Multiphase CT angiography perfusion maps for predicting target mismatch and ischemic lesion volumes.

The complexity of CT perfusion (CTP) acquisition protocols may limit the availability of target mismatch assessment at resource-limited hospitals. We compared CTP mismatch with a mismatch surrogate generated from a simplified dynamic imaging sequence comprising widely available non-contrast CT (NCCT) and multiphase CT angiography (mCTA). Consecutive patients with anterior circulation acute ischemic stroke who received NCCT, mCTA, and CTP were retrospectively included in this study. An mCTA-perfusion (mCTA-P) dynamic series was formed by co-registering NCCT and mCTA. We simulated an ideal mCTA-P study by down-sampling CTP (dCTP) dynamic images according to mCTA timing. Ischemic core and penumbra volumes were estimated by cerebral blood flow and Tmax thresholding, respectively, on perfusion maps calculated independently for CTP, dCTP, and mCTA-P by deconvolution. Concordance in target mismatch (core < 70 ml, penumbra ≥ 15 ml, mismatch ratio ≥ 1.8) determination by dCTP and mCTA-P versus CTP was assessed. Of sixty-one included patients, forty-six had a CTP target mismatch. Concordance with CTP profiles was 90% and 82% for dCTP and mCTA-P, respectively. Lower mCTA-P concordance was likely from differences in collimation width between NCCT and mCTA, which worsened perfusion map quality due to a CT number shift at mCTA. Moderate diagnostic agreement between CTP and mCTA-P was found and may improve with optimal mCTA scan parameter selection as simulated by dCTP. mCTA-P may be a pragmatic alternative where CTP is unavailable or the risks of additional radiation dose, contrast injections, and treatment delays outweigh the potential benefit of a separate CTP scan.

CT perfusion stroke lesion threshold calibration between deconvolution algorithms.

CTP is an important diagnostic tool in managing patients with acute ischemic stroke, but challenges persist in the agreement of stroke lesion volumes and ischemic core-penumbra mismatch profiles determined with different CTP post-processing software. We investigated a systematic method of calibrating CTP stroke lesion thresholds between deconvolution algorithms using a digital perfusion phantom to improve inter-software agreement of mismatch profiles. Deconvolution-estimated cerebral blood flow (CBF) and Tmax was compared to the phantom ground truth via linear regression for one model-independent and two model-based deconvolution algorithms. Using the clinical standard of model-independent CBF < 30% and Tmax > 6 s as reference thresholds for ischemic core and penumbra, respectively, we determined that model-based CBF < 15% and Tmax > 6 s were the corresponding calibrated thresholds after accounting for quantitative differences revealed at linear regression. Calibrated thresholds were then validated in 63 patients with large vessel stroke by evaluating agreement (concordance and Cohen's kappa, κ) between the two model-based and model-independent deconvolution methods in determining mismatch profiles used for clinical decision-making. Both model-based deconvolution methods achieved 95% concordance with model-independent assessment and Cohen's kappa was excellent (κ = 0.87; 95% confidence interval [CI] 0.72-1.00 and κ = 0.86; 95% CI 0.70-1.00). Our systematic method of calibrating CTP stroke lesion thresholds may help harmonize mismatch profiles determined by different software.

Feasibility of deconvolution-based multiphase CT angiography perfusion maps in acute ischemic stroke: Simulation and concordance with CT perfusion.

OBJECTIVES: Integration of CT perfusion (CTP) with requisite non-contrast CT and CT angiography (CTA) stroke imaging may allow efficient stroke lesion volume measurement. Using surrogate images from CTP, we simulated the feasibility of using multiphase CTA (mCTA) to generate perfusion maps and assess target mismatch profiles. MATERIALS AND METHODS: Patients with acute ischemic stroke who received admission CTP were included in this study. Four CTP images (surrogate mCTA, one pre-contrast and three post-contrast, starting at the arterial peak then at 8 s intervals) were selected according to the CTP arterial time-density curve to simulate non-contrast CT and mCTA images. Cerebral blood flow (CBF) and Tmax maps were calculated using the same model-based deconvolution algorithm for the standard CTP and surrogate mCTA studies. Infarct and penumbra were delineated with CBF < 20% and Tmax > 6 s threshold, respectively. Classification accuracy of surrogate mCTA target mismatch (infarct <70 ml; penumbra ≥15 ml; mismatch ratio ≥1.8) with respect to standard CTP was assessed. Agreement between infarct and penumbra volumes from standard CTP and surrogate mCTA maps were evaluated by Bland-Altman analysis. RESULTS: Of 34 included patients, 28 had target mismatch and 6 did not by standard CTP. Accuracy of classifying target mismatch profiles with surrogate mCTA was 79% with respect to that from standard CTP. Mean  ±  standard deviation of differences (standard CTP minus surrogate mCTA) of infarct and penumbra volumes were 9.8 ± 14.8 ml and 20.1 ± 45.4 ml, respectively. CONCLUSIONS: Surrogate mCTA ischemic lesion volumes agreed with those from standard CTP and may be an efficient alternative when CTP is not practical.

