the pairs n~, Dk) that is rea ched fro m a ny of the
most l ike ly pathways th at refer to this table . In mo st ins tances , an a dd i t iona l
-
8/10/2019 0004-3702(78)90015-2
19/28
M O D E L B A S E D M E D I C A L C O N S U L T A T I O N
6 3
c ons t r a i n t w il l be a dd e d t o t he s e a r c h o f t he c l a s si f i c a ti on t a b l e : w he n a s t a te n t
w i t h i n a t a b l e i s c on f i r m e d , i t m us t be t r a ve r s e d by a pa t hw a y ge ne r a t e d f r om a
m os t l i ke ly s t a r t i ng n ode t ha t r e f e rs t o t h i s t a bl e . O t he r w i s e , t he s t a t e m e n t f o r n~
is i na p p r op r i a t e ; o t he r pa t hw a ys m a y r e f e r t o n, i n a d if f e r e n t t a b l e. T he de e pe s t
s ta te i s app rop r ia te s ince any s ta t eme nt th a t i s fou nd ea r l i e r in the t ab le is fo r a
les s se r ious s tage of d i sease and can be igno red .
E xa m pl e 6. A c l a ss i f i ca t i on t a b l e f o r t he p r i m a r y ope n a ng l e m e c ha n i s m , na s ,
f rom Fig. 2 , i s given as :
(n25,
D1),
(n26, D2), (n3o, D3),
n31, D,,),
whe re :
D 1 = m i l d r i sk o f o p e n a n g l e g l a u c o m a
D 2 = h i gh ri s k o f ope n a ng l e g l a u c om a
D3 = ve ry h igh r i sk of ope n an gle g la uco ma; s ign i f i can t r i sk of v i sua l fi e ld loss
D 4 = ope n a ng l e g l a u c om a .
I f n3s i s s e lec ted as a mos t l ike ly s ta r t in g s ta te, and n25 and n26 a re conf i rme d but
no t n30, th en D2 is ap pro pr ia te . I f n2s , n26, n30, an d 1731 are co nf i rme d, th en D4
i s a pp r op r i a t e .
Wi t h in a t ab le , d i f f e r ing in tens i t i e s o f a d i sease proce~ss can be de t e rmi ned by
d i f f e r e nc es i n t he m a g n i t ud e , o r i n t e ns i t y o f t he s t a t es . I n g l a u c om a , d i f f e r en t
i n t e ns i t i e s o f p r e s s u r e m a y be d i s t i ngu i s h e d by de f i n i ng st a t es o f m ode r a t e l y
e l e va t e d p r e s s u r e o r e x t r e m e l y e l e va t e d p r e s s u r e . T he s e s t a te s , w he n f ound i n
c l a s s i fi c a t i on t a b l es , m a y t he n l e a d t o d i f f e r e n t c onc l us i on s .
I n s om e i n s t a nc e s, i t i s ne c e s s a r y t o ha ve c l a s si f i c a ti on ru l e s w h i c h i nd i c a t e t ha t
s pec i fi c s t a te s a r e de n i e d . T he s a m e no t a t i on a nd i n t e r p r e t a t i on f o r a c l as s i f ic a t i on
t a b l e i s u s e d , w he r e e a c h e n t r y i n t he t a b l e i s no t a c on f i r m e d s t at e , bu t r a t he r t he
r e qu i r e d t r u t h va l ue ( c on f i r m e d , de n i e d , o r unde t e r m i ne d ) f o r t ha t s t a te . M ul t i p l e
c a us e s f o r a pa r t i c u l a r p a t i e n t s d i s e a se , m a y e i t he r be i nde pe nde n t o r r e l a t e d . I f
t he y a r e i nde p e nde n t , s e pa r a t e c l a s si f i c a ti on t a b l e s a r e r e qu i r e d . I f t he y a r e r e l a t e d ,
t he s a m e c l a s si f i c a ti on t a b l e s a r e r e f e r e nc e d f o r e a c h o f t he m u l t i p l e c a use s . R u l e s
w h i c h a r e ba s e d on t r u t h va l ue s ( a nd no t c o n f i r m a t i on a l one ) a r e u s e d t o d i s t i ngu i s h
s i t ua t i ons w he r e m u l t i p l e c a us e s c a nn o t be c la s si f ie d i nde p e nde n t l y .
