Nonlinear Computer Diagnosis and the Problem of Pathology in the
Hepatopancreatoduodenal Area
Nonlinear Computer Diagnosis and the Problem of Pathology in the
Hepatopancreatoduodenal Area
S.P. Tokar, A.S. Davydova,
T.L. Guseva, V.I. Gusarov,
Z.F. Khabibullina, L.S. Pugacheva
The problem of pathology in the hepatopancreatoduodenal area still remains urgent and
explains why researchers are keen on the search for improved diagnosis methods, since
the diagnosis proper is the starting point for determining the approach to treatment.
Today the basic methods for diagnosing a pathology in this region are traditionally
methods of direct artificial contrasting of the pancreatobiliary system, such as
endoscopic retrograde cholangiopancreatography (ERCPG) and percutaneous
transhepatic cholangiography (PTCG), well established in both diagnosis and treatment
of a number of diseases such as cholelithiasis, cysts and tumors in the head of pancreas,
tumor and corrosive strictures of the biliary ducts, tumors of Vater’s papilla, etc.
At the same time, the padioendoscopic methods of investigation of the biliary ducts,
though characterized by a rich diagnostic informational content owing to their
invasiveness, still do not eliminate the danger of serious complications, such as acute
pancreatitis, hyperamylasermia, cholangitis, sepsis, and allergic reactions, biliary flux
into the abdominal cavity with developing biliary peritonitis, hemorrhages, etc.
Their incidence rate varies from 0.8 to 36%. Besides, in the course of ERCPG different
technical problems may arise (failure in the cannulation of Vater’s papilla, the
impossibility to enter the duodenoscope at esophagus diseases, such as strictures,
achalasia, etc.). In addition ERCPG requires involvement of certain specialists like radiologic
diagnosticians, surgical endoscopists and anesthelists.
The advent of new diagnostic techniques in radiology and first of all ultrasound
scanning (US) and computer tomography (CT), did not produce a great limiting impact on
the use of ERCPG so far as these methods were not successful in solving a number of
diagnostic issues related to pathologies in the biliary system and pancreas.
The development of nonlinear computer diagnosis (NLS) as a method for diagnosing
abdominal pathology, opened up new opportunities for detecting diseases in the
hepatopancreatoduodenal region, with obstructive jaundice being one of their main
clinical implications. With the development and adoption of a number of speedy
programs for obtaining NLS images, specifically NLS- cholangiopancreatography,
which enables to obtain an integrated virtual picture of the biliary system and
pancreatic ducts without administration of contrasting agents and intervention into the
biliary system, the method was attempted to put into active use as an alternative to
ERCPG.
Some published works dealing with NLS have some distinct trends to pay more
attention to this issue with a view of obtaining sufficiently convincing information, that
57would allow to draw a final conclusion about a new relationship between integrated X-
ray endoscopic examination and in the first place between ERCPG and NLS, when
detecting a pathology in the hepatopancreatoduodenal region. Some of the works
suggested that NLS be used as a method preceding endoscopic cholecystoectomy.
With all the above in view, this paper aims to present our data on the role and
significant of NLS at certain diseases in the hepatopancreatoduodenal region.
To achieve this aim the following tasks were performed:
1. Examination of the test group to study different versions of a standard NLS-picture
of the biliary tract;
2. Description of the principal NLS semiotics in the patients with a pathology in the
hepatopancreatoduodenal region;
3. Cross-comparison of MRT, ERCPG and NLS for a more objective assessment of
the collected data;
4. Definition of clinical indications and diagnostic potentials of the NLS method for
the patients with obstructed biliary ducts.
Subject and methods
The NLS investigation was performed on 54 patients, of them 19 made a test group and
35 had different pathologies in the hepatopancreatoduodenal region, with 89% of these
patients showing signs of obstructive jaundice. The patients were from 36 to 77 years
old. There were 20 women and 15 men in the group of 35. As a primary method of
investigation all the patients had a sonography which acted as a screening tool for
performing NLS. A relative comparison of the results of MRT, ERCPG and NLS was
made for 18 patients.
The NLS investigation was carried out using a unit equipped with a 4.9 Ghz trigger
sensor.
