Search

Background Whether the characteristics and outcome of secondary acute promyelocytic leukemia (s-APL) are similar to de no APL (dn-APL) remains unknown. Patients and Methods Using the SEER database, we identified 3877 patients with APL diagnosed from 2000 to 2014, including 465 s-APL and 3412 dn-APL. Results Compared with dn-APL, s-APL werecharacterized by older median age, and a higher early mortality rate. Multivariate Cox model showed s-APL, older age, earlier year of diagnosis, and male gender were independently associated with worse survival. Notably, s-APL had a significantly inferior survival regardless of gender, race, marital status, and year of diagnosis. However, the difference between the 2 cohorts was only evident in younger patients (≤ 65 years) but was lost in older patients (> 65 years). Additionally, the majority of index cancer type was breast and prostate in female and male s-APL, respectively. Latency Conclusions Inferior survival of s-APL points to the need for treatment improvement.

Due to limited information reported on the clinical characteristics and outcomes of Burkitt lymphoma (BL) patients with gastrointestinal (GI) involvement, here we used the Surveillance, Epidemiology, and End Results (SEER) database to perform our study in a population‐based scale. Extranodal GI involvement was categorized into gastric and intestinal primary sites. A total of 477 BL patients with GI involvement extracted from the SEER database between 2004 and 2015 were included in this study, 112 (23.5%) with the stomach and 365 (76.5%) with the intestine. Our study demonstrated that gastric involvement, older age, male gender, black race, advanced‐stage III/IV, no‐chemotherapy, and earlier years of diagnosis were associated with a significantly worse overall survival (OS) in GI BL patients after adjustment in multivariate analysis, whereas marital status did not significantly influence OS. Notably, BL Patients with gastric involvement had a significantly inferior 5‐year OS in both univariate and multivariate analysis, as compared to those with intestinal involvement (37.8% vs. 70.2%; Univariate: HR = 2.637, P, Our study included 477 Burkitt Lymphoma (BL) patients with gastrointestinal involvement extracted from the SEER database between 2004 and 2015, 112 (23.5%) with the stomach and 365 (76.5%) with the intestine. We demonstrated that BL patients with gastric involvement had an inferior prognosis compared to those with intestinal involvement (5‐year OS 37.8% vs. 70.2%). In subgroup analysis, gastric BL patients had a consistently worse OS than intestinal patients regardless of gender, clinical stage and year of diagnosis.

Purpose The study aims at a novel dosimetry methodology to reconstruct a 3D dose distribution as imparted to a virtual cylindrical phantom using an electronic portal imaging device (EPID). Methods A deep learning-based signal processing strategy, referred to as 3DosiNet, is utilized to learn a mapping from an EPID image to planar dose distributions at given depths. The network was trained with the volumetric dose exported from the clinical treatment planning system (TPS). Given the latent inconsistency between measurements and corresponding TPS calculations, unsupervised learning is formulated in 3DosiNet to capture abstractive image features that are less sensitive to the potential variations. Results Validation experiments were performed using five regular fields and three clinical intensity-modulated radiation therapy (IMRT) cases. The measured dose profiles and percentage depth dose (PDD) curves were compared with those measured using standard tools in terms of the 1D gamma index. The mean gamma pass rates (2%/2 mm) over the regular fields are 100% and 97.3% for the dose profile and PDD measurements, respectively. The measured volumetric dose was compared to the corresponding TPS calculation in terms of the 3D gamma index. The mean 2%/2 mm gamma pass rates are 97.9% for square fields and 94.9% for the IMRT fields. Conclusions The system promises to be a practical 3D dosimetric tool for pre-treatment patient-specific quality assurance and further developed for in-treatment patient dose monitoring.

