Andrew Schmidt

Andrew Schmidt

Auto-generated profile, based on publicly available information

Andrew Schmidt

Centre for Hip Health and Mobility

Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

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

Representative Publications

View All Publications
15 October 2019 in Lab on a Chip
Verified
Journal Article
Closed access
PDF file
IAIsmail Emre Araci
Ismail Emre Araci
SASevda Agaoglu
Sevda Agaoglu
JLJu Young Lee
Ju Young Lee
LYLaura Rivas Yepes
Laura Rivas Yepes
PDPriscilla Diep
Priscilla Diep
ASAndrew Schmidt
Andrew Schmidt
Dilatometric strain sensors (DSS) that work based on detection of volume change in microfluidic channels; i) are highly sensitive to biaxial strain, ii) can be fabricated using only soft and transparent materials, and iii) are easy to integrate with smart-phones. These features are especially attractive for contact lens based intraocular pressure (IOP) sensing applications. The inherent flow stabilization of the microfluidic systems is an additional advantage suitable for filtering out rapid fluctuations. Here, we have demonstrated that the low-pass filtering in microfluidic sensors improves the signal-to-noise-ratio for ophthalmic applications. We have fabricated devices with a time constant in the range of 1–200 seconds. We have demonstrated that the device architecture and working liquid viscosity (10–866 cSt) are the two independent factors that determine the sensor time constant. We have developed an equivalent circuit model for the DSS that accurately represents the experimental results thus can be used as a computational model for design and development of microfluidic sensors. For a sensor with the time constant of 4 s, we report that microfluidic signal filtering in IOP monitoring applications can suppress the rapid fluctuations (i.e., the noise due to ocular pulsation, blinking etc.) by 9 dB without the need for electronic components.

19 July 2022 in Journal of Magnetic Resonance Imaging
Verified
Journal Article
Closed access
ASAndrew M. Schmidt
Andrew M. Schmidt
ADArjun D. Desai
Arjun D. Desai
LWLauren E. Watkins
Lauren E. Watkins
HCHollis A. Crowder
Hollis A. Crowder
MBMarianne S. Black
Marianne S. Black
ACAkshay Chaudhari
Akshay Chaudhari

4 February 2023 in Bioengineering
Verified
Journal Article
Open access
PDF file
JDJeffrey Dominic
Jeffrey Dominic
NBNandita Bhaskhar
Nandita Bhaskhar
ADArjun D. Desai
Arjun D. Desai
ASAndrew Schmidt
Andrew Schmidt
ERElka Rubin
Elka Rubin
ACAkshay Chaudhari
Akshay Chaudhari
We systematically evaluate the training methodology and efficacy of two inpainting-based pretext tasks of context prediction and context restoration for medical image segmentation using self-supervised learning (SSL). Multiple versions of self-supervised U-Net models were trained to segment MRI and CT datasets, each using a different combination of design choices and pretext tasks to determine the effect of these design choices on segmentation performance. The optimal design choices were used to train SSL models that were then compared with baseline supervised models for computing clinically-relevant metrics in label-limited scenarios. We observed that SSL pretraining with context restoration using 32 × 32 patches and Poission-disc sampling, transferring only the pretrained encoder weights, and fine-tuning immediately with an initial learning rate of 1 × 10−3 provided the most benefit over supervised learning for MRI and CT tissue segmentation accuracy (p < 0.001). For both datasets and most label-limited scenarios, scaling the size of unlabeled pretraining data resulted in improved segmentation performance. SSL models pretrained with this amount of data outperformed baseline supervised models in the computation of clinically-relevant metrics, especially when the performance of supervised learning was low. Our results demonstrate that SSL pretraining using inpainting-based pretext tasks can help increase the robustness of models in label-limited scenarios and reduce worst-case errors that occur with supervised learning.

