Timothy Jameson Allen

Mr. Timothy Jameson Allen

orcid

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
18 October 2023 in Tomography
Verified
Journal Article
Open access
PDF file
TATimothy Jameson Allen
Timothy Jameson Allen
LBLeah C Henze Bancroft
Leah C Henze Bancroft
OUOrhan Unal
Orhan Unal
LELloyd D. Estkowski
Lloyd D. Estkowski
TCTy A. Cashen
Ty A. Cashen
JHJames H Holmes
James H Holmes
Deep learning (DL) reconstruction techniques to improve MR image quality are becoming commercially available with the hope that they will be applicable to multiple imaging application sites and acquisition protocols. However, before clinical implementation, these methods must be validated for specific use cases. In this work, the quality of standard-of-care (SOC) T2w and a high-spatial-resolution (HR) imaging of the breast were assessed both with and without prototype DL reconstruction. Studies were performed using data collected from phantoms, 20 retrospectively collected SOC patient exams, and 56 prospectively acquired SOC and HR patient exams. Image quality was quantitatively assessed via signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and edge sharpness. Qualitatively, all in vivo images were scored by either two or four radiologist readers using 5-point Likert scales in the following categories: artifacts, perceived sharpness, perceived SNR, and overall quality. Differences in reader scores were tested for significance. Reader preference and perception of signal intensity changes were also assessed. Application of the DL resulted in higher average SNR (1.2–2.8 times), CNR (1.0–1.8 times), and image sharpness (1.2–1.7 times). Qualitatively, the SOC acquisition with DL resulted in significantly improved image quality scores in all categories compared to non-DL images. HR acquisition with DL significantly increased SNR, sharpness, and overall quality compared to both the non-DL SOC and the non-DL HR images. The acquisition time for the HR data only required a 20% increase compared to the SOC acquisition and readers typically preferred DL images over non-DL counterparts. Overall, the DL reconstruction demonstrated improved T2w image quality in clinical breast MRI.

9 August 2024 in Magnetic Resonance in Medicine
Verified
Journal Article
Open access
GSGregory Simchick
Gregory Simchick
TATimothy Jameson Allen
Timothy Jameson Allen
DHDiego Hernando
Diego Hernando
Purpose: To evaluate reproducibility and interlobar agreement of intravoxel incoherent motion (IVIM) quantification in the liver across field strengths and MR scanners with different gradient hardware.Methods: Cramer‐Rao lower bound optimization was performed to determine optimized monopolar and motion‐robust 2D (b‐value and first‐order motion moment [M1]) IVIM‐DWI acquisitions. Eleven healthy volunteers underwent diffusion MRI of the liver, where each optimized acquisition was obtained five times across three MRI scanners. For each data set, IVIM estimates (diffusion coefficient (D), pseudo‐diffusion coefficients ( and ), blood velocity SDs (Vb1 and Vb2), and perfusion fractions [f1 and f2]) were obtained in the right and left liver lobes using two signal models (pseudo‐diffusion and M1‐dependent physical) with and without T2 correction (fc1 and fc2) and three fitting techniques (tri‐exponential region of interest–based full and segmented fitting and blood velocity SD distribution fitting). Reproducibility and interlobar agreement were compared across methods using within‐subject and pairwise coefficients of variation (CVw and CVp), paired sample t‐tests, and Bland–Altman analysis.Results: Using a combination of motion‐robust 2D (b‐M1) data acquisition, M1‐dependent physical signal modeling with T2 correction, and blood velocity SD distribution fitting, multiscanner reproducibility with median CVw = 5.09%, 11.3%, 9.20%, 14.2%, and 12.6% for D, Vb1, Vb2, fc1, and fc2, respectively, and interlobar agreement with CVp = 8.14%, 11.9%, 8.50%, 49.9%, and 42.0%, respectively, was achieved.Conclusion: Recently proposed advanced IVIM acquisition, signal modeling, and fitting techniques may facilitate reproducible IVIM quantification in the liver, as needed for establishment of IVIM‐based quantitative biomarkers for detection, staging, and treatment monitoring of diseases.