We assessed the condition of the lymph nodes, especially in the portal fissure projection, and
the hepaticoduodenal ligament on the virtual images. We used the “Metapathia IT”
, a
special computer program for acquisition of a virtual image of the biliary system and
Wirsung’s duct.
Analysis of results
The virtual model distinctly visualized the common bile duct, common hepatic duct, right
and left lobar ducts and gall bladder (GB). The segmentary and sub-segmentary intra-
hepatic ducts are not actually visualized even in a poly-projection examination. The
normal lumen of the common bile duct is 0.6 cm; the NLS-signal coming from it is
homogeneously normo-chromatic (1-2 points according to Fandler’s chromatic scale).
The anatomical variations and abnormal developments occur very seldom, yet we
observed 3 cases of this kind, of them 2 contained an abnormal drainage of the cystic duct
58and 1 an atypically high point of entry of the cystic duct into the common hepatic duct. An
insufficient detailing of the papillo-sphincter region is the basic limitation of NLS in our
investigation was.
Calculi are known to be the most frequent cause of the bile duct obstruction. According
to our investigations, cholecysto-choledocholithiasis comprised 34% of all diagnosed
pathologies in the hepatopancreatoduodenal region. Regardless of their location, the
concrements in the biliary ducts were visualized on the NLS images as individual or
multiple hyperchromogenic zones (5-6 points), rounded or oval-shaped. The sizes of the
concrements detected in the hepaticocholledochus and lobar hepatic ducts varied from 5
to 20 mm. 6 patients had single concrements, and 4 had multiple concrements, and the
entire lumen of the hepaticocholledochus “stuffed” with concrements was found in 1
patient.
The localization of the concrements was variable. In 2 observations the concrements
only localized in the gall bladder and in 5 cases they did in the hepaticocholledochus; in 1
case the clinic laboratory evidence of obstructive jaundice was not found, and in 2 cases
the concrements were visualized in both the choledochus and lobal biliary ducts. In 5
cases we observed a concurrence of concrements in the gall bladder and choledochus.
In the course of our observations we arrived at a conclusion that the NLS-diagnosis of
concrements in the gall bladder depended on their size. So, as compared to the US data, the
concrements under 5 mm in diameter were largely not visualized on NLS shots,
because the signal from them was overlapped by a hyperchromogenic signal from the
mucous membrane. Small multiple concrements in the gall bladder that produce a low
entropy density signal (3-4 points according to Fandler’s scale) on NLS-shots hamper their
differential diagnosis because of sediment and putty-like bile. We agree with some authors
who consider the ultrasound scanning to be the “golden standard” in detecting gall bladder
concrements which should not be replaced by NLS.
The major NLS disadvantages in diagnosing concrements in the hepaticocholledochus are
associated with certain difficulties in assessing the choledochus condition, when the
choledochus is fully filled with concrements.
In one observation the concrement localized in some distal areas of the hepatico-
cholledochus, and on NLS shots it looked like a hyper-chromogenic oval-shaped defect
with the upper outline looking like a concrete lens. The combination of NLS-shots with
conventional MR-tomograms in axial plane allowed us to specify the spatial relationship
between the choledochus and head of pancreas and the duodenum, in other words. It
enables to detail the localization of the concrement in the ampullar region of the common
biliary duct.
Papillosphincterotomy was done during ERCPG with concrement extraction.
The genesis of benign strictures of biliary ducts was related to their surgical lesion or
inflammation caused by lithiasis, chronic pancreatitis or papillostenosis in 90-95% of
59cases. The number of iatrogenic lesions of the biliary system ducts grew up with the
extensive application of the laparoscopic cholecystectomy; because the intraoperative
investigation of the common bile duct is more complicated during laparoscopy than
during open surgery. In this connection, in terms of preoperative preparation for
endoscopic cholecystectomy, it is necessary to specify the anatomy of the
pancreatobiliary system and assess its condition in order to prevent potential iatrogenic
lesions of the biliary ducts.