The automatic diagnosis of various retinal diseases from fundus images is important to support clinical decision-making. However, developing such automatic solutions is challenging due to the requirement of a large amount of human-annotated data. Recently, unsupervised/self-supervised feature learning techniques receive a lot of attention, as they do not need massive annotations. Most of the current self-supervised methods are analyzed with single imaging modality and there is no method currently utilize multi-modal images for better results. Considering that the diagnostics of various vitreoretinal diseases can greatly benefit from another imaging modality, e.g., FFA, this paper presents a novel self-supervised feature learning method by effectively exploiting multi-modal data for retinal disease diagnosis. To achieve this, we first synthesize the corresponding FFA modality and then formulate a patient feature-based softmax embedding objective. Our objective learns both modality-invariant features and patient-similarity features. Through this mechanism, the neural network captures the semantically shared information across different modalities and the apparent visual similarity between patients. We evaluate our method on two public benchmark datasets for retinal disease diagnosis. The experimental results demonstrate that our method clearly outperforms other self-supervised feature learning methods and is comparable to the supervised baseline., Comment: IEEE Transactions on Medical Imaging, code is at https://github.com/xmengli999/self_supervised

In a standard computed tomography (CT) image, pixels having the same Hounsfield Units (HU) can correspond to different materials and it is therefore challenging to differentiate and quantify materials. Dual-energy CT (DECT) is desirable to differentiate multiple materials, but DECT scanners are not widely available as singleenergy CT (SECT) scanners. Here we develop a deep learning approach to perform DECT imaging by using standard SECT data. The end point of the deep learning approach is a model capable of providing the high-energy CT image for a given input low-energy CT image. We retrospectively studied 22 patients who received contrast-enhanced abdomen DECT scan. The difference between the predicted and original high-energy CT images are 3.47 HU, 2.95 HU, 2.38 HU, and 2.40 HU for spine, aorta, liver and stomach, respectively. The difference between virtual non-contrast (VNC) images obtained from original DECT and deep learning DECT are 4.10 HU, 3.75 HU, 2.33 HU and 2.92 HU for spine, aorta, liver and stomach, respectively. The aorta iodine quantification difference between iodine maps obtained from original DECT and deep learning DECT images is 0.9%. This study demonstrates that highly accurate DECT imaging with single low-energy data is achievable by using a deep learning approach. The proposed method can significantly simplify the DECT system design, reducing the scanning dose and imaging cost.

Luminescence molecular tomography with Cherenkov excitation offers the ability to non-invasively image and quantify temporal changes in fluorescence throughout the body, and then further realize tumor localization. This can be done in radiotherapy to determine the response to treatment in fractionated therapy. To obtain high signal-to-background or signal-to-noise ratio measurement, it is critical to know the best post time point of in-vivo agent-based molecular imaging, which could account on a high signal ratio of target to skin (TSR). For this purpose, ex-vivo murine experiments were performed to quantify the biokinetics and biodistribution of the major organs, plasma, tumor, and skin.

This study aims to demonstrate a low-cost camera-based radioluminescence imaging system (CRIS) for high-quality beam visualization that encourages accurate pre-treatment verifications on radiation delivery in external beam radiotherapy. To ameliorate the optical image that suffers from mirror glare and edge blurring caused by photon scattering, a deep learning model is proposed and trained to learn from an on-board electronic portal imaging device (EPID). Beyond the typical purposes of an on-board EPID, the developed system maintains independent measurement with co-planar detection ability by involving a cylindrical receptor. Three task-aware modules are integrated into the network design to enhance its robustness against the artifacts that exist in an EPID running at the cine mode for efficient image acquisition. The training data consists of various designed beam fields that were modulated via the multi-leaf collimator (MLC). Validation experiments are performed for five regular fields ranging from 2 × 2 cm2 to 10 × 10 cm2 and three clinical IMRT cases. The captured CRIS images are compared to the high-quality images collected from an EPID running at the integration-mode, in terms of gamma index and other typical similarity metrics. The mean 2%/2 mm gamma pass rate is 99.14% (range between 98.6% and 100%) and 97.1% (ranging between 96.3% and 97.9%), for the regular fields and IMRT cases, respectively. The CRIS is further applied as a tool for MLC leaf-end position verification. A rectangular field with introduced leaf displacement is designed, and the measurements using CRIS and EPID agree within 0.100 mm ± 0.072 mm with maximum of 0.292 mm. Coupled with its simple system design and low-cost nature, the technique promises to provide viable choice for routine quality assurance in radiation oncology practice.

