Honors & Awards
Stanford Molecular Imaging Scholar (SMIS Fellow), Stanford (2018-2021)
Student Travel Stipends, World Molecular Imaging Congress (WMIC) (2017)
Outstanding Translational Research Award, Department of Biomedical Engineering at Georgia Institute of Technology and Emory University (2016)
Rising Stars in Biomedical Engineering and Science, Massachusetts Institute of Technology, Cambridge (2016)
Outstanding Research Poster Award (Third Place), Inauguration Workshop and Launch of the Integrative Cancer Imaging Research Initiative, Emory&Gatech (2016)
Coulter Scholarship, The Wallace H. Coulter Foundation (2009-2012)
Excellent Academic Scholarship, Shanghai Jiao Tong University (2009-2012)
BOSCH Academic Scholarship, Shanghai Jiao Tong University (2011)
BlackBerry Academic Scholarship, Shanghai Jiao Tong University (2011)
Outstanding Academic Scholarship, Xidian University (2008)
First Class Academic Scholarship, Xidian University (2005–2007)
National Scholarship of China, Ministry of Education of the P. R. China (2005)
Doctor of Philosophy, Georgia Institute of Technology and Emory University, Biomedical Engineering (2016)
Master of Science, Georgia Institute of Technology, Electrical and Computer Engineering (2012)
Master of Science, Shanghai Jiao Tong University, Precision Instrument and Machinery (2012)
Bachelor of Science, Xidian University, Measurement Control Technology and Instrumentation (2009)
Eben Rosenthal, Postdoctoral Faculty Sponsor
Deep convolutional neural networks for classifying head and neck cancer using hyperspectral imaging.
Journal of biomedical optics
2017; 22 (6): 60503
Surgical cancer resection requires an accurate and timely diagnosis of the cancer margins in order to achieve successful patient remission. Hyperspectral imaging (HSI) has emerged as a useful, noncontact technique for acquiring spectral and optical properties of tissue. A convolutional neural network (CNN) classifier is developed to classify excised, squamous-cell carcinoma, thyroid cancer, and normal head and neck tissue samples using HSI. The CNN classification was validated by the manual annotation of a pathologist specialized in head and neck cancer. The preliminary results of 50 patients indicate the potential of HSI and deep learning for automatic tissue-labeling of surgical specimens of head and neck patients.
View details for DOI 10.1117/1.JBO.22.6.060503
View details for PubMedID 28655055
View details for PubMedCentralID PMC5482930
Detection and delineation of squamous neoplasia with hyperspectral imaging in a mouse model of tongue carcinogenesis.
Journal of biophotonics
yperspectral imaging (HSI) holds the potential for the noninvasive detection of cancers. Oral cancers are often diagnosed at a late stage when treatment is less effective and the mortality and morbidity rates are high. Early detection of oral cancer is, therefore, crucial in order to improve the clinical outcomes. To investigate the potential of HSI as a non-invasive diagnostic tool, an animal study was designed to acquire hyperspectral images of in vivo and ex vivo mouse tongues from a chemically induced tongue carcinogenesis model. A variety of machine-learning algorithms, including discriminant analysis, ensemble learning, and support vector machines, were evaluated for tongue neoplasia detection using HSI and were validated by the reconstructed pathological gold-standard maps. The diagnostic performance of HSI, autofluorescence imaging, and fluorescence imaging were compared in this study. Color-coded prediction maps were generated to display the predicted location and distribution of premalignant and malignant lesions. This study suggests that hyperspectral imaging combined with machine-learning techniques can provide a non-invasive tool for the quantitative detection and delineation of squamous neoplasia. Detection and delineation of tongue neoplasia with hyperspectral imaging validated by the pathological gold standard.
View details for DOI 10.1002/jbio.201700078
View details for PubMedID 28921845
Detection of Head and Neck Cancer in Surgical Specimens Using Quantitative Hyperspectral Imaging.
Clinical cancer research : an official journal of the American Association for Cancer Research
2017; 23 (18): 5426–36
Purpose: This study intends to investigate the feasibility of using hyperspectral imaging (HSI) to detect and delineate cancers in fresh, surgical specimens of patients with head and neck cancers.Experimental Design: A clinical study was conducted in order to collect and image fresh, surgical specimens from patients (N = 36) with head and neck cancers undergoing surgical resection. A set of machine-learning tools were developed to quantify hyperspectral images of the resected tissue in order to detect and delineate cancerous regions which were validated by histopathologic diagnosis. More than two million reflectance spectral signatures were obtained by HSI and analyzed using machine-learning methods. The detection results of HSI were compared with autofluorescence imaging and fluorescence imaging of two vital-dyes of the same specimens.Results: Quantitative HSI differentiated cancerous tissue from normal tissue in ex vivo surgical specimens with a sensitivity and specificity of 91% and 91%, respectively, and which was more accurate than autofluorescence imaging (P < 0.05) or fluorescence imaging of 2-NBDG (P < 0.05) and proflavine (P < 0.05). The proposed quantification tools also generated cancer probability maps with the tumor border demarcated and which could provide real-time guidance for surgeons regarding optimal tumor resection.Conclusions: This study highlights the feasibility of using quantitative HSI as a diagnostic tool to delineate the cancer boundaries in surgical specimens, and which could be translated into the clinic application with the hope of improving clinical outcomes in the future. Clin Cancer Res; 23(18); 5426-36. ©2017 AACR.
