Aim: The aim of this study was to test the feasibility for differentiation of serous fluid collections and abscess fluid with diffusion-weighted imaging and to compare three different diffusion sequences.
Material and methods: Thirteen puncture samples (6 serous, 7 purulent) and 19 patients with either serous fluid collections (n = 12) or abscesses (n = 7) were examined with diffusion-weighted spin-echo (SE), stimulated-echo (STE), and fatsuppressed SE (SE-SPIR) sequences. For in vitro studies 4 different b-factors (0, 87, 355, 798 s/mm2 in SE and SE-SPIR and 0, 51, 204, 460 s/mm2 in STE) were chosen and the apparent diffusion coefficient (ADC) was calculated. For in vivo measurements identical sequences with two b-factors (0.598 s/mm2 in SE and SE-SPIR and 0.360 s/mm2 in STE) were applied and the normalized signal attenuation was calculated. A navigator-echo technique and peripheral pulse triggering was used for motion artifact reduction.
Results: The in vitro study yielded an ADC of serous fluid, which was close to that of free water, whereas for purulent fluid a significantly lower ADC was calculated. During in vivo examinations, serous fluids showed a strong signal attenuation (down to 22-32% of basic value) compared to a minor signal attenuation in purulent fluids (down to 86-94% of basic value) (p < 0.05).
Conclusions: In summary, with all three investigated diffusion schemes serous and purulent fluids can be clearly differentiated.