W e c a n no w s um m a r i z e ou r d i a gnos t i c m e t ho d a s a s er ie s o f t r a ns f o r m a t i ons .
As t e s t r e su l t s a re r ece ived , they a re r e la ted to in d iv idu a l s ta te s. Th ese s ta te s a re
t he n o r ga n i z e d i n t o pa t hw a ys i n f e r r e d f r om c on f i gu r a t i ons o f a s t a te ne t w or k .
T he ge ne r a t e d pa t h w a ys a r e t he n r e l a t e d t o c l a s si f i ca t i on t a b l e s c on t a i n i ng t he
de t a i l e d d i a gnos t i c c a t e go r i e s .
T r e a t m e n t R e c o m m e n d a t i o n s
I n s o m e c a s e s a t h e r a p y r e c o m m e n d a t i o n c a n b e e x p l i c i t l y l i n k e d t o a s p e c i f i c
d i a g n o s t i c c o n c l u s i o n T h e r e m a y b e a u n i q u e t r e a t m e n t f o r a g i v e n c o n d i t i o n
-
8/10/2019 0004-3702(78)90015-2
20/28
164 S.M. WEISS, C. A. KULIKOWSKI, A. SAFIR AND S. AMAREL
I n o t he r i n s t a nc e s , a c a t e go r y o f t r e a t m e n t s m a y be de s c r i be d ( f o r e xa m p l e , t he
c l s s
o f m i o t i c m e d i c a t i o n s ) w i t h o u t a n i n d i c a t i o n o f a s pe c if ic m e d i c a t i o n . F o r
t he se s i m p li f i e d s i t ua t i ons , t he r e c om m e nd a t i on o f a t r e a t m e n t is a c on t i nu a t i on
o f t he d i a gno s t i c s t a t e m e n t f ound i n a c l a s si f i c a ti on ta b l e .
E xa m pl e 7 . T he c l a s s i f i ca t i on t a b l e o f E xa m pl e 6 m a y be a ugm e n t e d t o i nc l ud e
t r e a tm e n t r e c o m m e n d a t i o n s :
( 125, D1, T1), (/226, 9 2 , T2), (n 30, O 3, T3), (h a1, D., 7 4)
w he r e t h e t r e a t m e n t r e c o m m e n d a t i o n s a r e :
TI = re t ur n vis i t in 6 mo nt hs
T2 = c a r e f u l f o l l o w up w i t h r e pe a t e d t e n s i on r e a d i ngs
7 3 = ca re fu l fo l low up or a t he r ape ut i c t r i a l w i t h p i lo ea r p in e I ~o Q . I . D .
7 4 = m i o t i c t he r a py ( or , i f m e d i c a l l y unc on t r o l l e d , s u r ge r y )
The pa i r (D~, T~) a re l ink ed tog e th e r for th i s t ab le ; they may not a lway s be fou nd
t oge t he r f o r o t he r t a b l e s a nd o t he r m e c h a n i s m s o f d i s ea s e .
A r e c om m e nda t i on f o r t he r a py i s u s ua l l y a m or e c o m pl e x p r ob l e m t h a n de -
s c ri bed a bove . W hi l e t he num b e r o f po t e n t i a l t r e a t m e n t s t ha t a r e a pp l i c a b l e t o a
pa t i e n t m a y be g r e a t l y r e duc e d by t he p r e c i s i on o f t he d i a gnos i s , m a ny t r e a t m e n t s
m a y s ti ll r e m a i n f e a s i bl e . O ne o f t he s e t r e a t m e n t s m us t t he n be s e le c t e d. I n a dd i t i o n ,
onc e a t r e a t m e n t is r e c om m e nde d a nd g i ve n t o t he pa t i e n t , i t is i m por t a n t t o
e va l ua t e a nd m on i t o r t he e ff e c t i ve ne s s o f t ha t t r e a t m e n t .
