Diagnostic accuracy of tissue harmonic imaging in obstructive jaundice.
Introduction: Tissue harmonic imaging (THI) is a new ultrasound imaging technique providing the images which are much superior in quality compare to conventional ultrasound images by improving lateral resolution, signal to noise ratio and reducing side-lobe artifacts.
Aim: Our aim is to estimate diagnostic accuracy of THI in diagnosing the cause of obstructive jaundice and to compare it with previous studies.
Material and method: 125 cases of obstructive jaundice were evaluated with THI for the cause of obstructive jaundice by three radiologists separately during period of two year interval on Phillips HD 11 XE ultrasound scanner. Follow up done with ERCP, surgery and histopathology. CT and MRCP done wherever needed. Sensitivity, specificity, PPV, NPV, accuracy, positive likelihood ratio, negative likelihood ratio, diagnostic odds ratio and Youden’s index were calculated and compared with previous studies.
Result: In our study we found choledocholithiasis as most common cause for obstructive jaundice. Sensitivity of 80%, 57.49%, 70.97% and specificity of 90%, 90.59%, 89.36% was noted for choledocholitiasis, neoplasm and benign stricture respectively. Accuracy of 86.4%, 80%, 84.8% was found for choledocholitiasis, neoplasm and benign stricture respectively. PPV of 81.81%, 74.19%, 68.75% and NPV of 88.88%, 81.91%, 90.32% was observed for
choledocholithiasis, neoplasm and benign stricture respectively. Positive likelihood ratio of 8, 6.10, 6.67 and negative likelihood ratio of 0.22, 0.46, 0.32 was noted for choledocholithiasis, neoplasm and benign stricture respectively. Diagnostic odds ratio and Youden’s index of 36 and 0.7, 13.02 and 0.48, 20.53 and 0.60 was seen for choledocholithiasis, neoplasm and benign stricture respectively.
Conclusion: Introduction of THI in the modern ultrasound equipment had made improvement in the diagnosis of cause for biliary tract obstruction. Till today even with THI ultrasound doesn’t have higher diagnostic value.
Key words: Tissue harmonic imaging, obstructive jaundice, choledocholithiasis, neoplasm, benign stricture.
Jaundice is a symptom characterized by yellowish discoloration of tissues and body fluids due to an increase in the bile pigments.1 It is of two types: obstructive (surgical) or nonobstructive (medical). Obstructive jaundice may be attributable to multiple causes including stones, intrinsic tumor, stricture or compression by extrinsic masses.2,3,4 It can present with jaundice with or without pain, dark urine, pruritus, pale stools, weight loss and anorexia.5
There are various investigations which could be carried out for the diagnosis of obstructive jaundice like ultrasonography, CT, MRCP, ERCP. After Laboratory investigations ultrasonography of abdomen is considered first preliminary investigation of choice for evaluation of obstructive jaundice due to its accessibility, speed, ease of performance and low cost.6 Modern ultrasound equipments with inbuilt THI facility has contributed to significant improvement in sonographic performance for the assessment of biliary tract.7
Tissue harmonic imaging was first utilized as a technique for the detection of nonlinear vibrations of microbubble contrast agent.8 It is based on the phenomenon known as nonlinear propagation.9 Now THI is also used to image body tissue. Conventional ultrasound waves are generated at the surface of the transducer and its intensity decrease progressively as they traverse the body. In comparison harmonic waves are generated within the tissue and build up to a point of maximum intensity before they decrease due to attenuation. Harmonic wave frequencies are multiples of the transmitted frequency and present technology uses only second harmonic for imaging.10
Relatively small amplitude of the harmonic waves reduces the detection of echoes from multiple scattering events and side lobe artifacts are less likely to occur in THI. Shorter wavelength and improve focusing with higher frequencies had improved the axial and lateral resolution respectively.10
New ultrasound equipments with THI has made it necessary to redefined the diagnostic value of ultrasound with tissue harmonic imaging in detection of cause for obstructive jaundice. It is important preoperatively to determine the cause of obstruction because ill-chosen procedure can lead to high morbidity and mortality. We describe our experience regarding the diagnostic accuracy of sonography in detection of the cause of obstructive jaundice using modern equipment with tissue harmonic imaging.
