Super Resolution for Improved Multislice Prostate MRI - PROJECT SUMMARY / ABSTRACT Prostate cancer (PCa) is the most common and second most deadly non-cutaneous cancer in men in the USA, but five-year survival rates approaching 100% are possible if detected early. Prostate MRI has become increasingly important in cancer detection and localization. Two-dimensional (2D) multislice T2-weighted spin-echo (T2SE) and diffusion-weighted (DWI) acquisitions are mainstays of clinical prostate MRI, but radi- ological interpretation is hindered because through-plane resolution is 3–8× coarser than inplane resolution. This is particularly limiting in the transition zone (TZ) of the prostate where differences in relaxation times between malignant and normal tissue are small, and interpretation is often based on sharpness-related criteria such as encapsulation. Per PI-RADSv2.1, T2SE is the critical sequence. Reported sensitivity is far less than ideal, 30 to 70%. In addition to the specific need for improved diagnostic accuracy in the TZ, there is an initiative to make prostate MRI exams shorter and more efficient. Thus, there is a need to reconcile these competing demands of improved resolution and contrast via reduced partial volume effects with higher effi- ciency in prostate MRI. We have developed a super resolution (SR) method for T2SE which uses overlapped, standard-thickness (3mm) axial slices to form 1mm slices with reconstruction in kZ-space, dubbed kZ-multislice (KZM). We have shown in vivo the superior sharpness of the resultant images vs. 3mm slices for equal inplane resolution. More recently we have developed methods which virtually eliminate the slice-to-slice in- consistency disrupting current KZM reconstruction and degrading image quality. Our central hypothesis is that this KZM approach shown in feasibility studies can now be markedly improved technically to simultane- ously provide improved through-plane resolution and allow reduced scan time, resulting in an efficient tech- nique enabling improved performance of clinical prostate T2SE and DWI MRI. We propose to: 1. Develop techniques for rapid, high quality, multislice acquisition suitable for SR prostate T2SE. We will incorporate non-traditional RF excitation pulses for improved fidelity in kZ. We will use simultaneous multi- slice (SMS) to reduce scan time and statistical reconstruction to retain SNR. We will develop means to measure the small, 1-2 mm A/P prostate motion during the scan, allowing accurate motion correction. 2. Extend the SR methodology developed for T2SE to DWI of the prostate. Segmented acquisition will provide for slice-to-slice positional consistency. SR reconstruction will be tuned to the individual b-values used. Sta- tistical reconstruction will retain SNR, allowing thin-slice images of the ADC. 3. Apply SR techniques to multislice prostate MRI. We will evaluate our technical developments in clinical evaluations, comparing conventional T2SE with our new SR method for a Super Abbreviated Exam (SAE). We will also evaluate our T2SE KZM method on improved accuracy in detecting PCa in the transition zone.