A5076562643- Advanced Neuroimaging Techniques and Applications
- Fetal and Pediatric Neurological Disorders
- Medical Imaging Techniques and Applications
- Neonatal and fetal brain pathology
- AI in cancer detection
- Advanced MRI Techniques and Applications
- Advanced X-ray and CT Imaging
- Agriculture and Farm Safety
- MRI in cancer diagnosis
- Image and Signal Denoising Methods
- Medical Image Segmentation Techniques
- Radiomics and Machine Learning in Medical Imaging
- Advanced Neural Network Applications
- Domain Adaptation and Few-Shot Learning
- Soil Mechanics and Vehicle Dynamics
- Vehicle Dynamics and Control Systems
- Bone and Joint Diseases
- Prostate Cancer Diagnosis and Treatment
- Medical Imaging and Analysis
- Advanced Image Processing Techniques
- Machine Learning and Data Classification
- Sparse and Compressive Sensing Techniques
- COVID-19 diagnosis using AI
- Remote Sensing in Agriculture
- Leaf Properties and Growth Measurement
Harvard University
2019-2025
Boston Children's Hospital
2019-2025
Boston Children's Museum
2024
University of British Columbia
2015-2021
Boston University
2021
Harvard University Press
2020
Iranian Research Organization for Science and Technology
2019
University of Manitoba
2008-2012
Florida Department of Citrus
2011
University of Florida
2011
The Hausdorff Distance (HD) is widely used in evaluating medical image segmentation methods. However, the existing methods do not attempt to reduce HD directly. In this paper, we present novel loss functions for training convolutional neural network (CNN)-based with goal of reducing We propose three estimate from probability map produced by a CNN. One method makes use distance transform boundary. Another based on applying morphological erosion difference between true and estimated maps....
Visual inspection of histopathology images stained biopsy tissue by expert pathologists is the standard method for grading prostate cancer (PCa). However, this process time-consuming and subject to high inter-observer variability. Machine learning-based methods have potential improve efficient throughput large volumes slides while decreasing variability, but they are not easy develop because require substantial amounts labeled training data. In paper, we propose a deep classification...
<h3>Importance</h3> Proper evaluation of the performance artificial intelligence techniques in analysis digitized medical images is paramount for adoption such by community and regulatory agencies. <h3>Objectives</h3> To compare several cross-validation (CV) approaches to evaluate a classifier automatic grading prostate cancer histopathologic when trained using data from 1 expert multiple experts. <h3>Design, Setting, Participants</h3> This quality improvement study used tissue microarray...
ABSTRACT There is a growing interest in using diffusion MRI to study the white matter tracts and structural connectivity of fetal brain. Recent progress data acquisition processing suggests that this imaging modality has unique role elucidating normal abnormal patterns neurodevelopment utero. However, there have been no efforts quantify prevalence crossing bottleneck regions, important issues investigated for adult brains. In work, we determined brain regions with bottlenecks between 23 36...
Limitations in the accuracy of brain pathways reconstructed by diffusion MRI (dMRI) tractography have received considerable attention. While technical advances spearheaded Human Connectome Project (HCP) led to significant improvements dMRI data quality, it remains unclear how these should be analyzed maximize accuracy. Over a period two years, we engaged community IronTract Challenge, which aims answer this question leveraging unique dataset. Macaque brains that both tracer injections and ex...
Diffusion-weighted Magnetic Resonance Imaging (dMRI) is increasingly used to study the fetal brain in utero. An important computation enabled by dMRI streamline tractography, which has unique applications such as tract-specific analysis of white matter and structural connectivity assessment. However, due low data quality challenging nature existing methods tend produce highly inaccurate results. They generate many false streamlines while failing reconstruct that constitute major tracts. In...
Diffusion-weighted magnetic resonance imaging (DW-MRI) of fetal brain is challenged by frequent motion and signal to noise ratio that much lower than non-fetal imaging. As a result, accurate robust parameter estimation in DW-MRI remains an open problem. Recently, deep learning techniques have been successfully used for subjects. However, none those prior works has addressed the because obtaining reliable training data challenging. To address this problem, work we propose novel methodology...
PURPOSE Prostate cancer (PCa) represents a highly heterogeneous disease that requires tools to assess oncologic risk and guide patient management treatment planning. Current models are based on various clinical pathologic parameters including Gleason grading, which suffers from high interobserver variability. In this study, we determine whether objective machine learning (ML)–driven histopathology image analysis would aid us in better stratification of PCa. MATERIALS AND METHODS We propose...
Abstract Diffusion-weighted magnetic resonance imaging (dMRI) of the brain offers unique capabilities including noninvasive probing tissue microstructure and structural connectivity. It is widely used for clinical assessment disease injury, neuroscience research. Analyzing dMRI data to extract useful information medical scientific purposes can be challenging. The measurements may suffer from strong noise artifacts, exhibit high intersession interscanner variability in data, as well...
Abstract Diffusion magnetic resonance imaging (dMRI) is pivotal for probing the microstructure of rapidly-developing fetal brain. However, motion during scans and its interaction with field inhomogeneities result in artifacts data scattering across spatial angular domains. The effects those are more pronounced high-angular resolution dMRI, where signal-to-noise ratio very low. Those lead to biased estimates compromise consistency reliability dMRI analysis. This work presents HAITCH, first...
Diffusion MRI (dMRI) provides unique insights into fetal brain microstructure in utero. Longitudinal and cross-sectional dMRI studies can reveal crucial neurodevelopmental changes but require precise spatial alignment across scans subjects. This is challenging due to low data quality, rapid development, limited anatomical landmarks. Existing registration methods, designed for high-quality adult data, struggle with these complexities. To address this, we introduce FetDTIAlign, a deep learning...
ABSTRACT During the second and third trimesters of human gestation, brain undergoes rapid neurodevelopment thanks to critical processes such as neuronal migration, radial glial scaffolding, synaptic sprouting. Unfortunately, gathering high‐quality MRI data on healthy fetal is complex, making it challenging understand this development. To address issue, we conducted a study using motion‐corrected diffusion tensor imaging (DTI) analyze changes in cortical gray matter (CP) sub‐cortical white...
Fetal brain imaging is essential for prenatal care, with ultrasound (US) and magnetic resonance (MRI) providing complementary strengths. While MRI has superior soft tissue contrast, US offers portable inexpensive screening of neurological abnormalities. Despite the great potential synergy combined fetal MR to enhance diagnostic accuracy, little effort been made integrate these modalities. An step towards this integration accurate automatic spatial alignment, which technically very...
Citrus greening, also known as Huanglongbing or HLB, is a major threat to the U.S. citrus industry. Currently, scouting and visual inspection are used for screening infected trees. However, this time-consuming expensive method HLB disease detection. Moreover, it subjective, current exhibits high detection error rates. The objective of study was investigate potential visible near-infrared (VIS-NIR) spectroscopy identifying HLB-infected spectral data from healthy orange trees Valencia variety...
Abstract This work presents detailed anatomic labels for a spatiotemporal atlas of fetal brain Diffusion Tensor Imaging (DTI) between 23 and 30 weeks post‐conceptional age. Additionally, we examined developmental trajectories in fractional anisotropy (FA) mean diffusivity (MD) across gestational ages (GA). We performed manual segmentations on DTI atlas. labeled 14 regions interest (ROIs): cortical plate (CP), subplate (SP), Intermediate zone‐subventricular zone‐ventricular zone (IZ/SVZ/VZ),...