W hi l e t he pu r pos e o f t he r a py is t o c on t r o l a n d i f pos s i b l e t o c u r e d i se a s e , t he
r e c o m m e n d a t i o n o f t r e a t m e n t o f t e n i n t r o d u c e s fa c t o rs w h i c h a r e e x t e r n al t o t h e
o r i g ina l d i a gnos i s . Spec i fi c t r e a t m e n t s m a y be c on t r a i n d i c a t e d be c a us e o f pa r t i c u l a r
c c nd i t i ons o f t he pa t i e n t t ha t do no t r e l at e d i r e c t l y t o t he d i s ea s e s t ha t a r e m ode l e d .
T he s e f a c t o r s m us t be c ons i de r e d be f o r e a t r e a t m e n t is r e c om m e nde d . Fo r e xa m pl e ,
age, a ll e rg ies , and h i s t ory of o the r i l lne sses may a ll p lay an i mpo r ta n t ro le in the
r e c om m e nda t i on o f a m e d i c a t i on . A t r e a t m e n t f o r d i s e a se m a y i n i t s el f c a us e ne w
pr oc e ss e s o f dys f u nc t i o n t ha t a r e un r e l a t e d t o t he o r i g i na l d i a gnos i s . M a ny
m e d i c a t i on s a r e know n t o c a use s i de e ff ec ts , a nd u nw a n t e d c om pl i c a t i ons m a y
e ns ue f r om s u r gi c a l p r oc e du r e s .
A p l a n o f a c t i on c a n be de s i g ne d t o se l ec t t r e a t m e n t s f o r pa t i e n t s w ho f a l l i n t o
a pa r t i c u l a r d i a gn os t i c c a t e go r y . A s t r a t e gy o f t r e a t m e n t s e l e c t i on a da p t s t he ge ne r a l
t r e a t m e n t p l a n t o t he s pe ci fi c c i r c um s t a nc e s o f a pa t i e nt . T he t r e a t m e n t p l a n m us t
t a k e i n t o a c c o u n t :
( a) t he e f f e ct i ve ne s s o f t he c u r r e n t t r e a t m e n t a nd
(b ) i nd i c a t i ons o r c on t r a i nd i c a t i o ns f o r va r i ous t he r a p i e s .
D i a gno s t i c c onc l us i ons f o r a pa t i e n t a r e f ou nd by i n t e r p r e t i ng t he s pe c if ic
obs e r va t i on s w i t h i n t he m ode l o f d i s ea s e . T he s e d i a gno s t i c c onc l u s i ons w i ll c on -
s i de r s e ve ri t y a nd p r og r e s s i o n o f d i s ea s e . T he d i a gno s t i c s t a t e m e n t s m a y t he n
po i n t t o one o r m or e t r e a t m e n t p l a ns . T he s e a r e s how n i n F i g . 4 . E a c h p l a n
-
8/10/2019 0004-3702(78)90015-2
21/28
M O D E L - B A S E D M E D I C A L C O N S U L TA T I O N 65
c o n s i s t s o f a n o r d e r e d l is t o f t r e a t m e n t s : T i t , T~2, . T ~,. T h e l i s t i s o r d e r e d b y
p r e f e r e n c e : t r e a t m e n t 1 i s t r i e d b e f o r e t r e a t m e n t 2 , w h i c h w i l l b e t r i e d b e f o r e
t r e a t m e n t 3 , e tc . T h i s p l a n r e p r e s e n t s a p r o t o t y p i c a l s e q u e n c e o f t r e a t m e n t s f o r
p a t i e n t s i n t h e a p p r o p r i a t e d i a g n o s t i c c a t e g o r i e s , a s a g r e e d t o i n a d v a n c e b y t h e
e x p e r t s i n t h e d o m a i n . A s t r a t e g y f o r r e c o m m e n d i n g t r e a t m e n t f o r a n i n d i v i d u a l
w i l l u s u a l l y f o l l o w t h e o r d e r o f t r e a t m e n t p r e f e r e n c e s . H o w e v e r , t h e o r d e r i n g m a y
b e c h a n g e d i n r e s p o n s e t o p a r t i c u l a r o b s e r v a t i o n s n o t e d i n t h e i n d i v i d u a l p a t i e n t .