Laminar optical tomography (LOT) is a new mesoscopic functional optical imaging technique. Currently, the forward problem of LOT image reconstruction is generally solved on the basis of Monte-Carlo (MC) methods. However, considering the nonlinear nature of the image reconstruction in LOT with the increasing number of source positions, methods based on MC take too much computation time. This letter develops a fast image reconstruction algorithm based on perturbation MC (pMC) for reconstructing the absorption or scattering image of a slab medium, which is suitable for LOT or other functional optical tomography system with narrow source-detector separation and dense sampling. To calculate the pMC parameters, i.e., the path length passed by a photon and the collision numbers experienced in each voxel with only one baseline MC simulation, we propose a scheme named as the trajectory translation and target voxel regression (TT&TVR) based on the reciprocity principle. To further speed up the image reconstruction procedure, the weighted average of the pMC parameters for all survival photons is adopted and the region of interest (ROI) is extracted from the raw data to save as the prior information of the image reconstruction. The method is applied to the absorption reconstruction of the layered inhomogeneous media. Results demonstrate that the reconstructing time is less than 20 s with the X-Y section of the sample subdivided into 50 \times 50 voxels, and the target size quantitativeness ratio can be obtained in a satisfying accuracy in the source-detector separations of 0.4 and 1.25 mm, respectively.

In a typical laminar optical tomography (LOT) system, the dip-angle between the incident light (or the emitting light) and the normal of the detection plane randomly changes during raster-scanning. The inconstant dip-angle causes consistency between the measurement and the light transportation model where a fixed dip-angle of the incident light is generally required. To eliminate the effect from this dip angle, methods such as keeping the angle unchangeable by moving the phantom instead of scanning the light were investigated. In this paper, a LOT system with small dip-angle over the whole detection range is developed. Simulation and experimental evaluation show that the dip-angle of the modified system is much smaller than that of the traditional system. For example, the relative angle between the two incident light at (x=0mm, y=0mm) and (x=0mm, y=2.5mm) on the image plane is about 0.7° for the traditional system while that is only about 0.02° for the modified system. The main parameters of the system are also evaluated and an image reconstruction algorithm is developed based on Monte Carlo simulation. The reconstructed images show that the spatial resolution and quantitative ratio is improved by the modified system without loss of the scanning speed.

Most analytical methods for describing light propagation in turbid medium exhibit low effectiveness in the near-field of a collimated source. Motivated by the Charge Simulation Method in electromagnetic theory as well as the established discrete source based modeling, we have reported on an improved explicit model, referred to as "Virtual Source" (VS) diffuse approximation (DA), to inherit the mathematical simplicity of the DA while considerably extend its validity in modeling the near-field photon migration in low-albedo medium. In this model, the collimated light in the standard DA is analogously approximated as multiple isotropic point sources (VS) distributed along the incident direction. For performance enhancement, a fitting procedure between the calculated and realistic reflectances is adopted in the nearfield to optimize the VS parameters (intensities and locations). To be practically applicable, an explicit 2VS-DA model is established based on close-form derivations of the VS parameters for the typical ranges of the optical parameters. The proposed VS-DA model is validated by comparing with the Monte Carlo simulations, and further introduced in the image reconstruction of the Laminar Optical Tomography system.

Radiation therapy produces Cherenkov optical emission in tissue, and this light can be utilized to activate molecular probes. The feasibility of sensing luminescence from a tissue molecular oxygen sensor from within a human body phantom was examined using the geometry of the axillary lymph node region. Detection of regions down to 30-mm deep was feasible with submillimeter spatial resolution with the total quantity of the phosphorescent sensor PtG4 near 1 nanomole. Radiation sheet scanning in an epi-illumination geometry provided optimal coverage, and maximum intensity projection images provided illustration of the concept. This work provides the preliminary information needed to attempt this type of imaging in vivo.