11 November 2024 in Magnetic Resonance in Medicine
Verified
Journal Article
Open access
TATimothy Jameson Allen
Timothy Jameson Allen
RHRianne A. van der Heijden
Rianne A. van der Heijden
GSGregory Simchick
Gregory Simchick
DHDiego Hernando
Diego Hernando
Purpose: Cardiac‐induced liver motion can bias liver ADC measurements and compromise reproducibility. The purpose of this work was to enable motion‐robust DWI on multiple MR scanners and assess reproducibility of the resulting liver ADC measurements. Methods: First moment‐optimized diffusion imaging (MODI) was implemented on three MR scanners with various gradient performances and field strengths. MODI‐DWI and conventional Stejskal‐Tanner monopolar (MONO) DWI were acquired in eight ( N = 8) healthy volunteers on each scanner, and DWI repetitions were combined using three different averaging methods. For each combination of scanner, acquisition, and averaging method, ADC measurements from each liver segment were collected. Systematic differences in ADC values between scanners and methods were assessed with linear mixed effects modeling, and reproducibility was quantified via reproducibility coefficients. Results: MODI reduced left–right liver lobe ADC bias from 0.43 × 10 −3 mm 2 /s (MONO) to 0.19 × 10 −3 mm 2 /s (MODI) when simple (unweighted) repetition averaging was used. The bias was reduced from 0.23 × 10 −3 mm 2 /s to 0.06 × 10 −3 mm 2 /s using weighted averaging, and 0.14 × 10 −3 mm 2 /s to 0.01 × 10 −3 mm 2 /s using squared weighted averaging. There was no significant difference in ADC measurements between field strengths or scanner gradient performance. MODI improved reproducibility coefficients compared to MONO: 0.84 × 10 −3 mm 2 /s vs. 0.63 × 10 −3 mm 2 /s (MODI vs. MONO) for simple averaging, 0.66 × 10 −3 mm 2 /s vs. 0.50 × 10 −3 mm 2 /s for weighted averaging, and 0.61 × 10 −3 mm 2 /s vs. 0.47 × 10 −3 mm 2 /s for squared weighted averaging. Conclusion: The feasibility of motion‐robust liver DWI using MODI was demonstrated on multiple MR scanners. MODI improved interlobar agreement and reproducibility of ADC measurements in a healthy cohort.

10 May 2023 in Tomography
Verified
Journal Article
Open access
PDF file
TATimothy Jameson Allen
Timothy Jameson Allen
LBLeah C Henze Bancroft
Leah C Henze Bancroft
KWKang Wang
Kang Wang
PWPing Ni Wang
Ping Ni Wang
OUOrhan Unal
Orhan Unal
JHJames H Holmes
James H Holmes
Graphically prescribed patient-specific imaging volumes and local pre-scan volumes are routinely placed by MRI technologists to optimize image quality. However, manual placement of these volumes by MR technologists is time-consuming, tedious, and subject to intra- and inter-operator variability. Resolving these bottlenecks is critical with the rise in abbreviated breast MRI exams for screening purposes. This work proposes an automated approach for the placement of scan and pre-scan volumes for breast MRI. Anatomic 3-plane scout image series and associated scan volumes were retrospectively collected from 333 clinical breast exams acquired on 10 individual MRI scanners. Bilateral pre-scan volumes were also generated and reviewed in consensus by three MR physicists. A deep convolutional neural network was trained to predict both the scan and pre-scan volumes from the 3-plane scout images. The agreement between the network-predicted volumes and the clinical scan volumes or physicist-placed pre-scan volumes was evaluated using the intersection over union, the absolute distance between volume centers, and the difference in volume sizes. The scan volume model achieved a median 3D intersection over union of 0.69. The median error in scan volume location was 2.7 cm and the median size error was 2%. The median 3D intersection over union for the pre-scan placement was 0.68 with no significant difference in mean value between the left and right pre-scan volumes. The median error in the pre-scan volume location was 1.3 cm and the median size error was −2%. The average estimated uncertainty in positioning or volume size for both models ranged from 0.2 to 3.4 cm. Overall, this work demonstrates the feasibility of an automated approach for the placement of scan and pre-scan volumes based on a neural network model.