الفهرس 
INTRODUCTIONOnly 14 pages are availabe for public view
 |  | from | 59 |  |  |
| | | from | 59 | | |
Abstract
Evaluation of lymph nodes by different imaging modalities has always been a diagnostic challenge; particularly differentiating between benign and malignant ones. If imaging can reliably differentiate between them, we can reduce the number of biopsies and the associated morbidities. Many studies showed the ADC value derived from DWI can be beneficial n characterization of lymph nodes. Less studies were made on CSI and their ability in characterization of LNs; they all showed promising results and required further research.
The aim of the study is to evaluate the potentiality of CSI and DWI in characterization of LNs.
Our study included 31 patients referred from the Otorhinolaryngology and head and neck ward with known head and neck primary for assessment of metastasis or clinically detected LNs to the Radio-diagnosis department of Alexandria Main University Hospitals for MR imaging. The MR diagnoses were correlated with histolopathological findings from biopsy or, in some cases, with clinical follow up. The biopsy was taken ultrasound guided from the studied node. Excisional biopsy was only useful if the whole level was benign or malignant as pathology reports does not specify which node was found malignant; they are mentioned together.
Our study showed that mean ADCs of malignant LNs (mean ADC 0.85 ± 0.24×10-3 mm2/s) were significantly lower than benign LNs (mean ADC 1.06 ± 0.25×10-3 mm2/s) (p-value = P<0.041, significant if P≤0.05). An ADC cutoff point of 0.9 mm2/s was best for differentiating benign from malignant LNs; it provided 72.77% sensitivity, 88.7 % specificity.
On studying the role CSI, we studied a ratio between in the In and Out of phase image to study the signal drop. Our study showed that mean In/Out phase ratio of malignant LNs (mean 0.96 ± 0.06) were significantly higher than benign LNs (mean 0.75 ± 0.16) (p-value = P<0.001, significant if P≤0.05) denoting the loss of the fatty hilum. A cutoff point of 0.9 was best for differentiating benign from malignant LNs; it provided 95.45 % sensitivity, 88.89 % specificity.
We studied an inflammatory LN; known patient with Tuberculosis (it was not included in the statistical analysis). Both DWI and CSI failed to differentiate it from malignant nodes and perceived same findings as malignant nodes. Thus, limiting the role of CSI and DWI in assessing the nodal status in known primaries rather than initial approach of the cervical lymphadenopathy.
Our study showed that CSI is a strong diagnostic tool in characterization of LNs. It could be used alone or along with DWI providing better results than DWI.