W i t h i n a t r e a t m e n t p l a n , d e v i a t i o n s f r o m t h e p r o t o t y p e r e s u l t f r o m c h a n g e s i n t h e
d e g r e e s o f p r e f e r e n c e o r c o n t r a i n d i c a t i o n s f o r s p e c i f ic t r e a t m e n t s . I n c e r t a i n
s i t u a t i o n s , n o w e l l - e s t a b l i s h e d s e t o f p r e f e r e n c e s e x i s t s a n d t h e s e l e c t i o n o f a
t r e a t m e n t f r o m w i t h i n t h e g e n e r a l p l a n i s a l m o s t c o m p l e t e l y d e t e r m i n e d b y t h e
p a t t e r n o f o b s e r v a t i o n s f o r e a c h p a t i e n t .
T h e s t r a t e g y f o r t r e a t m e n t s d l e c t io n i s d e s c r i b e d a s f o l l o w s : W i t h i n a t r e a t m e n t
p l a n
Tk
e a c h s p e c i f i c t r e a t m e n t
Tkj
h a s a s s o c i a t e d w i t h i t a p r e f e r e n c e m e a s u r e
P f( T k ~ ) , w h i c h i s a s s i g n e d f r o m d i r e c t o b s e r v a t i o n s o f t h e p a t i e n t . E a c h o b s e r v a t i o n
t~ t h a t a f f e c t s t h e p r e f e r e n c e o f
Tkj
c o n t r i b u t e s a m e a s u r e P f ~j ( - - 1 ~< P f o ~< + 1),
w h i c h i s a s s i g n e d i n a m a n n e r s i m i l a r to t h e Q~m f o r r e l a t i n g o b s e r v a t i o n s t o t h e
d i s e a s e s t a t e s nm . F o r e x a m p l e , a d r u g i n t o l e r a n c e m a y a s s o c i a t e a n e g a t i v e
p r e f e r e n c e w i t h a p a r t i c u l a r t r e a t m e n t . F o r g l a u c o m a , a v ~ r y h i g h t e n s i o n r e a d i n g
a f t e r t r e a t m e n t w o u l d i n d i c a t e i n e f f e c t i v e c o n t r o l o f t h e d i s e a s e , a n d c o n t r i b u t e
a n e g a t i v e p r e f e r e n c e t o t h e c u r r e n t t r e a t m e n t . B e i n g in a p a r t i c u l a r a g e g r o u p m a y
i n c r e a s e t h e p r e f e r e n c e m e a s u r e o f o n e t r e a t m e n t o v e r a n o t h e r . T h e o v e r a l l
p r e f e r e n c e m e a s u r e P f i s c o m p u t e d b y t h e s a m e r u l e s u s e d t o c o m p u t e t h e c o n f i -
d e n c e m e a s u r e C f f o r t h e d i s e a s e s t a t e s . O n c e t h e P f h a v e b e e n c o m p u t e d , t h e
r u l e f o r s e l e c t i n g a s p e c i f ic t r e a t m e n t
Tki
f r o m w i t h i n i t s p l a n T k c a n b e s u m m a r i z e d
a s
( a ) S e l e c t
Tkj
s u c h t h a t P f ( T k i ) = m a x [ P f ( T k i) ]
( b ) I f t h e r e i s m o r e t h a n a s i n g l e t r e a t m e n t w i t h m a x i m u m P f , s e le c t t h e o n e
w i t h s m a l l e s t i n d e x j i n t h e a - p r i o r i p r o t o t y p i c a l o r d e r i n g .