View details for DOI 10.1158/1078-0432.CCR-17-0906
View details for PubMedID 28611203
A Minimum Spanning Forest-Based Method for Noninvasive Cancer Detection With Hyperspectral Imaging
IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING
2016; 63 (3): 653-663
The purpose of this paper is to develop a classification method that combines both spectral and spatial information for distinguishing cancer from healthy tissue on hyperspectral images in an animal model.An automated algorithm based on a minimum spanning forest (MSF) and optimal band selection has been proposed to classify healthy and cancerous tissue on hyperspectral images. A support vector machine classifier is trained to create a pixel-wise classification probability map of cancerous and healthy tissue. This map is then used to identify markers that are used to compute mutual information for a range of bands in the hyperspectral image and thus select the optimal bands. An MSF is finally grown to segment the image using spatial and spectral information.The MSF based method with automatically selected bands proved to be accurate in determining the tumor boundary on hyperspectral images.Hyperspectral imaging combined with the proposed classification technique has the potential to provide a noninvasive tool for cancer detection.
View details for DOI 10.1109/TBME.2015.2468578
View details for Web of Science ID 000371933800021
View details for PubMedID 26285052
View details for PubMedCentralID PMC4791052
Framework for hyperspectral image processing and quantification for cancer detection during animal tumor surgery
JOURNAL OF BIOMEDICAL OPTICS
2015; 20 (12)
Hyperspectral imaging (HSI) is an imaging modality that holds strong potential for rapid cancer detection during image-guided surgery. But the data from HSI often needs to be processed appropriately in order to extract the maximum useful information that differentiates cancer from normal tissue. We proposed a framework for hyperspectral image processing and quantification, which includes a set of steps including image preprocessing, glare removal, feature extraction, and ultimately image classification. The framework has been tested on images from mice with head and neck cancer, using spectra from 450- to 900-nm wavelength. The image analysis computed Fourier coefficients, normalized reflectance, mean, and spectral derivatives for improved accuracy. The experimental results demonstrated the feasibility of the hyperspectral image processing and quantification framework for cancer detection during animal tumor surgery, in a challenging setting where sensitivity can be low due to a modest number of features present, but potential for fast image classification can be high. This HSI approach may have potential application in tumor margin assessment during image-guided surgery, where speed of assessment may be the dominant factor.
View details for DOI 10.1117/1.JBO.20.12.126012
View details for Web of Science ID 000368440300037
View details for PubMedID 26720879
View details for PubMedCentralID PMC4691647
Spectral-spatial classification for noninvasive cancer detection using hyperspectral imaging
JOURNAL OF BIOMEDICAL OPTICS
2014; 19 (10)
Early detection of malignant lesions could improve both survival and quality of life of cancer patients. Hyperspectral imaging (HSI) has emerged as a powerful tool for noninvasive cancer detection and diagnosis, with the advantage of avoiding tissue biopsy and providing diagnostic signatures without the need of a contrast agent in real time. We developed a spectral-spatial classification method to distinguish cancer from normal tissue on hyperspectral images. We acquire hyperspectral reflectance images from 450 to 900 nm with a 2-nm increment from tumor-bearing mice. In our animal experiments, the HSI and classification method achieved a sensitivity of 93.7% and a specificity of 91.3%. The preliminary study demonstrated that HSI has the potential to be applied in vivo for noninvasive detection of tumors.
View details for DOI 10.1117/1.JBO.19.10.106004
View details for Web of Science ID 000345837200021
View details for PubMedID 25277147
View details for PubMedCentralID PMC4183763
Medical hyperspectral imaging: a review
JOURNAL OF BIOMEDICAL OPTICS
2014; 19 (1)
Hyperspectral imaging (HSI) is an emerging imaging modality for medical applications, especially in disease diagnosis and image-guided surgery. HSI acquires a three-dimensional dataset called hypercube, with two spatial dimensions and one spectral dimension. Spatially resolved spectral imaging obtained by HSI provides diagnostic information about the tissue physiology, morphology, and composition. This review paper presents an overview of the literature on medical hyperspectral imaging technology and its applications. The aim of the survey is threefold: an introduction for those new to the field, an overview for those working in the field, and a reference for those searching for literature on a specific application.
