Performance Of Imaging System Research Articles

Assessing the performance of ultrasound imaging systems using images from relatively high-density random spherical void phantoms: A simulation study.

The development of clinically meaningful, objective, and quantitative methods for assessing the performance of ultrasound imaging systems represents a continuing area of interest. One approach has been to image phantoms with randomly distributed spherical voids. The objectives of this study were: (1) to explore the potential of using relatively high-volume fraction random spherical void (RSV) phantoms as an approach for quantitatively assessing the performance of ultrasound imaging systems; (2) to identify potential metrics that can be used to provide quantitative assessments of images obtained from relatively high-volume fraction RSV phantoms; and (3) to demonstrate changes in the quantitative metrics that can occur as image features are degraded. A series (10 each) of computer-simulated RSV phantoms with a range of RSV volume fractions (0.05, 0.15, and 0.25) were generated. To determine the number of image planes necessary to provide robust measurements, a series of consecutive planes (ranging from 1 to 150) within each type of simulated phantom were analyzed. The observed circular cross-section radii histogram distributions (representing the intersection of each plane with the local distribution of spherical voids) were compared with the theoretical histogram distribution. Simulated phantom images were produced by adding speckle and degradation of imaging system performance was modeled by averaging 1 to 9 neighboring planes to represent increasing elevation plane thicknesses. Quantification of the performance of the imaging system was determined by measuring the: (1) mean number of circular cross-sections detected per image frame; (2) mean fractional area of circular cross-sections detected per image frame; (3) agreement of observed circular cross-section radii histogram distribution with the theoretical distribution (Chi-square statistic); and (4) contrast and contrast-to-noise ratio as a function of observed circular cross-section radius. Results suggest that analyses of a sufficient number of image planes (providing over approximately 3000 total circular cross-sectional areas) provides excellent agreement between the observed and theoretical histogram distributions (mean Chi-square<0.004). For the 0.15 volume fraction series of simulated RSV phantoms, using 150 image plane analyses, phantom images show decreasing mean number of circle cross-sections detected per frame (31.5 ± 0.3, 28.4 ± 0.3, 28.2 ± 0.3, 26.3 ± 0.3, and 25.3 ± 0.3); decreasing mean fractional area of circle cross-sections per frame (0.157 ± 0.002, 0.133 ± 0.001, 0.133 ± 0.001, 0.111 ± 0.001, and 0.108 ± 0.001); and a decreasing agreement with the theoretical histogram distribution of radii (Chi-square values: 0.070 ± 0.004, 0.140 ± 0.005, 0.149 ± 0.007, 0.379 ± 0.011, and 0.518 ± 0.010) for 1, 3, 5, 7, and 9 plane averages, respectively. Contrast and contrast-to-noise measurements as a function of observed circular cross-section radius also demonstrate marked changes with simulated image degradation. Results of this simulation study suggest that analyses of images obtained from relatively high-density RSV phantoms may offer a promising approach for assessing ultrasound imaging systems. The proposed measurements appear to provide reproducible, robust, quantitative metrics that can be compared with corresponding theoretical values to provide quantifiable, objective metrics of imaging system performance.

Extended aberration analysis in symmetry-free optical systems – part II: evaluation and application

In part I [Opt. Express 29, 39967 (2021)10.1364/OE.439862] of this publication, the working principle of a mixed ray-tracing (MRT) method is introduced for surface-decomposed transverse aberration contributions, and the settings and various calculation options are discussed. Following part I, this paper is a supplement with the evaluation of the calculation results and more illustrations of practical applications, which focuses on the support of the MRT method during the symmetry-free optical system design process. The comprehensive analysis of a group of lithography systems as an application example proves that the MRT method is a powerful tool for imaging system performance assessment, as well as for relative sensitivity analysis among the surfaces in the optical system.