Material & method:
Our study population included 125 patients referred from gastroenterology department of our institute for abdominal ultrasound examination over a period of past two years intervals having confirmed diagnostic cause of obstructive jaundice on follow up. All patients were above 18 year of age. Patients having previous history of pancreatico-biliary surgery, patients with inconclusive ultrasound finding, patient who lost follow up were excluded from the study.
All the scans were obtained on sonographic equipment with tissue harmonic imaging capability, Phillips HD11XE Scanner. Sonographic probe used was C5-2 MHz. Harmonic mode setting done by means of a toggle switch on scanner control panel. All patients underwent abdominal ultrasound examination in which cause of obstruction was observed by three expert radiologists separately. These radiologists were had designation assistant professor and above. Bile duct was measured from inner border to inner border. CBD greater than 8mm was labeled as abnormal. For USG of pancreatico-biliary tree, fasting for six hours was ensured. Scanning was done in supine and left lateral position. Pressure was applied wherever needed to displace the bowel gases. In some cases oral water was administered for better demonstration of pancreas. When anterior epigastric approach fail to demonstrate distal CBD, right lateral or anterolateral approach with patient in left posterior oblique position was helpful. Out of three radiologists the cause confirmed by two or more was labeled as a cause on ultrasound.
Any intraluminal, hyperechoic structure with or without posterior acoustic shadowing was labeled as calculus. Intraluminal or intra-extraluminal hypoechoic or isoechoic mass causing abrupt interruption of dilated bile duct was labeled as neoplasm. Smooth tapering of dilated bile duct without any mass lesion was labeled as benign stricture.
All the patients were followed for ERCP, surgery and histopathology. CT and MRCP were done wherever needed for diagnosis. Data was analyzed on statistical software SPSS.
Out of 125 patients youngest was 20 year old and oldest was 83 year old. Maximum numbers of patients were in the age group 50-59 year. Out of 125 patient studied 68 were male and 57 were female. Major presenting complaints were jaundice (100%), abdominal pain (60%), nausea & vomiting (58.7%), dark colored urine (52.5%), pruritus (43.7%). In our study, we had found choledocholithiasis in 45, pancreatico-biliary neoplasm in 40, benign strictures in 31 and miscellaneous causes in 9 as a cause for obstruction.
Out of total 45 cases of choledocholithiasis (Figure 1), 36 were detected on THI (Figure 4). Nine patients were false negative and eight patients were false positive on THI. We found sensitivity, specificity, PPV, NPV, accuracy of 80%, 90%, 81.81% 88.88% and 86.4% respectively in case of choledocholithiasis. Positive and negative likelihood ratios were 8 and 0.22 respectively. Diagnostic odds ratio of 36 and Youden’s index of 0.70 was observed for choledocholithiasis (Table 1).
Figure 1: Ultrasound image (A) showing hyperechoic structure in common hepatic duct giving posterior acoustic shadowing which on ERCP (B) seen as oval filling defect representing a calculus which was removed during ERCP.
THI had detected 23 cases of pancreatico-biliary neoplasm (Figure 2) out of total 40 cases (Figure 4). Eight cases were false positive and seventeen cases were false negative on THI. For pancreatico-biliary neoplasms we had found sensitivity, specificity, PPV, NPV, accuracy of 57.49%, 90.59%, 74.19%, 81.91% and 80% respectively. Positive likelihood ratio of 6.10 and negative likelihood ratio of 0.46 was observed. Diagnostic odds ratio of 13.02 and Youden’s index of 0.48 was found for pancreatico-biliary neoplasm (Table 1).