In this paper, a cervical cancer screening system developed in this group is introduced. 476 patients were clinically examined and total coincidence rate to the histopathology is 72.1%.

The cervical cancer screening at a pre-cancer stage is beneficial to reduce the mortality of women. An opto-electronic joint detection system based on DSP aiming at early cervical cancer screening is introduced in this paper. In this system, three electrodes alternately discharge to the cervical tissue and three light emitting diodes in different wavelengths alternately irradiate the cervical tissue. Then the relative optical reflectance and electrical voltage attenuation curve are obtained by optical and electrical detection, respectively. The system is based on DSP to attain the portable and cheap instrument. By adopting the relative reflectance and the voltage attenuation constant, the classification algorithm based on Support Vector Machine (SVM) discriminates abnormal cervical tissue from normal. We use particle swarm optimization to optimize the two key parameters of SVM, i.e. nuclear factor and cost factor. The clinical data were collected on 313 patients to build a clinical database of tissue responses under optical and electrical stimulations with the histopathologic examination as the gold standard. The classification result shows that the opto-electronic joint detection has higher total coincidence rate than separate optical detection or separate electrical detection. The sensitivity, specificity, and total coincidence rate increase with the increasing of sample numbers in the training set. The average total coincidence rate of the system can reach 85.1% compared with the histopathologic examination.

Image reconstruction in the most model-based biophotonic imaging modalities essentially poses an ill-posed nonlinear inverse problem, which has been effectively tackled in the diffusion-approximation-satisfied scenarios such as diffuse optical tomography. Nevertheless, a nonlinear implementation in high-resolution laminar optical tomography (LOT) is normally computationally-costly due to its strong dependency on a dense source-detector configuration and a physically-rigorous photon-transport model. To circumvent the adversity, we herein propose a practical nonlinear LOT approach to the absorption reconstruction. The scheme takes advantage of the numerical stability of the singular value decomposition (SVD) for the ill-posed linear inversion, and is accelerated by adopting an explicitly recursive strategy for the time-consuming repeated SVD inversion, which is based on a scaled expression of the sensitivity matrix. Experiments demonstrate that the proposed methodology can perform as well as the traditional nonlinear one, while the computation time of the former is merely 26.27% of the later on average.

Laminar optical tomography (LOT) is a new mesoscopic functional optical imaging technique. Currently, the forward problem of LOT image reconstruction is generally solved on the basis of Monte-Carlo (MC) methods. However, considering the nonlinear nature of the image reconstruction in LOT, with the increasing number of source positions, methods based on MC takes too much computation time. Based on the scheme of trajectory translation and target voxel regression (TT&TVR) proposed by our group, this paper develops a fast 3D image reconstruction algorithm. The algorithm is applied to the absorption reconstruction of the layered inhomogeneous media. Results demonstrate that the reconstructing time is less than 15min with the X-Y-Z section of the sample subdivided into 50 × 50 × 10 voxels, and the target size and quantitativeness ratio can be obtained in a satisfying accuracy.

Laminar optical tomography (LOT) is a new mesoscopic functional optical imaging technique, which is an extension of a confocal microscope and diffuse optical tomography to acquire both the coaxial and off-axis scattered light at the same time. In this paper, a LOT system with a larger detection area aiming at the in vivo detection of early cervical cancer is developed. The field of view of our system is 10 mm x 10 mm. In order to improve the image quality of the system, two methods were performed: the correction of image distortion and the restriction of returning light. The performance of the system with aperture stop was assessed by liquid phantom experiments. Comparing with the Monte Carlo simulation, the measurement results show that the average relative errors of eight different source-detector distances corresponding to 4 source points are lower than the errors of the system taking the frame of objective lens as the aperture stop by 5.7%， 4.8%，6.1%，6.1% respectively. Moreover, the experiment based on the phantom with specified structure and optical parameters to simulate the cervix demonstrates that the system perform well for the cervix measurement.