View details for DOI 10.1117/1.JBO.19.1.010901
View details for Web of Science ID 000331892700001
View details for PubMedID 24441941
View details for PubMedCentralID PMC3895860
Determination of Tumor Margins with Surgical Specimen Mapping Using Near-Infrared Fluorescence.
For many solid tumors, surgical resection remains the gold standard and tumor-involved margins are associated with poor clinical outcomes. Near-infrared (NIR) fluorescence imaging using molecular agents has shown promise for in situ imaging during resection. However, for cancers with difficult imaging conditions, surgical value may lie in tumor-mapping of surgical specimens. We thus evaluated a novel approach for real-time, intraoperative tumor margin assessment. 21 adult patients with biopsy-confirmed squamous cell carcinoma arising from the head and neck (HNSCC) scheduled for standard-of-care surgery were enrolled. Cohort 1 (n=3) received panitumumab-IRDye800CW at an intravenous microdose of 0.06 mg/kg, cohort 2A (n=5) received 0.5mg/kg, cohort 2B (n=7) received 1mg/kg, and cohort 3 (n=6) received 50 mg. Patients were followed 30 days post-infusion and adverse events were recorded. Imaging was performed using several closed- and wide-field devices. Fluorescence was histologically correlated to determine sensitivity and specificity. In situ imaging demonstrated tumor-to-background ratio (TBR) of 2-3, compared to ex vivo specimen imaging TBR of 5-6. We obtained clear differentiation between tumor and normal tissue, with a three-fold signal difference between positive and negative specimens (p<0.05). We achieved high correlation of fluorescence intensity with tumor location with sensitivities and specificities >89%; fluorescence predicted distance of tumor tissue to the cut surface of the specimen. This novel method of detecting tumor-involved margins in surgical specimens using a cancer-specific agent provides highly sensitive and specific, real-time, intraoperative surgical navigation in resections with complex anatomy which are otherwise less amenable to image guidance.
View details for DOI 10.1158/0008-5472.CAN-18-0878
View details for PubMedID 29967260
Ultrasound Imaging Technologies for Breast Cancer Detection and Management: A Review.
Ultrasound in medicine & biology
2018; 44 (1): 37–70
Ultrasound imaging is a commonly used modality for breast cancer detection and diagnosis. In this review, we summarize ultrasound imaging technologies and their clinical applications for the management of breast cancer patients. The technologies include ultrasound elastography, contrast-enhanced ultrasound, 3-D ultrasound, automatic breast ultrasound and computer-aided detection of breast ultrasound. We summarize the study results seen in the literature and discuss their future directions. We also provide a review of ultrasound-guided, breast biopsy and the fusion of ultrasound with other imaging modalities, especially magnetic resonance imaging (MRI). For comparison, we also discuss the diagnostic performance of mammography, MRI, positron emission tomography and computed tomography for breast cancer diagnosis at the end of this review. New ultrasound imaging techniques, ultrasound-guided biopsy and the fusion of ultrasound with other modalities provide important tools for the management of breast patients.
View details for DOI 10.1016/j.ultrasmedbio.2017.09.012
View details for PubMedID 29107353
Label-free reflectance hyperspectral imaging for tumor margin assessment: a pilot study on surgical specimens of cancer patients.
Journal of biomedical optics
2017; 22 (8): 1–7
A label-free, hyperspectral imaging (HSI) approach has been proposed for tumor margin assessment. HSI data, i.e., hypercube (x,y,λ), consist of a series of high-resolution images of the same field of view that are acquired at different wavelengths. Every pixel on an HSI image has an optical spectrum. In this pilot clinical study, a pipeline of a machine-learning-based quantification method for HSI data was implemented and evaluated in patient specimens. Spectral features from HSI data were used for the classification of cancer and normal tissue. Surgical tissue specimens were collected from 16 human patients who underwent head and neck (H&N) cancer surgery. HSI, autofluorescence images, and fluorescence images with 2-deoxy-2-[(7-nitro-2,1,3-benzoxadiazol-4-yl)amino]-D-glucose (2-NBDG) and proflavine were acquired from each specimen. Digitized histologic slides were examined by an H&N pathologist. The HSI and classification method were able to distinguish between cancer and normal tissue from the oral cavity with an average accuracy of 90%±8%, sensitivity of 89%±9%, and specificity of 91%±6%. For tissue specimens from the thyroid, the method achieved an average accuracy of 94%±6%, sensitivity of 94%±6%, and specificity of 95%±6%. HSI outperformed autofluorescence imaging or fluorescence imaging with vital dye (2-NBDG or proflavine). This study demonstrated the feasibility of label-free, HSI for tumor margin assessment in surgical tissue specimens of H&N cancer patients. Further development of the HSI technology is warranted for its application in image-guided surgery.