So, owing to its noninvasiveness and high resolution, NLS can be a diagnosis-
determining method for this kind of patient. Unlike ERCPG, NLS enables to visualize
the bile ducts above and below the obstructive level, which is displayed on both MRT
and NLS shots. The latter method gives a virtual physiological picture of the condition of
hepatic and pancreatic ducts as compared to ERCPG, in which the administration of a
contrasting agent overstates the extent of duct dilatation.
In all of our observations NLS allowed to define the accurate extent of the arctation, its
length and cause. In 2 cases the arctations localized at the cystic duct level, which was
indicative of their iatrogenic genesis. In 1 case it was an actation hepatico-
cunoanastomosis. In 5 observations the arctations from 1.5 to 2.0 cm long were located
at the confluence and in the proximal region of the hepaticocholledochus.
In assessing the arctation extent in the case where the lumen was not visualized on NLS-
shots, we always analyzed the native MRT scans and supplemented the investigation with
thin sections, which allowed forming a more exact opinion about the arctation extent. At
the same time, comparing the results of nonlinear diagnosis to ERCPG one must admit
that the latter method is more exact in determining the extent of duct affection.
However the essential criterion in deciding on the surgical correction method enables not
only to detect the level and length of a structure, but also clearly specify the pattern of
cholledochus deformation in presence of an arctation, which also determines the surgical
approach to the reconstructive operations.
The combination of conventional MRT and NLS considerably enhances the diagnostic
potential of this method as opposed to ERCPG in diagnosing chronic or acute
pancreatitis, because it enables not only to investigate the condition of the ducts of the
pancreatobiliary system, but also to assess both the pancreas proper and the adjacent
organs and structures. Of 4 our observations of chronic pancreatitis in 1 case we had a
frank contraction of the intrapancreatic part of the cholledochus cause by a chronic
inflammatory process, in 3 patients the constriction of the distal part of the cholledochus
was caused by a cyst in the head of pancreas. On the NLS-shots the obstruction of the
biliary duct looked cone-shaped, and its affected part could be visualized all the way
along the head of the pancrease including Vater’s papilla area, and was assessed at 5-6
points according to Flandler’s chromatic scale.
In all the cases a moderate chromogenic density of dilated biliary ducts and a
heterochromous response of Wirsung duct occurred. The cysts, where they existed, were
60depicted on MRT and NLS shots and the constricted area of the common biliary duct had an
arc-shaped route because of being forced back by the cyst. The NLS allowed
detection of a relation between the Wirsung duct and pancreatic cysts.
The most common and well-known causes of the biliary tract obstruction are the tumors
localised in different organs: liver, biliary extra hepatic ducts, head of pancreas, major
duodenal pupilla, as well as metastases into the hepatoduodenal ligament and portal
fissure. The tumors concentrated in these locations are conventionally called the “tumors
of the hepatopancreatoduodenal region”
. The reason for that is common clinical
implications relate to the obstruction of biliary and pancreatic ducts. In patients affected
by malignant tumors localized in this area the primary symptom of the disease is
generally obstructive jaundice. A decision on the expediency of surgery for this kind of
patients necessitates an assessment of a clinical prognosis depending on the tumor state
according to the TNM system. So, if a malignant pattern of obstructive jaundice was
suspected, then together with the elimination analysis, which is a special program, we
always carried out standard investigations in the AUTO TUNE mode which enabled
localization of the tumor and assessment of its spread to the adjacent structures as well as
defining hematogenic and lymphogenous metastases.
We observed 4 cases of cholangiocarcinoma with obturation of intra- or extrahepatic
ducts including 2 cases with a tumor localized at the common hepatic duct level and 2
cases with affected intrahepatic ducts and liver parenchyma. In all of the 4 cases the NLS
allowed to precisely localize the lesion level and define its length. Both cases of
cholangiocarcinoma displayed a spectral similarity to the reference standard process “liver
carcinoma” (D from 0.213 to0. 418).
