Preclinical Molecular Imaging technologies have an increasingly broader application base while they at the same time are becoming more user-friendly. Tail vein injections are a routine but critical step in most imaging applications, with poor injections greatly affecting

the experimental results. The high skills and experience required to perform successful tail vein injections leave many preclinical imaging scientists ill-suited to perform this task. In a recent study, we found that trained routine injectors left on average 14% of the injected probe in the tail tissue. Improvements in injection accuracy, injection consistency, safety, and a reduction in time required to perform the task are needed in preclinical molecular imaging. To achieve these goals, we have devised a semi-automated vascular access system (VAS) to facilitate injections and eventually blood sampling from the mouse tail. We have eliminated much of the human error involved in the manual approach by using a computer-controlled mechanically moving micro-needle.

To make use of the VAS, one places an anesthetized mouse onto the temperature controlled mouse bed and secures the tail on a heated tail holder. The VAS uses NIR light, cross-polarizers, and a basic CCD camera to image the tail. The reflection image is processed and the vein is located using edge detection methods. The vein location is plotted and over-laid onto the live video feed of the mouse tail. Using a custom designed user interface, the user properly aligns the needle to the tail vein by employing computer-controlled motors. Once the needle is properly aligned, it begins to penetrate the tail tissue and enter the vein. A pressure transducer attached to the needle detects when the needle has entered the vein, and automatically stops further progression of the needle. With the needle inside the vein, probes can be injected manually, with a liquid handling system, or via a syringe pump. The VAS was first validated using a mouse tail phantom. The phantom was a PDMS chip with channels equivalent to the dimensions of a mouse tail vein (300um). The channels were filled with water and pressurized to variable pressures. With the phantom, the ability for the VAS to align a needle according to an image, insert a needle into a desired location, and stop the progression of the needle based on a pressure signal were tested and verified. Mouse studies were also performed with the VAS. These studies showed that the accuracy of the device, as measured by the percentage of injected probe left in the tail, is 3.4% (+/- 4.5). The VAS reduces the operator skill requirements and training, has the potential to improve injection accuracy, reduces the time required to perform a tail vein injection, and is potentially safer for users and mice in comparison to current manual methods.