A5068099928- Radiomics and Machine Learning in Medical Imaging
- AI in cancer detection
- Lung Cancer Diagnosis and Treatment
- Medical Imaging Techniques and Applications
- Ultrasound Imaging and Elastography
- Multiple Sclerosis Research Studies
- Cancer Genomics and Diagnostics
- Brain Tumor Detection and Classification
- Lung Cancer Treatments and Mutations
- Advanced X-ray and CT Imaging
- Intracerebral and Subarachnoid Hemorrhage Research
- Image Processing Techniques and Applications
- COVID-19 diagnosis using AI
- Cancer Immunotherapy and Biomarkers
- Medical Image Segmentation Techniques
- Machine Learning in Healthcare
- CAR-T cell therapy research
- Influenza Virus Research Studies
- Digital Imaging for Blood Diseases
- Ultrasound and Hyperthermia Applications
- Genetic factors in colorectal cancer
- COVID-19 and healthcare impacts
- Cell Image Analysis Techniques
- Fetal and Pediatric Neurological Disorders
The University of Texas MD Anderson Cancer Center
2022-2024
The University of Texas Health Science Center at Houston
2018-2022
Icahn School of Medicine at Mount Sinai
2019
Only around 20-30% of patients with non-small-cell lung cancer (NCSLC) have durable benefit from immune-checkpoint inhibitors. Although tissue-based biomarkers (eg, PD-L1) are limited by suboptimal performance, tissue availability, and tumour heterogeneity, radiographic images might holistically capture the underlying biology. We aimed to investigate application deep learning on chest CT scans derive an imaging signature response immune checkpoint inhibitors evaluate its added value in...
Abstract While we recognize the prognostic importance of clinicopathological measures and circulating tumor DNA (ctDNA), independent contribution quantitative image markers to prognosis in non-small cell lung cancer (NSCLC) remains underexplored. In our multi-institutional study 394 NSCLC patients, utilize pre-treatment computed tomography (CT) 18 F-fluorodeoxyglucose positron emission (FDG-PET) establish a habitat imaging framework for assessing regional heterogeneity within individual...
Background Enhancing lesions on MRI scans obtained after contrast material administration are commonly thought to represent disease activity in multiple sclerosis (MS); it is desirable develop methods that can predict enhancing without the use of material. Purpose To evaluate whether deep learning Materials and Methods This study involved prospective analysis existing data. A convolutional neural network was used for classification unenhanced scans. performed each slice, slice scores were...
BackgroundPredicting relapse and overall survival in early-stage non-small cell lung cancer (NSCLC) patients remains challenging. Therefore, we hypothesized that detection of circulating tumor DNA (ctDNA) can identify with increased risk integrating radiological volume measurement along ctDNA detectability improves prediction outcome.Patients MethodWe analyzed 366 serial plasma samples from 85 who underwent surgical resections assessed using a next-generation sequencing liquid biopsy assay,...
Objective: To investigate the performance of deep learning (DL) based on fully convolutional neural network (FCNN) in segmenting brain tissues a large cohort multiple sclerosis (MS) patients. Methods: We developed FCNN model to segment tissues, including T2-hyperintense MS lesions. The training, validation, and testing were ~1000 magnetic resonance imaging (MRI) datasets acquired relapsing–remitting patients, as part phase 3 randomized clinical trial. Multimodal MRI data (dual-echo, FLAIR,...
Background Deep learning (DL) is a promising methodology for automatic detection of abnormalities in brain MRI. Purpose To automatically evaluate the quality multicenter structural MRI images using an ensemble DL model based on deep convolutional neural networks (DCNNs). Study Type Retrospective. Population The study included 1064 autism patients and healthy controls from Autism Brain Imaging Data Exchange (ABIDE) database. data 110 multiple sclerosis CombiRx were independent testing....
[18F]Fluorodeoxyglucose positron emission tomography (FDG-PET) and computed (CT) are indispensable components in modern medicine. Although PET can provide additional diagnostic value, it is costly not universally accessible, particularly low-income countries. To bridge this gap, we have developed a conditional generative adversarial network pipeline that produce FDG-PET from CT scans based on multi-center multi-modal lung cancer datasets (n = 1,478). Synthetic images validated across...
