Avantika Sinha
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Avantika Sinha has a total of 17 co-authors. Here are the top co-authors by number of shared publications:
- Peder Eric Zufall Larson
Department of Radiology and Biomedical Imaging, University of California, San Francisco, CA 94158, USA
2 shared publications - Renuka Sriram
Department of Radiology and Biomedical Imaging, University of California, San Francisco, San Francisco, California
2 shared publications - Sule Sahin
UC Berkeley–UCSF Graduate Program in Bioengineering University of California, Berkeley and University of California, San Francisco California USA
2 shared publications - Abhejit Rajagopal
Department of Radiology and Biomedical Imaging, University of California-San Francisco, San Francisco, CA 94158, USA
1 shared publication - Ernesto Diaz
Department of Radiology and Biomedical Imaging, University of California-San Francisco, San Francisco, CA 94158, USA
1 shared publication - Henry Salkever
Department of Radiology and Biomedical Imaging, University of California-San Francisco, San Francisco, CA 94158, USA
1 shared publication
Latest Publications
- Open Access
- 22 February 2025
- Published by MDPI AG in Tomography
- Vol. 11 (3) , 21
Deep Learning-Based Tumor Segmentation of Murine Magnetic Resonance Images of Prostate Cancer Patient-Derived Xenografts
Background/Objective: Longitudinal in vivo studies of murine xenograft models are widely utilized in oncology to study cancer biology and develop therapies. Magnetic resonance imaging (MRI) of these tumors is an invaluable tool for monitoring tumor growth and characterizing the tumors as well. Methods: In this work, a pipeline for automating the segmentation of xenografts in mouse models was developed. T2-weighted (T2-wt) MRI images from mice implanted with six different prostate cancer patient-derived xenografts (PDX) in the kidneys, liver, and tibia were used. The segmentation pipeline included a slice classifier to identify the slices that had tumors and subsequent training and validation using several U-Net-based segmentation architectures. Multiple combinations of the algorithm and training images for different sites were evaluated for inference quality. Results and Conclusions: The slice classifier network achieved 90% accuracy in identifying slices containing tumors. Among the various segmentation architectures tested, the dense residual recurrent U-Net achieved the highest performance in kidney tumors. When evaluated across the kidneys, tibia, and liver, this architecture performed the best when trained on all data as compared to training on only data from a single site (and inferring on a multi-site tumor images), achieving a Dice score of 0.924 across the test set.
- 22 February 2025
- Published by MDPI AG in Tomography
- Vol. 11 (3) , 21
Open AccessDeep Learning-Based Tumor Segmentation of Murine Magnetic Resonance Images of Prostate Cancer Patient-Derived Xenografts
Background/Objective: Longitudinal in vivo studies of murine xenograft models are widely utilized in oncology to study cancer biology and develop therapies. Magnetic resonance imaging (MRI) of these tumors is an invaluable tool for monitoring tumor growth and characterizing the tumors as well. Methods: In this work, a pipeline for automating the segmentation of xenografts in mouse models was developed. T2-weighted (T2-wt) MRI images from mice implanted with six different prostate cancer patient-derived xenografts (PDX) in the kidneys, liver, and tibia were used. The segmentation pipeline included a slice classifier to identify the slices that had tumors and subsequent training and validation using several U-Net-based segmentation architectures. Multiple combinations of the algorithm and training images for different sites were evaluated for inference quality. Results and Conclusions: The slice classifier network achieved 90% accuracy in identifying slices containing tumors. Among the various segmentation architectures tested, the dense residual recurrent U-Net achieved the highest performance in kidney tumors. When evaluated across the kidneys, tibia, and liver, this architecture performed the best when trained on all data as compared to training on only data from a single site (and inferring on a multi-site tumor images), achieving a Dice score of 0.924 across the test set.
- Open Access
- 27 March 2023
- Published by MDPI AG in Tomography
- Vol. 9 (2) , 736-749
Metabolite-Specific Echo Planar Imaging for Preclinical Studies with Hyperpolarized 13C-Pyruvate MRI
Metabolite-specific echo-planar imaging (EPI) sequences with spectral–spatial (spsp) excitation are commonly used in clinical hyperpolarized [1-13C]pyruvate studies because of their speed, efficiency, and flexibility. In contrast, preclinical systems typically rely on slower spectroscopic methods, such as chemical shift imaging (CSI). In this study, a 2D spspEPI sequence was developed for use on a preclinical 3T Bruker system and tested on in vivo mice experiments with patient-derived xenograft renal cell carcinoma (RCC) or prostate cancer tissues implanted in the kidney or liver. Compared to spspEPI sequences, CSI were found to have a broader point spread function via simulations and exhibited signal bleeding between vasculature and tumors in vivo. Parameters for the spspEPI sequence were optimized using simulations and verified with in vivo data. The expected lactate SNR and pharmacokinetic modeling accuracy increased with lower pyruvate flip angles (less than 15°), intermediate lactate flip angles (25° to 40°), and temporal resolution of 3 s. Overall SNR was also higher with coarser spatial resolution (4 mm isotropic vs. 2 mm isotropic). Pharmacokinetic modelling used to fit kPL maps showed results consistent with the previous literature and across different sequences and tumor xenografts. This work describes and justifies the pulse design and parameter choices for preclinical spspEPI hyperpolarized 13C-pyruvate studies and shows superior image quality to CSI.
- 27 March 2023
- Published by MDPI AG in Tomography
- Vol. 9 (2) , 736-749
Open AccessMetabolite-Specific Echo Planar Imaging for Preclinical Studies with Hyperpolarized 13C-Pyruvate MRI
Metabolite-specific echo-planar imaging (EPI) sequences with spectral–spatial (spsp) excitation are commonly used in clinical hyperpolarized [1-13C]pyruvate studies because of their speed, efficiency, and flexibility. In contrast, preclinical systems typically rely on slower spectroscopic methods, such as chemical shift imaging (CSI). In this study, a 2D spspEPI sequence was developed for use on a preclinical 3T Bruker system and tested on in vivo mice experiments with patient-derived xenograft renal cell carcinoma (RCC) or prostate cancer tissues implanted in the kidney or liver. Compared to spspEPI sequences, CSI were found to have a broader point spread function via simulations and exhibited signal bleeding between vasculature and tumors in vivo. Parameters for the spspEPI sequence were optimized using simulations and verified with in vivo data. The expected lactate SNR and pharmacokinetic modeling accuracy increased with lower pyruvate flip angles (less than 15°), intermediate lactate flip angles (25° to 40°), and temporal resolution of 3 s. Overall SNR was also higher with coarser spatial resolution (4 mm isotropic vs. 2 mm isotropic). Pharmacokinetic modelling used to fit kPL maps showed results consistent with the previous literature and across different sequences and tumor xenografts. This work describes and justifies the pulse design and parameter choices for preclinical spspEPI hyperpolarized 13C-pyruvate studies and shows superior image quality to CSI.