الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 16 |  |  |
| | | from | 16 | | |
Abstract
Magnetic resonance (MR) imaging is the favored modality for evaluation of soft tissue tumors and tumorlike conditions because of its superior soft tissue contrast, multiplanar imaging capability, and lack of radiation exposure. MR imaging is valuable for lesion detection, diagnosis, and staging. (1) For many patients, conventional MR imaging is unable to provide a specific histologic diagnosis or assess the true extent of viable or necrotic malignant tumor, factors that are important to determine the response to treatment and patient prognosis.(2) The role of magnetic resonance (MR) imaging in the evaluation of musculoskeletal tumors continues to evolve as newer pulse sequences emerge, MR imaging may also be used for the detection, characterization, and assessment of a tumor after treatment either after neoadjuvant therapy before surgery for the assessment of treatment regression as well as postsurgical assessment of residual or recurrent disease.(3) This study was conducted upon 110 patients with suspected bone or soft tissue tumour as well as for a known patient of bone or soft tissue tumour on chemo-radiotherapy or following surgical resection for initial detection and characterization as well as for post therapeutic surveillance. Osseous lesions showed mean ADC value of benign lesions ranged 0.9 - 3.2 X 10-3mm2/sec of mean ±SD (1.9±0.6). In malignant lesions, the mean ADC value ranged 0.6 – 1.9 X 10-3mm2/sec of mean ±SD (1.1±0.4). Benign and malignant tumours show statistically significant ADC values at P<0.0001.Differentiation between malignant and benign tumours at a threshold of 1.1 X 10-3mm2/sec was possible with sensitivity, specificity, positive predictive value, negative predictive value and accuracy of 94.1%, 70.3%, 59.3%, 96.3%, 77.8% respectively. SIR for benign lesions ranged 0.2-1.0 with mean ±SD (0.6±0.3). For malignant lesions SIR ranged 0.1-1.2 with mean ±SD (0.8±0.3). Benign and malignant tumours show statistically significant SIR values at P<0.022. Differentiation between malignant and benign tumours at a threshold of 0.75 was possible with sensitivity, specificity, positive predictive value, negative predictive value and accuracy of 70.3%, 76.5%, 86.7%, 54.2% and 72.2% respectively.
Regarding perfusion, the most common curve type in benign tumours was type II (n=15, 83%), whereas the most common curve in malignant tumors was type IV (n=9, 60. Type (V) curve was the most common type in post therapy malignant tumours follow up (n=4, 45%). Although there was an overlap in types of curves, results are statically significant in differentiation between benign and malignant tumors with P<0.0001 (using Monte Carlo corrected P-value significant at P≤0.05).
All follow up osseous tumours showed no diffusion restriction and higher ADC values, low SIR values as well as type II-V perfusion dynamic curves in case of good responders. Poor responders, showed low ADC values, high SIR values as well as type IV perfusion dynamic curves. For benign soft tissue tumours, diffusion restriction occurred in benign six tumours. The lowest value was for giant cell tumour. Median value of wash in rate and maximum enhancement of benign tumours was 4 and 1630 respectively. Malignant soft tissue tumours, diffusion restriction occurred in (75%) of tumours, the lowest value was for rhabdomyosarcoma Type IV dynamic curve was the most common among malignant tumours (45%). Median value of wash in rate and maximum enhancement of malignant tumours was 21 and 525 respectively. Perfusion showed higher sensitivity than diffusion in cauterization of soft tissue tumours.