Dynamic Data Driven Multistatic Radio Frequency Imaging

Multistatic radio frequency (RF) imaging systems utilize distributed RF sensors which transmit waveforms to illuminate a target scene and estimate the dielectric properties of the region of interest from the received echoes. This article applies the principles of dynamic data driven systems to improve the performance of multistatic RF imaging system in terms of power efficiency and image reconstruction accuracy. Target location information derived from initial image reconstruction is applied to dynamically reconfigure the imaging system. Based on the geolocation of targets relative to the distributed RF sensors, Fisher information matrices associated with the targets and multistatic sensor pairs are calculated and applied to derive optimum power allocation strategies under different constraints. Image reconstruction of multistatic RF imaging often suffers from artifacts and ghost targets caused by multipath propagation and multiorder reflections.A grouped-coordinate descent type reconstruction algorithm is developed to exploit target location information. The iterative optimization alternates between the groups of target space parameters and nontarget space parameters. The dictionary of reconstruction is dynamically updated to include the secondary reflections from target locations. The improved model fidelity leads to more accurate reconstructions without a significant increase in computational complexity. Numerical simulations demonstrate that the proposed power allocation strategies are effective in energy saving. The proposed reconstruction algorithm converges faster than sequential parameter update algorithms, and the reconstruction accuracy is superior to that using a fixed dictionary containing the first order reflections only.

Fluorescent nanodiamonds for characterization of nonlinear microscopy systems.

Characterizing the performance of fluorescence microscopy and nonlinear imaging systems is an essential step required for imaging system optimization and quality control during longitudinal experiments. Emerging multimodal nonlinear imaging techniques require a new generation of microscopy calibration targets that are not susceptible to bleaching and can provide a contrast across the multiple modalities. Here, we present a nanodiamond-based calibration target for microscopy, designed for facilitating reproducible measurements at the object plane. The target is designed to support day-to-day instrumentation development efforts in microscopy laboratories. The images of a phantom contain information about the imaging performance of a microscopy system across multiple spectral windows and modalities. Since fluorescent nanodiamonds are not prone to bleaching, the proposed imaging target can serve as a standard, shelf-stable sample to provide rapid reference measurements for ensuring consistent performance of microscopy systems in microscopy laboratories and imaging facilities.

Temperature environmental adaptability on electrical and stress properties of Ge film in space

The Ge-film-based photon-counting imaging detector is a key component of optical payload in space weather observation. It's important to study the space temperature adaptability of Ge film since the performances of Ge film are extremely sensitive to the temperature. Herein, the in-situ tests are proposed to measure the resistance and stress of the Ge film at the operating temperature. It is demonstrated that the resistance decreases from 520 MΩ/□ to 124 MΩ/□ linearly between −20 °C and 80 °C, which can meet the requirement of the imaging system well. And the stress gradually releases in the thermal cycle. The ex-situ test is adopted to investigate the limited temperature adaptability and the intrinsic mechanism of performance variations. When the temperature increased to 600 °C, Ge film failed and cracked because of resistance reducing to 30 KΩ/□ and stress increasing to 504 Mpa. Meanwhile, methods were proposed to enhance the temperature adaptability of Ge film by analyzing crystal phase and optical band gap. Thus, the resistance and stress changes of Ge film in space have little effect on the performances of imaging system. The Ge film with high temperature adaptability and low stress we prepared is promising in the space observation.

A Micro-flow Phantom for Superficial Micro-vasculature Imaging

Visualization of cutaneous micro-vasculatures is a powerful approach assisting in the diagnosis of skin vascular disorders. These minute structures can be visualized by high-frequency ultrasound (HFUS) using ultrafast Doppler imaging. Ultrasound flow phantoms have been used as assessment tools to evaluate the performance of the ultrasound imaging system, however, to optimize the imaging system for visualization of micro-structures, flow phantom with micro-channels is required which are usually difficult to fabricate. Here, we design a simple approach for micro-flow phantom which is easy to fabricate and cast for detection of micro-circulation in superficial micro-structures. The proposed approach features (i) the micro-channels of 200-micron at the depth of 4 mm (ii) casted in the cryogel mixture of Poly-vinyl alcohol (PVA) and (iii) infused at flow speed of 30 mm/s using infusion pump. Visualization of micro-flow channel in power Doppler image obtained by HFUS ultrafast Doppler imaging reveals that the proposed micro-flow phantom could serve as a viable assessment tool for optimizing the system for in-vivo cutaneous micro-vasculature imaging.

