Sensitivity of ACL volume and T21 relaxation time to magnetic resonance imaging scan conditions

Anterior cruciate ligament (ACL) volume and T₂¹ relaxation times from magnetic resonance (MR) images have been previously shown to predict the structural properties of healing ligaments. We investigated whether MR imaging scan resolution and condition (in vivo, in situ, or ex vivo) affected ACL volume and T₂¹ relaxation times in intact ligaments. ACLs of 14 pigs were imaged using a 3 T scanner and a six-channel flexcoil using at least two of four possible scan conditions: (1) in vivo moderate resolution (n = 14); (2) in vivo high resolution (n = 7); (3) in situ high resolution acquired within 60 minutes of euthanasia (n = 6); and (4) ex vivo high resolution following hind limb disarticulation and one freeze-thaw cycle (n = 7). T₂¹ relaxation times were mapped to the ACL voxels. The total ACL volume was then divided into four sub-volumes (Vol_1–4) based on predetermined increasing ranges of T₂¹ times. ACL T₂¹ statistics (first quartile, median, and standard deviation (SD)) were computed. Scan resolution had no effect on the total ACL volume, but Vol₁ and first quartile T₂¹ times decreased with high resolution and in situ/ex vivo scan conditions. The most dramatic differences in T₂¹ summary statistics were between in vivo moderate and ex vivo high resolution scan conditions that included a freeze-thaw cycle: ACL T₂¹ SD increased by over 50% in 9 animals, and more than 90% in 4 animals. Our results indicated that T₂¹-based prediction models to quantify in vivo structural properties of healing ligaments should be based on high resolution in vivo MR scan conditions.