View details for DOI 10.1117/1.JBO.22.8.086009
View details for PubMedID 28849631
View details for PubMedCentralID PMC5572439
Functional MRI of the Eustachian Tubes in Patients With Nasopharyngeal Carcinoma: Correlation With Middle Ear Effusion and Tumor Invasion
AMERICAN JOURNAL OF ROENTGENOLOGY
2016; 206 (3): 617-622
We sought to combine the Valsalva maneuver with MRI to evaluate eustachian tube function in patients with nasopharyngeal carcinoma (NPC) and to correlate the extent of tumor invasion with the presence of middle ear effusion (MEE) and eustachian tube dysfunction (ETD).We performed MRI along the lengths of the eustachian tubes, before and after the Valsalva maneuver was performed, in 53 patients with untreated NPC. The images were reviewed by two radiologists.A total of 106 eustachian tubes and middle ears were studied. There was dysfunction in 37 eustachian tubes, which was always ipsilateral to the NPC. There was MEE in 26 ears of 22 patients. In all cases of MEE, there was ipsilateral ETD. ETD was correlated with tumor invasion of the ipsilateral pharyngeal recess (p < 0.001), pharyngeal opening of the eustachian tube (p < 0.001), the cartilaginous eustachian tube (p < 0.001), the eustachian cartilage (p < 0.001), Ostmann fat pad (p < 0.001), the levator veli palatine muscle (p < 0.001), and the tensor veli palatine muscle (p < 0.001). There was a strong correlation between the grade of parapharyngeal space invasion and ETD (r = 0.809; p < 0.001) and MEE (r = 0.693; p < 0.001).Combining the Valsalva maneuver with MRI is helpful in assessing the function of the eustachian tube in patients with NPC. The cause of MEE in patients with NPC is dysfunction of the eustachian tube opening, which is associated with tumor invasion around the eustachian tube.
View details for DOI 10.2214/AJR.15.14751
View details for Web of Science ID 000370848400028
View details for PubMedID 26901020
Simulating cardiac ultrasound image based on MR diffusion tensor imaging
2015; 42 (9): 5144-5156
Cardiac ultrasound simulation can have important applications in the design of ultrasound systems, understanding the interaction effect between ultrasound and tissue and setting the ground truth for validating quantification methods. Current ultrasound simulation methods fail to simulate the myocardial intensity anisotropies. New simulation methods are needed in order to simulate realistic ultrasound images of the heart.The proposed cardiac ultrasound image simulation method is based on diffusion tensor imaging (DTI) data of the heart. The method utilizes both the cardiac geometry and the fiber orientation information to simulate the anisotropic intensities in B-mode ultrasound images. Before the simulation procedure, the geometry and fiber orientations of the heart are obtained from high-resolution structural MRI and DTI data, respectively. The simulation includes two important steps. First, the backscatter coefficients of the point scatterers inside the myocardium are processed according to the fiber orientations using an anisotropic model. Second, the cardiac ultrasound images are simulated with anisotropic myocardial intensities. The proposed method was also compared with two other nonanisotropic intensity methods using 50 B-mode ultrasound image volumes of five different rat hearts. The simulated images were also compared with the ultrasound images of a diseased rat heart in vivo. A new segmental evaluation method is proposed to validate the simulation results. The average relative errors (AREs) of five parameters, i.e., mean intensity, Rayleigh distribution parameter σ, and first, second, and third quartiles, were utilized as the evaluation metrics. The simulated images were quantitatively compared with real ultrasound images in both ex vivo and in vivo experiments.The proposed ultrasound image simulation method can realistically simulate cardiac ultrasound images of the heart using high-resolution MR-DTI data. The AREs of their proposed method are 19% for the mean intensity, 17.7% for the scale parameter of Rayleigh distribution, 36.8% for the first quartile of the image intensities, 25.2% for the second quartile, and 19.9% for the third quartile. In contrast, the errors of the other two methods are generally five times more than those of their proposed method.The proposed simulation method uses MR-DTI data and realistically generates cardiac ultrasound images with anisotropic intensities inside the myocardium. The ultrasound simulation method could provide a tool for many potential research and clinical applications in cardiac ultrasound imaging.
View details for DOI 10.1118/1.4927788
View details for Web of Science ID 000360645000017
View details for PubMedID 26328966
View details for PubMedCentralID PMC4537486