Low-dose CT Perfusion with Sparse-view Filtered Back Projection in Acute Ischemic Stroke.

RATIONALE AND OBJECTIVES: Radiation dose associated with computed tomography (CT) perfusion (CTP) may discourage its use despite its added diagnostic benefit in quantifying ischemic lesion volume. Sparse-view CT reduces scan dose by acquiring fewer X-ray projections per gantry rotation but is contaminated by streaking artifacts using filtered back projection (FBP). We investigated the achievable dose reduction by sparse-view CTP with FBP without affecting CTP lesion volume estimations. MATERIALS AND METHODS: Thirty-eight consecutive patients with acute ischemic stroke and CTP were included in this simulation study. CTP projection data was simulated by forward projecting original reconstructions with 984 views and adding Gaussian noise. Full-view (984 views) and sparse-view (492, 328, 246, and 164 views) CTP studies were simulated by FBP of simulated projection data. Cerebral blood flow (CBF) and time-to-maximum of the impulse residue function (Tmax) maps were generated by deconvolution for each simulated CTP study. Ischemic volumes were measured by CBF<30% relative to the contralateral hemisphere and Tmax > 6 s. Volume accuracy was evaluated with respect to the full-view CTP study by the Friedman test with post hoc multiplicity-adjusted pairwise tests and Bland-Altman analysis. RESULTS: Friedman and multiplicity-adjusted pairwise tests indicated that 164-view CBF < 30%, 246- and 164-view Tmax > 6 s volumes were significantly different to full-view volumes (p < 0.001). Mean difference ± standard deviation (sparse minus full-view lesion volume) ranged from -1.0 ± 2.8 ml to -4.1 ± 11.7 ml for CBF < 30% and -2.9 ± 3.8 ml to -12.5 ± 19.9 ml for Tmax > 6 s from 492 to 164 views, respectively. CONCLUSION: By ischemic volume accuracy, our study indicates that sparse-view CTP may allow dose reduction by up to a factor of 3.

Short-duration dynamic [18F]DCFPyL PET and CT perfusion imaging to localize dominant intraprostatic lesions in prostate cancer: validation against digital histopathology and comparison to [18F]DCFPyL PET/MR at 120 minutes.

PURPOSE: Localized prostate cancer (PCa) in patients is characterized by a dominant focus in the gland (dominant intraprostatic lesion, DIL). Accurate DIL identification may enable more accurate diagnosis and therapy through more precise targeting of biopsy, radiotherapy and focal ablative therapies. The goal of this study is to validate the performance of [18F]DCFPyL PET and CT perfusion (CTP) for detecting and localizing DIL against digital histopathological images. METHODS: Multi-modality image sets: in vivo T2-weighted (T2w)-MRI, 22-min dynamic [18F]DCFPyL PET/CT, CTP, and 2-h post-injection PET/MR were acquired in patients prior to radical prostatectomy. The explanted gland with implanted fiducial markers was imaged with T2w-MRI. All images were co-registered to the pathologist-annotated digital images of whole-mount mid-gland histology sections using fiducial markers and anatomical landmarks. Regions of interest encompassing DIL and non-DIL tissue were drawn on the digital histopathological images and superimposed on PET and CTP parametric maps. Logistic regression with backward elimination of parameters was used to select the most sensitive parameter set to distinguish DIL from non-DIL voxels. Leave-one-patient-out cross-validation was performed to determine diagnostic performance. RESULTS: [18F]DCFPyL PET and CTP parametric maps of 15 patients were analyzed. SUVLate and a model combining Ki and k4 of [18F]DCFPyL achieved the most accurate performance distinguishing DIL from non-DIL voxels. Both detection models achieved an AUC of 0.90 and an error rate of < 10%. Compared to digital histopathology, the detected DILs had a mean dice similarity coefficient of 0.8 for the Ki and k4 model and 0.7 for SUVLate. CONCLUSIONS: We have validated using co-registered digital histopathological images that parameters from kinetic analysis of 22-min dynamic [18F]DCFPyL PET can accurately localize DILs in PCa for targeting of biopsy, radiotherapy, and focal ablative therapies. Short-duration dynamic [18F]DCFPyL PET was not inferior to SUVLate in this diagnostic task. CLINICAL TRIAL REGISTRATION NUMBER: NCT04009174 (ClinicalTrials.gov).