E x a m p l e 8 . A t r e a t m e n t p l a n T 4 c o r r e s p o n d i n g t o a c o n f i r m e d c a s e o f o p e n
a n g l e g l a u c o m a ( D a ) , a s i n d i c a t e d i n E x a m p l e 7 , i s s h o w n i n F i g . 4 . T h e P f ( T 4 j )
a r e c o m p u t e d f r o m t h e o b s e r v a t i o n s o f a p a t i e n t , s o m e o f w h i c h a r e i l l u s t r a t e d i n
t h e f i g u r e . T h e p a t i e n t s h o w n i s c u r r e n t l y u n d e r t r e a t m e n t T 4~ , y e t t h e o b s e r v e d
t e n s i o n o f 2 7 m m o f H g i n d i c a t e s a n u n c o n t r o l l e d i n t r a o c u l a r p r e s s u r e . T h i s
a s s i g n s a d e c r e a s e d p r e f e r e n c e m e a s u r e o f - - 0 . 5 t o t h e c u r r e n t t r e a t m e n t T 4L a n d
t h e r e l a t e d t r e a t m e n t T a2 . T h e p a t i e n t a l so s h o w e d p r o g r e s s i o n o f f ie l d l o s s , w h i c h
d e c r e a s e s t h e p r e f e r e n c e ( - 0 .8 ) f o r t h e c u r r e n t m e d i c a t i o n T 4 ~ e v e n m o r e s t r o n g l y .
B e c a u s e t h e p a t i e n t i s u n d e ~ 3 0 , a s y s t e m i c m e d i c a t i o n s u c h a s D i a m o x i s l e s s
p r e f e r r e d ( - - 0 . 3 ) . T h e r e l a t i v e l y h i g h e r r i s k o f s u r g e r y v s. m e d i c a t i o n r e s u l t s i n t h e
a s s i g n m e n t t o T4 8 o f a d e c r e a s e d p r e f e r e n c e , - - 0 . 7 . A s a r e s u l t o f c o m p a r i n g t h e s e
-
8/10/2019 0004-3702(78)90015-2
22/28
66 S Mo ~TEtSS~ C A KUL IKO WSKI ~ A SAFI R AND S AMA REL
c~
-
8/10/2019 0004-3702(78)90015-2
23/28
M O D E L B A S E D M E D I C A L C O N S U L T A T I O N
6 7
and o ther Pf ' s der ived f rom the observat ions , the t rea tment wi th the maximum
preference for th is pat ient is T#5, which is recommended.
R e s u l t s a n d i s c u s s i o n
A genera l met hod fo r so lv ing a c lass o f d iagnost ic and therapy se lec t ion p rob lems
has been p resen ted . These ideas on model-ba sed in te rp re t a t ion have been pu t in to
pract ice th rou gh the impl eme nta t io n o f a com put er sys tem for medical decision-
making . Muc h exper ience has been gained in the deve lopmen t o f a model fo r the
d iagnosis a nd t rea tme nt o f g laucoma, which has led to the des ign o f a sys tem wi th
a h igh level o f exper t i se . Th is has in f luenced our genera l appr oac h towards
k n o wled g e rep resen ta t io n an d reaso n in g p ro cedu res . T h e co n su l t a t io n sys tem,
however, is not specif ic to g la u0oma. Oth er models of d isease have been developed
for the anemias , thyr o id dysfunct ion , diabetes, and hyper tens ion . Glauc oma,
however , i s the one app l ica t ion that has been pursued in dep th and has undergone
clin ical test ing.
The des ign o f a consu l t a t ion sys tem can be b roken own in to two imp o r tan t
tasks . These are the des ign and repres en ta t i on o f models , an d the design o f genera l
p rob lem so lv ing a lgor i thms which use a su i tab ly def iged model fo r decis ion-
making . In the genera l CASNET sys tem, a medical
xp rt
describes or modifies
a model , b u t does no t a l ter th~ reason i ng p rocedures that se lec t d iagnost ic in ter -
p re ta t io n an d t r ea tmen t p lan s . T wo sep ara te co mp u te r p ro g rams h av e b een
deve loped : the mode l ing p rog ram for des ign ing app l ica t ion models [19], and the
co n su l t a t i o n p ro g ram th a t u ses mod e l s for r each in g d iag no ses an d rec o mmen d in g
therapies [20].