1 September 2021 in The Bone & Joint Journal
Verified
Journal Article
Closed access
DSDavid J. Stockton
David J. Stockton
ASAndrew M. Schmidt
Andrew M. Schmidt
AYAndrew Yung
Andrew Yung
JDJane Desrochers
Jane Desrochers
HZHonglin Zhang
Honglin Zhang
DWDavid R. Wilson
David R. Wilson
Aims: Anterior cruciate ligament (ACL) rupture commonly leads to post-traumatic osteoarthritis, regardless of surgical reconstruction. This study uses standing MRI to investigate changes in contact area, contact centroid location, and tibiofemoral alignment between ACL-injured knees and healthy controls, to examine the effect of ACL reconstruction on these parameters. Methods: An upright, open MRI was used to directly measure tibiofemoral contact area, centroid location, and alignment in 18 individuals with unilateral ACL rupture within the last five years. Eight participants had been treated nonoperatively and ten had ACL reconstruction performed within one year of injury. All participants were high-functioning and had returned to sport or recreational activities. Healthy contralateral knees served as controls. Participants were imaged in a standing posture with knees fully extended. Results: Participants’ mean age was 28.4 years (SD 7.3), the mean time since injury was 2.7 years (SD 1.6), and the mean International Knee Documentation Subjective Knee Form score was 84.4 (SD 13.5). ACL injury was associated with a 10% increase (p = 0.001) in contact area, controlling for compartment, sex, posture, age, body mass, and time since injury. ACL injury was associated with a 5.2% more posteriorly translated medial centroid (p = 0.001), equivalent to a 2.6 mm posterior translation on a representative tibia with mean posteroanterior width of 49.4 mm. Relative to the femur, the tibiae of ACL ruptured knees were 2.3 mm more anteriorly translated (p = 0.003) and 2.6° less externally rotated (p = 0.010) than healthy controls. ACL reconstruction was not associated with an improvement in any measure. Conclusion: ACL rupture was associated with an increased contact area, posteriorly translated medial centroid, anterior tibial translation, and reduced tibial external rotation in full extension. These changes were present 2.7 years post-injury regardless of ACL reconstruction status. Cite this article: Bone Joint J 2021;103-B(9):1505–1513.

18 October 2023 in Journal of Magnetic Resonance Imaging
Verified
Journal Article
Closed access
ERElka B. Rubin
Elka B. Rubin
ASAndrew M. Schmidt
Andrew M. Schmidt
MKMatthew F. Koff
Matthew F. Koff
FKFeliks Kogan
Feliks Kogan
KGKenneth Gao
Kenneth Gao
GGGarry E. Gold
Garry E. Gold

14 March 2022 in arXiv
Verified
Preprint
Open access
PDF file
ADArjun D Desai
Arjun D Desai
ASAndrew M Schmidt
Andrew M Schmidt
ERElka B Rubin
Elka B Rubin
CSChristopher M Sandino
Christopher M Sandino
MBMarianne S Black
Marianne S Black
ACAkshay Chaudhari
Akshay Chaudhari
Magnetic resonance imaging (MRI) is a cornerstone of modern medical imaging. However, long image acquisition times, the need for qualitative expert analysis, and the lack of (and difficulty extracting) quantitative indicators that are sensitive to tissue health have curtailed widespread clinical and research studies. While recent machine learning methods for MRI reconstruction and analysis have shown promise for reducing this burden, these techniques are primarily validated with imperfect image quality metrics, which are discordant with clinically-relevant measures that ultimately hamper clinical deployment and clinician trust. To mitigate this challenge, we present the Stanford Knee MRI with Multi-Task Evaluation (SKM-TEA) dataset, a collection of quantitative knee MRI (qMRI) scans that enables end-to-end, clinically-relevant evaluation of MRI reconstruction and analysis tools. This 1.6TB dataset consists of raw-data measurements of ~25,000 slices (155 patients) of anonymized patient MRI scans, the corresponding scanner-generated DICOM images, manual segmentations of four tissues, and bounding box annotations for sixteen clinically relevant pathologies. We provide a framework for using qMRI parameter maps, along with image reconstructions and dense image labels, for measuring the quality of qMRI biomarker estimates extracted from MRI reconstruction, segmentation, and detection techniques. Finally, we use this framework to benchmark state-of-the-art baselines on this dataset. We hope our SKM-TEA dataset and code can enable a broad spectrum of research for modular image reconstruction and image analysis in a clinically informed manner. Dataset access, code, and benchmarks are available at https://github.com/StanfordMIMI/skm-tea.