Researchers observed an increased chromogenic density (5-6 points) in intra-hepatic
ducts more proximal to the arctation. In one of the observations, in the projection of the
constriction of the common biliary duct, the MRT scans displayed a soft tissue structure,
up to 3 cm in diameter with a medium intensity signal, which enveloped the duct in a
sleeve-like manner at the lesion level and was indicative of a tumor etiology of the
structure. Another observation at stenosis of the common biliary duct with no imaging of
the tumor tissue, detected enlarged lymph nodes in the lesser omentum region and a
single metastasis into the liver which allowed us to correctly interpret the pattern of the
lesion confirmed by histological investigation of the biopsy material acquired during
trans-hepatic drainage for decompressing the biliary ducts. In 1 of the 2 observations of
cholangio-carcinoma of intra-hepatic ducts also histologically confirmed later, we drew
an erroneous conclusion, because the pattern of MR-changes, i.e. a small ectasia of the
intra-hepatic ducts by a varicose type above the moderately constricted common hepatic
duct and unaffected hepatic duct more distal from the structure, a rather long anamnesis
of the disease (the patient had had itching fits, occasionally icteric integument, de-colored
feces and dark urine for ten years) and the obscure clinical presentation were interpreted
by us (and during ERCPG) as manifestations of a primary sclerogenic cholangitis.
In all three our observations of pancreas head cancer the obstruction of the biliary duct
looked on NLS shots like a progressively growing chromogenic density of the ecstatic
61biliary duct at the level of its intra-pancreatic area. A frank hyper-chromogenic pattern of
the intra-hepatic ducts was concurrently noted. The Wirsung duct was unevenly hyper-
chromatic too. The standard AUTO TUNE shots allowed us to evaluate the spread of the
tumor to the adjacent structures and determine some hematogenic and lymphogenous
metastases. The structure of the tumor itself could be better visualized on the front shots.
In one of the cases we also detected both metastases into the liver, and enlarged lymph
nodes in the supra-pancreatic cellular tissue. The histological investigation confirmed the
adencarcinoma in all three cases.
Differential diagnosis of pancreas head cancer and chronic pseudotumor pancreatitis is a
very complicated task and until now has been a problem yet to be solved. Integrated
abdominal NLS-investigation with visual simulation of changes in the head of pancreas
(not always pathognomonic for a tumor) and its spectral similarity to blastic process (D<
0.425) allowed us to produce a more defined opinion about the tumor pattern of the
process. In addition, the value of the NLS consists in its capability to assess the
spreading of the tumor to the cystic duct which is very essential for making a decision on
the expediency of cholecystodigestive anastomosis.
According to some literary evidence, differential diagnosis of diseases of the
papillosphincteral region by means of NLS investigation is very difficult. In one of our
observations of a patient which clinical lab manifestations of obstructive jaundice the US
detected a distal block of cholledochus with dilated superjacent regions of the biliary
system. The duodenoscopy detected signs of frank papillate of the major duodenal
papilla and suspected infiltration of the longitudinal fold of the duodenum. The
percutaneous transhepatic cholangiography (PTCG) that was performed additionally
showed a picture of stenosed papillosphincteral region. The NLS displayed
hyperchromatic intrahepatic ducts and hepaticocholledochus, with the latter traceable as
far as the level of entering the duodenum. The NLS investigation performed in a sagittal
projection allowed to detect an additional structure with a moderately chromogenic signal
in the ampullar region of the cholledochus. It allowed a supposition about a tumor lesion
of the major duodenal papilla. The patient died of DVS syndrome. In section: poorly
differentiated adenocarcinoma of the major duodenal papilla. We would like to use this
example to stress the importance of a correct methodical approach for carrying out an
investigation in a specific diagnostic situation.
Thus, our results were able to show that NLS had great prospects for diagnosing a
number of diseases with such a common anatomic-functional concept as a pathology in
the hepatopancreatoduodenal region. So, thanks to its noninvasiveness, no need for
administering contrasting agents and intervening into the biliary system as well as the
ability to combine with conventional MRT, which substantially increases the diagnostic
importance of the method, in quite a lot of cases the NLS investigation can replace radio-
endoscopic methods of investigation (ERCPG and PTCG) as a primary diagnosis method
although the other methods can retain their therapeutic functions. NLS may also become
a method of diagnosis choice for patients who have some counter indications to X-ray-
endoscopic investigation. We believe it necessary to actively expand the use of NLS with
a view to making an ultimate decision on its clinical diagnostic importance.