The cur rent g l au com a cons ult at ion system has more t han 100 states, 400 tests ,
75 classif icat ion tables, and 200 diagnos t ic and tr ea tme nt statements. Results must
be in ter pret ed for e ach eye, so that , in effect , twice the numb er of rules are involved
in any oph t hal molo g ic al model . There are a lso many special ru les fo r b inoc u lar
comp ar i sons o f s ta tes , tes ts , and d iagnost ic and t r ea tm ent s ta tements . A set o f
the program's conclusions for a sample case is g iven in Fig . 5 . This session
i l lus tra tes the level o f per fo r manc e which lhe p rogr am has a t ta i ned in reason ing
abo ut complex cases o f g laucoma.
The co nsu l ta t ion p rogr am has been des igned fo r ef f ic ien t perfo rmanc e. Hum an
enginee r ing aspects o f p rogr am des ign have also been emphasized . The p r ogra m
has been developed pr im ar i ly as a too l fo r research o f medical decis ion-ma king
b y co mp u te r . Ho wev er , o u r ap p ro a ch to p ro g ram d ev e lo p men t in v olv ed th e
co l l ab o ra t io n o f a n e two rk o f ph y s ic ian s wi th min i mal p r io r ex per i en ce in th e
use o f computers . Thei r ac t ive par t ic ip at ion in the p ro ject requ i red careful a t ten t i on
to p ro g rammin g d e ta i l s wh ich wo u ld a l lo w o u r co l l ab o ra to r s an d o th e r
oph th almol og is ts to use the p rogra ms wi th l i t tle di f ficul ty. Th is impl ies that on ly
l imi ted typ ing would be requ i red , and qu ick response t ime, even fo r complex
d iagnost ic in terp re ta t ions would be essen t ia l .
-
8/10/2019 0004-3702(78)90015-2
24/28
68 S- ML WEISS C . A. KULI KOWS KI A. SAFIR AND S . AMA REL
*****VISIT
1 :
*
R IGH T EYE:
*
( 1 ) PR ESENT D IAGN OSTIC STATU S:
PIGNE NTARY GLAUCONA. OPEN ANGLE GLAUCOMA.
C H AR AC TER IST IC V ISU AL F IEL D L OSS WITH
CORRESPONDING DISC CHANGES. EARLY FIELD LOSS.
(2 ) TREATMENT RECOMH ENDATI ONS:
P IL OC AR PIN E Z~ OID .
* * * * R ESEAR C H STU D IES* * * *
AL TER N ATIVE IN TER PR ETATION S OF
GLAUCOMA=
.SECONDARY GLAUCOMA
.PRIMARY OPEN ANGLE GLAUCOMA
PIGMEN TAR Y
* * * * * V IS IT 7 :
R I G H T E Y E : *
(1 ) PRESENT DIAGNO STIC STATU S:
PIGM ENTA RY GLAUCOMA. OPEN ANGLE GLAUCOMA.
CHARACT ERIST IC VISUA L FIELD LOSS WITH
CORRESPONDING DISC CHANGES. ADVANCED FIELD
LOS S. CURRENT HED ICA TIO N HAS NOT CONTROLLE D
ZOP IN THIS EYE. (AS INDICA TED BY PROGRESSION
OF CUPPIN G) (AS INDICA TED BY VISUA L FIELD LOSS
PR OGR ESSION )
(2 ) TREATHENT
RECOMHENDATIDNS:
FILI ERI NG SURGERY IS INDICATED. AS AN
ALTERNATIVE, PHOSPHOLINE MAY BE TRIED BUT I;3T
USED 2 WEEKS BEFORE SURGERY).
l ~ o . 5. E x a m p l ~ o f p r o g r a m - g e n e r a t e d d c c i s i o n s f o r a c a s e o f p i g m e n t a r y g l a u c o m a a b s t r a c t e C
~om a seq ueno cof seve n v i si t s .
R E F E R E N C E S =
I . W H E N PIGMENTARY GLAUCOMA WAS FIRST
DESCRIBED IT WAS THOUGHT TO B E A FORM O F
SECONDARY GLAUCONA CAUSED BY PLU GGI NG OF THE
TRA BEC ULA R HESHWORK BY THE SAME PI GH ENT THAT
FORMED THE KRU KEN BER C S SPIND LES . HOWEVER, AN
INC REA SIN G NUMBER OF OBSERVERS NOW BEL IEV