Evaluation and consistency calibration of hyperspectral imaging system based on liquid crystal tunable filter for fabric color measurement

AbstractThe Hyperspectral Imaging System can obtain the spectral response value of each pixel in a fabric sample image while being applied to fabric color measurement. Compared with the common grating Hyperspectral Imaging System, the Hyperspectral Imaging System based on Liquid Crystal Tunable Filter (LCTF) has the advantages of small size, high spatial resolution, and easy integration, which has potential application prospects in textile industry. In this article, the performance of the LCTF Hyperspectral Imaging System for fabric color measurement was evaluated, and an improved model for calibrating the consistency between it and Datacolor 650 was proposed. First, the 210 Pantone cotton color cards were used as samples to study the color repeatability and reproducibility of the LCTF Hyperspectral Imaging System. For evaluation its color measurement ability, the measured reflectance values were converted to the CIELAB color space and compared with those of Datacolor 650. Then, a Radial Basis Function neural network model was proposed to calibrate the color consistency between LCTF Hyperspectral Imaging System and Datacolor 650 which can fully use the reflectance information of the whole band instead of only adjacent wavelengths. Experimental results show that the LCTF Hyperspectral Imaging System can accurately measure the fabric color with good repeatability and reproducibility, and the consistency between the two instruments was significantly improved based on the Radial Basis Function neural network model compared with the R model.

Reflection-mode acousto-optic imaging using plane wave ultrasound pulses.

.Significance: Performance of an acousto-optic imaging system is limited by light fluence rate and acoustic pressure field distributions characteristics. In optically scattering media, the former determines the achievable contrast, whereas the latter the imaging resolution. The system parameters can be shaped by changing relative positions of ultrasound (US) transducer array and optodes. However, in the case of many potential clinical applications, optimization possibilities in this regard are limited, as a sample is accessible from one side only and using a water tank for coupling is not feasible.Aim: We investigate the possibilities of improving performance of an acousto-optic imaging system operating in reflection mode geometry with linear US array in direct contact with a sample using plane wave instead of focused US pulses.Approach: Differences in acoustic pressure field distributions for various transducer excitation patterns were determined numerically and experimentally. Acousto-optic images of phantoms with and without optically absorbing inclusions were acquired by measuring laser speckle contrast decrease due to the light modulation by plane wave and focused US pulses with different apodization patterns.Results: The residual acoustic pressure field components occupy relatively large volume and contribute to light modulation. Using nonsteered plane wave US pulses instead of focused ones allows one to mitigate their influence. It also allows one to obtain clear two-dimensional reconstructions of light fluence rate maps by shifting transducer apodization along the lateral direction.Conclusions: Using nonsteered plane wave US pulses allows one to achieve better imaging performance than with focused pulses in the assumed system geometry.

Fabrication of microcalcifications for insertion into phantoms used to evaluate x-ray breast imaging systems

Physical breast phantoms can be used to evaluate x-ray imaging systems such as mammography, digital breast tomosynthesis and dedicated breast computed tomography (bCT). These phantoms typically attempt to mimic x-ray attenuation properties of adipose and fibroglandular tissues within the breast. In order to use these phantoms for task-based objective assessment of image quality, relevant diagnostic features should be modeled within the phantom, such as mass lesions and/or microcalcifications. Evaluating imaging system performance in detecting microcalcifications is of particular interest due to its’ clinical significance. Many previously-developed phantoms have used materials that model microcalcifications using unrealistic chemical composition, which do not accurately portray their desired x-ray attenuation and scatter properties. We report here on a new method for developing real microcalcification simulants that can be embedded in breast phantoms. This was achieved in several steps, including cross-linking hydroxyapatite and calcium oxalate powders with a binder called polyvinylpyrrolidone (PVP), and mechanical compression. The fabricated microcalcifications were evaluated by measuring their x-ray attenuation and scatter properties using x-ray spectroscopy and x-ray diffraction systems, respectively, and were demonstrated with x-ray mammography and bCT images. Results suggest that using these microcalcification models will make breast phantoms more realistic for use in evaluating task-based detection performance of the abovementioned breast imaging techniques, and bode well for extending their use to spectral imaging and x-ray coherent scatter computed tomography.