Semi-automatic measurement of intracranial hemorrhage growth on non-contrast CT.

BACKGROUND: Manual segmentations of intracranial hemorrhage on non-contrast CT images are the gold-standard in measuring hematoma growth but are prone to rater variability. AIMS: We demonstrate that a convex optimization-based interactive segmentation approach can accurately and reliably measure intracranial hemorrhage growth. METHODS: Baseline and 16-h follow-up head non-contrast CT images of 46 subjects presenting with intracranial hemorrhage were selected randomly from the ANNEXA-4 trial imaging database. Three users semi-automatically segmented intracranial hemorrhage to measure hematoma volume for each timepoint using our proposed method. Segmentation accuracy was quantitatively evaluated compared to manual segmentations by using Dice similarity coefficient, Pearson correlation, and Bland-Altman analysis. Intra- and inter-rater reliability of the Dice similarity coefficient and intracranial hemorrhage volumes and volume change were assessed by the intraclass correlation coefficient and minimum detectable change. RESULTS: Among the three users, the mean Dice similarity coefficient, Pearson correlation, and mean difference ranged from 76.79% to 79.76%, 0.970 to 0.980 (p < 0.001), and -1.5 to -0.4 ml, respectively, for all intracranial hemorrhage segmentations. Inter-rater intraclass correlation coefficients between the three users for Dice similarity coefficient and intracranial hemorrhage volume were 0.846 and 0.962, respectively, and the corresponding minimum detectable change was 2.51 ml. Inter-rater intraclass correlation coefficient for intracranial hemorrhage volume change ranged from 0.915 to 0.958 for each user compared to manual measurements, resulting in an minimum detectable change range of 2.14 to 4.26 ml. CONCLUSIONS: We spatially and volumetrically validate a novel interactive segmentation method for delineating intracranial hemorrhage on head non-contrast CT images. Good spatial overlap, excellent volume correlation, and good repeatability suggest its usefulness for measuring intracranial hemorrhage volume and volume change on non-contrast CT images.

Dual-domain cascade of U-nets for multi-channel magnetic resonance image reconstruction.

The U-net is a deep-learning network model that has been used to solve a number of inverse problems. In this work, the concatenation of two-element U-nets, termed the W-net, operating in k-space (K) and image (I) domains, were evaluated for multi-channel magnetic resonance (MR) image reconstruction. The two-element network combinations were evaluated for the four possible image-k-space domain configurations: a) W-net II, b) W-net KK, c) W-net IK, and d) W-net KI. Selected four element (WW-nets) and six element (WWW-nets) networks were also examined. Two configurations of each network were compared: 1) each coil channel was processed independently, and 2) all channels were processed simultaneously. One hundred and eleven volumetric, T1-weighted, 12-channel coil k-space datasets were used in the experiments. Normalized root mean squared error, peak signal-to-noise ratio and visual information fidelity were used to assess the reconstructed images against the fully sampled reference images. Our results indicated that networks that operate solely in the image domain were better when independently processing individual channels of multi-channel data. Dual-domain methods were better when simultaneously reconstructing all channels of multi-channel data. In addition, the best cascade of U-nets performed better (p < 0.01) than the previously published, state-of-the-art Deep Cascade and Hybrid Cascade models in three out of four experiments.