Background The dependence of deep‐learning (DL)‐based segmentation accuracy brain MRI on the training size is not known. Purpose To determine required for a desired in multiple sclerosis (MS) using DL. Study Type Retrospective analysis data acquired as part multicenter clinical trial. Population In all, 1008 patients with clinically definite MS. Field Strength/Sequence MRIs were at 1.5T and 3T scanners manufactured by GE, Philips, Siemens dual turbo spin echo, FLAIR, T 1 ‐weighted echo...
<title>Abstract</title> A panzootic of H5Nx avian influenza viruses has severely affected poultry and wild bird populations, resulting in multiple mammalian spillovers, including mild to fatal human infections caused by antigenically distinct diverse virus clades. The unpredictable nature spillover events drifted hinders our ability effectively respond the heightened risk human-to-human transmission. Stockpiled strain-specific H5 whole virus-based vaccines provide limited breadth reduced...
Automated evaluation of image quality is essential to assure accurate diagnosis and effective patient management. This particularly important for multi-center studies, typically employed in clinical trials, which the data are acquired on different machines with protocols. Visual assessment magnetic resonance imaging (MRI) subjective impractical large datasets. Data-intensive deep learning methods such as convolutional neural networks (CNNs) promising tools processing large-scale datasets...
Multiple sclerosis (MS) is a demyelinating disease that affects the central nervous system (CNS) and characterized by presence of CNS lesions. Volumetric measures tissues, including lesions, on magnetic resonance imaging (MRI) play key roles in clinical management treatment evaluation MS patient. Recent advances deep learning (DL) show promising results for automated medical image segmentation. In this work, we used convolutional neural networks (CNNs) brain tissue classification MRI...
Cancer patients have worse outcomes from the COVID-19 infection and greater need for ventilator support elevated mortality rates than general population. However, previous artificial intelligence (AI) studies focused on without cancer to develop diagnosis severity prediction models. Little is known about how AI models perform in patients. In this study, we aim a computational framework particularly population further compare it head-to-head
<h3>Background</h3> Adoptive cell therapy (ACT) is a newly emerging treatment modality for solid tumors with heterogenous outcomes and limited biomarkers to guide patient selection. The relationship between tumor burden ACT efficacy poorly understood. We sought investigate whether CT-based volumetric analysis helps predict after the aim of improving <h3>Methods</h3> This retrospective study included consecutive patients advanced who received on protocol in Department Investigational Cancer...
<h3>Background</h3> Immune checkpoint inhibitors (ICIs) are a cornerstone of modern oncological treatments, particularly in the management various cancers through immunotherapy. Despite their clinical success, ICIs often associated with several immune-related adverse events (irAEs), among which pneumonitis is significant due to its potential severity. Accurately differentiating ICI-induced from normal pulmonary findings and other interstitial lung abnormalities (ILAs) at an early stage...
<h3>Background</h3> Immune checkpoint inhibitors (ICIs) have revolutionized the treatment of non-small cell lung cancer (NSCLC), benefiting 20–30% patients. The current clinical standard for initiating ICI therapy is assessment Programmed Death-Ligand 1 (PD-L1) status via immunohistochemistry (IHC) on biopsy specimens. However, this invasive procedure presents risks and limitations, highlighting need a non-invasive alternative. This issue critical as it affects patient outcomes...
Histopathologic whole-slide images (WSI) are generally considered the gold standard for cancer diagnosis and prognosis. Survival prediction based on WSI has recently attracted substantial attention. Nevertheless, it remains a central challenge owing to inherent difficulties of predicting patient prognosis effectively extracting informative survival-specific representations from with highly compounded gigapixels. In this study, we present fully automated cellular-level dual global fusion...
Abstract Introduction: Atypical adenomatous hyperplasia (AAH) is the only recognized preneoplasia of lung adenocarcinoma, which can progress to adenocarcinoma in situ (AIS), minimally invasive (MIA) and eventually (ADC). A more complete understanding early carcinogenesis cancer critical for detection interception. However, studying these precursors challenging because lesions are often insufficient molecular immune profiling. Artificial intelligence (AI)-based studies on H&E...
Background: Prior observational studies suggest lower in-hospital mortality (IHM) among intracerebral hemorrhage (ICH) patients treated at higher levels of care (LOC). However, comparisons across non-certified (NC), primary certified (PSC), and comprehensive centers (CSC) have not been reported; particularly accounting for changing certification status. Also, unmeasured confounding in designs can bias estimates. We performed an instrumental variable (IV) analysis population level TX state...