7. An algorithm for using deep learning convolutional neural networks with 3-dimensional depth sensor imaging in scoliosis detection

<h3>BACKGROUND CONTEXT</h3> Timely intervention in growing individuals, such as brace treatment, relies on early detection of adolescent idiopathic scoliosis (AIS). To this end, several screening methods have been implemented. However, these methods have limitations in predicting the Cobb angle. <h3>PURPOSE</h3> This study aimed to evaluate the performance of a 3-dimensional depth sensor imaging system with a deep learning algorithm (DLA), in predicting the Cobb angle in AIS. <h3>STUDY DESIGN/SETTING</h3> Retrospective analysis of prospectively collected, consecutive, nonrandomized series of patients at 5 scoliosis centers in Japan. <h3>PATIENT SAMPLE</h3> One hundred sixty human subjects suspected to have AIS were included. <h3>OUTCOME MEASURES</h3> Patient demographics, radiographic measurements, and predicted Cobb angle derived from the DLA were the outcome measures for this study. <h3>METHODS</h3> One hundred sixty data files were shuffled into 5 datasets with 32 data files at random (dataset 1, 2, 3, 4 and 5) and 5-fold cross validation was performed. The relationships between the actual and predicted Cobb angles were calculated using Pearson's correlation coefficient analyses. The prediction performances of the network models were evaluated using mean absolute error (MAE) and root mean square error (RMSE) between the actual and predicted Cobb angles. The shuffling into five datasets and five-fold cross validation was conducted ten times. There were no study-specific biases related to conflicts of interest. <h3>RESULTS</h3> The mean age was 14.7 ± 2.4 years, and the average Cobb angle was 30° (range: 0˚ to 64˚). The range of correlation coefficients was 0.87 to 0.89 in the five-fold cross validation with 10 repeats. The correlation between the actual Cobb angle and the mean predicted Cobb angle with 10 repeats was 0.91. The range of MAEs was 4.4˚ to 4.7˚in the 5-fold cross validation with 10 repeats. The range of RMSEs was 5.8˚ to 6.3˚. The MAE and RMSE between the actual Cobb angle and the mean predicted Cobb angle with 10 repeats was 4.0˚ and 5.4˚ respectively. At a Cobb angle of 10˚, which confirms a diagnosis of scoliosis, the mean predicted Cobb angle with 10 repeats showed a sensitivity of 0.99, a specificity of 0.42, PPV of 0.95, NPV of 0.71, accuracy of 0.94, PLR of 1.69, and an NLR of 0.03. <h3>CONCLUSIONS</h3> The 3D depth sensor imaging system with its newly innovated CNN for regression is objective and has significant ability to predict the Cobb angle in children and adolescents. This system is expected to be used for screening scoliosis in clinics or physical examination at schools. In addition, this is also a possible surrogate for radiographs to monitor curve progression, preventing unnecessary X-rays for mild case of scoliosis. <h3>FDA DEVICE/DRUG STATUS</h3> This abstract does not discuss or include any applicable devices or drugs.

Impact of iterative reconstruction algorithms on the applicability of Fourier-based detectability index for x-ray CT imaging.

The increasing application of iterative reconstruction algorithms in clinical computed tomography to improve image quality and reduce radiation dose, elicits strong interest, and needs model observers to optimize CT scanning protocols objectively and efficiently. The current paradigm for evaluating imaging system performance relies on Fourier methods, which presuppose a linear, wide-sense stationary system. Long-range correlations introduced by iterative reconstruction algorithms may narrow the applicability of Fourier techniques. Differences in the implementation of reconstruction algorithms between manufacturers add further complexity. The present work set out to quantify the errors entailed by the use of Fourier methods, which can lead to design decisions that do not correlate with detectability. To address this question, we evaluated the noise properties and the detectability index of the ideal linear observer using the spatial approach and the Fourier-based approach. For this purpose, a homogeneous phantom was imaged on two scanners: the Revolution CT (GE Healthcare) and the Somatom Definition AS+ (Siemens Healthcare) at different exposure levels. Images were reconstructed using different strength levels of IR algorithms available on the systems considered: Adaptative Statistical Iterative Reconstruction (ASIR-V) and Sinogram Affirmed Iterative Reconstruction (SAFIRE). Our findings highlight that the spatial domain estimate of the detectability index is higher than the Fourier domain estimate. This trend is found to be dependent on the specific regularization used by IR algorithms as well as the signal to be detected. The eigenanalysis of the noise covariance matrix and of its circulant approximation yields explanation about the evoked trends. In particular, this analysis suggests that the predictive power of the Fourier-based ideal linear observer depends on the ability of each basis analyzed to be relevant to the signal to be detected. The applicability of Fourier techniques is dependent on the specific regularization used by IR algorithms. These results argue for verifying the assumptions made when using Fourier methods since Fourier-task-based detectability index does not always correlate with signal detectability.

Analysis of Compound Stained Cervical Cell Images Using Multi-Spectral Imaging

Multi-spectral imaging technique plays an important role in real-world applications such as medicine and medical detections. This paper proposes a cervical cancer cell screening method to simultaneously adopt TBS classification and DNA quantitative analysis for a single cell smear. Through using compound staining on a smear, the cytoplasm is stained by Papanicolauo and the nucleus is stained by Feulgen. The main evaluation parameter is the DNA content of the nucleus, not the subjective description of cell morphology, which is more objective than the TBS classification method and reduces the chances of missing a diagnosis due to subjective factors. Each nucleus has its own DI value and color image of the whole cell, which is convenient for doctors as it allows them to review and confirm the morphology of cells with a nucleus DI of over 2.5. Mouse liver smears and cervical cases are utilized as the measuring specimens to evaluate the performance of the microscope multi-spectral imaging system; illustrative results demonstrate that the proposed system qualifies, with high accuracy and reliability, and further presents wide application prospects in the early diagnosis of cervical cancer.

Highly sensitive CMOS plasmon detector with a low‐gain buffer amplifier for terahertz imaging systems

AbstractA CMOS plasmon detector with a low‐gain buffer amplifier, which enhances the characteristics of the detector core, is proposed for improving the performance of THz imaging systems. A high voltage gain (AV) of the main amplifier integrated with the detector improves the voltage responsivity (RV) but not the performances of the detector core itself. The increase in the ratio of the RV to the noise‐equivalent power (NEP) in the proposed detector with a low‐gain buffer can improve the signal‐to‐noise ratio (SNR) of the THz images. The proposed detector with a − 1.5 dB AV buffer amplifier is fabricated by TSMC 0.25 μm CMOS technology and measured to have the RV of 357.1 kV/W and NEP of 57.3 pW/√Hz at the gate bias of 0.17 V for 0.2 THz input signals. The measurement results indicate that the proposed detector has an RV lower than that of a detector with an AV of 11.2 dB exhibiting a similar NEP. The obtained 0.2 THz raster scanning images show that the image SNR can be improved by 14.8 dB using in the proposed detector.

CMOS Detector Using Customized Bolt-Wrench Capacitor on Backend Oxide Layer

A CMOS detector with a customized capacitor is proposed to improve the detection performance for terahertz imaging systems. The proposed gate-drain capacitor is formed based on the bolt-wrench shape integrated into the layout design of the detector core. The proposed capacitor is implemented using a metal-insulator-metal stack in the backend oxide layer of the TSMC 0.25- μm CMOS technology. The detector using the customized capacitor exhibits an 18% improvement in the voltage responsivity simulation compared to the detector using the capacitor in the process design kit (PDK). The measurement results of the detector, including the integrated antenna and amplifiers, indicate a voltage responsivity of 2.99 MV/W and a noise-equivalent power of 46.3 pW/ √Hz. The 200-GHz images obtained using the proposed detector demonstrate a 30% correlation improvement with the sample compared to the images obtained using the detector with the PDK capacitor.

No-reference quality assessment for neutron radiographic image based on a deep bilinear convolutional neural network

Neutron imaging (NI) has been widely employed in non-destructive investigations. Since the image quality assessment (IQA) method can be beneficial in reflecting the performance of imaging systems and image processing algorithms, we propose a proof-of-concept IQA method for the NI system based on a deep bilinear convolutional neural network (CNN) framework with two designed datasets. Due to the lack of neutron IQA database, different levels of authentic distortion induced by NI are first simulated on the natural and neutron radiographic images to generate the pre-training and fine-tuning datasets, respectively. Then, the gradient magnitude similarity deviation (GMSD) algorithm and transfer learning method are respectively employed to label the above datasets and optimize the prediction performance. Experimental results demonstrate that the proposed method can maintain good consistency with human perception in predicting the quality scores of both the authentic and enhanced neutron radiographic images.

A Low-Cost Method to Assess the Performance of Surface Guidance Imaging Systems at Non-Zero Couch Angles.

A procedure is presented to assess performance at non-zero couch angles and perform routine quality assurance (QA) on surface-guided radiotherapy (SGRT) imaging systems used for stereotactic radiosurgery (SRS). A low-cost anthropomorphic phantom was used to assess the system under patient-like conditions. The phantom is embedded with a tungsten ball bearing (BB) to facilitate the use of surface imaging (SI) with concurrent megavoltage (MV) imaging to cross-compare and validate SI-reported offsets. Data analysis is done via in-house software that utilized the SGRT system’s log files for automated analysis. This procedure enables users to assess and inter-compare MV-reported offsets with their SGRT system. The analysis provides SGRT system residual error so that users are aware of inherent offsets present in addition to increases in translational offsets due to couch walkout. The procedure was validated with two commercial SGRT systems. The procedure can be used with any surface imaging system and linear accelerator system.

An Efficient Dual-Channel Data Storage and Access Method for Spaceborne Synthetic Aperture Radar Real-Time Processing

With the development of remote sensing technology and very large-scale integrated circuit (VLSI) technology, the real-time processing of spaceborne Synthetic Aperture Radar (SAR) has greatly improved the ability of Earth observation. However, the characteristics of external memory have led to matrix transposition becoming a technical bottleneck that limits the real-time performance of the SAR imaging system. In order to solve this problem, this paper combines the optimized data mapping method and reasonable hardware architecture to implement a data controller based on the Field-Programmable Gate Array (FPGA). First of all, this paper proposes an optimized dual-channel data storage and access method, so that the two-dimensional data access efficiency can be improved. Then, a hardware architecture is designed with register manager, simplified address generator and dual-channel Double-Data-Rate Three Synchronous Dynamic Random-Access Memory (DDR3 SDRAM) access mode. Finally, the proposed data controller is implemented on the Xilinx XC7VX690T FPGA chip. The experimental results show that the reading efficiency of the data controller proposed is 80% both in the range direction and azimuth direction, and the writing efficiency is 66% both in the range direction and azimuth direction. The results of a comparison with the recent implementations show that the proposed data controller has a higher data bandwidth, is more flexible in its design, and is suitable for use in spaceborne scenarios.

Simple target acquisition model based on Fλ/d

The US Army Night Vision and Electronic Sensors Directorate’s Night Vision Integrated Performance Model (NVIPM) is a robust, comprehensive, and sophisticated model that calculates the performance of an infrared imager in the task of target acquisition. The inputs are specifics about the target, atmosphere, optics, detector, electronics, and display and the outputs are probabilities of target detection, recognition, and identification. NVIPM is complicated and takes a good bit of time to master obtaining correct answers based on numerous assumptions and modeling experience. For midwave infrared (MWIR) and longwave infrared (LWIR) sensors, past research has shown that imaging system performance is strongly related to Fλ / d, where F is the f-number, λ is the wavelength, and d is the detector pitch. Fλ / d provides a metric that relates how closely to diffraction-limited performance a sensor operates. We use the past Fλ / d work to develop a simple model that provides probability of target discrimination as a function of range that can be performed with a simple hand calculator or spreadsheet. We compare this model to the NVIPM calculations on 10 very disparate MWIR and LWIR sensors to show robust agreement. We also describe the conditions under which the simple model is valid.

Performance of a novel protease-activated fluorescent imaging system for intraoperative detection of residual breast cancer during breast conserving surgery

PurposeSafe breast cancer lumpectomies require microscopically clear margins. Real-time margin assessment options are limited, and 20–40% of lumpectomies have positive margins requiring re-excision. The LUM Imaging System previously showed excellent sensitivity and specificity for tumor detection during lumpectomy surgery. We explored its impact on surgical workflow and performance across patient and tumor types.MethodsWe performed IRB-approved, prospective, non-randomized studies in breast cancer lumpectomy procedures. The LUM Imaging System uses LUM015, a protease-activated fluorescent imaging agent that identifies residual tumor in the surgical cavity walls. Fluorescent cavity images were collected in real-time and analyzed using system software.ResultsCavity and specimen images were obtained in 55 patients injected with LUM015 at 0.5 or 1.0 mg/kg and in 5 patients who did not receive LUM015. All tumor types were distinguished from normal tissue, with mean tumor:normal (T:N) signal ratios of 3.81–5.69. T:N ratios were 4.45 in non-dense and 4.00 in dense breasts (p = 0.59) and 3.52 in premenopausal and 4.59 in postmenopausal women (p = 0.19). Histopathology and tumor receptor testing were not affected by LUM015. Falsely positive readings were more likely when tumor was present < 2 mm from the adjacent specimen margin. LUM015 signal was stable in vivo at least 6.5 h post injection, and ex vivo at least 4 h post excision.ConclusionsIntraoperative use of the LUM Imaging System detected all breast cancer subtypes with robust performance independent of menopausal status and breast density. There was no significant impact on histopathology or receptor evaluation.

An algorithm for using deep learning convolutional neural networks with three dimensional depth sensor imaging in scoliosis detection

BACKGROUND CONTEXTTimely intervention in growing individuals, such as brace treatment, relies on early detection of adolescent idiopathic scoliosis (AIS). To this end, several screening methods have been implemented. However, these methods have limitations in predicting the Cobb angle. PURPOSEThis study aimed to evaluate the performance of a three-dimensional depth sensor imaging system with a deep learning algorithm, in predicting the Cobb angle in AIS. STUDY DESIGNRetrospective analysis of prospectively collected, consecutive, nonrandomized series of patients at five scoliosis centers in Japan. PATIENT SAMPLEOne hundred and-sixty human subjects suspected to have AIS were included. OUTCOME MEASURESPatient demographics, radiographic measurements, and predicted Cobb angle derived from the deep learning algorithm were the outcome measures for this study. METHODSOne hundred and sixty data files were shuffled into five datasets with 32 data files at random (dataset 1, 2, 3, 4, and 5) and five-fold cross validation was performed. The relationships between the actual and predicted Cobb angles were calculated using Pearson's correlation coefficient analyses. The prediction performances of the network models were evaluated using mean absolute error and root mean square error between the actual and predicted Cobb angles. The shuffling into five datasets and five-fold cross validation was conducted ten times. There were no study-specific biases related to conflicts of interest. RESULTSThe correlation between the actual and the mean predicted Cobb angles was 0.91. The mean absolute error and root mean square error were 4.0° and 5.4°, respectively. The accuracy of the mean predicted Cobb angle was 94% for identifying a Cobb angle of ≥10° and 89% for that of ≥20°. CONCLUSIONSThe three-dimensional depth sensor imaging system with its newly innovated convolutional neural network for regression is objective and has significant ability to predict the Cobb angle in children and adolescents. This system is expected to be used for screening scoliosis in clinics or physical examination at schools.