Phase aberration measurements in medical ultrasound: human studies

Using a modified real-time phased array sector scanner, phase aberrations and amplitude fluctuations across the imaging aperture have been measured in a number of human subjects. Data from these subjects were classified into two categories based on the quality of conventional longitudinal images of the liver. Measured phase aberrations were very small in all subjects exhibiting high quality images. In contrast, large phase aberrations were measured in subjects producing low quality images. However, there were no significant amplitude variations across the array for all subjects studied. These results suggest that the absence of significant phase aberrations is a necessary condition for high quality phased array imaging. If so, improvements in clinical image quality in such subjects may be possible.     

In vivo measurements of ultrasonic beam distortion in the breast

The phase aberrations encountered by ultrasonic pulses propagating through breast tissue in twenty-two female volunteers were measured. The experiments were designed to assess the impact of these aberrations on clinical ultrasonic image quality for a variety of transducer and imaging geometries. The phase aberration profiles of a given patient were correlated with the amount of parenchymal tissue determined from that patient's mammogram. These data are useful in assessing the image quality achievable with conventional ultrasonic imaging systems, and the potential application of adaptive ultrasonic imaging systems. The results indicate that phase aberrations significantly degrade breast image quality for typical transducer frequencies and sizes.     

Phase aberration correction using echo signals from moving targets. II: Experimental system and results.

A method for correcting errors due to near-field tissue inhomogeneities in phased array ultrasound images is evaluated experimentally. The method uses the brightness of a moving speckle-generating target, such as blood, as a quality factor to correct for unknown phase aberrations. A real time experimental system utilizing the technique has been constructed and is described. Initial results from in vitro studies using a flow phantom are compared to theoretical predictions. The results indicate that the technique can provide significant improvements in image quality when imaging through aberrating media, and may find application in clinical imaging through skull and fatty layers.     

Simultaneous 4-phase-shifted full-field optical coherence microscopy

A new method is presented for full-field optical coherence tomography imaging, which permits capturing single shot phase sensitive imaging through simultaneous acquisition of four phase-shifted images with a single camera using unpolarized light for object illumination. Our method retains the full dynamic range of the camera by using different areas of a single camera sensor to capture each image. We demonstrate the performance of our method by imaging phantoms and live cultures of fibroblast, cancer, and macrophage cells to achieve 59 dB sensitivity with isotropic resolution down to 1 μm, and displacement sensitivity down to 0.1 nm. Our method can serve as a platform for developing high resolution imaging systems because when used in conjunction with broadband spatially incoherent light sources, the resolution is not affected by optical aberrations or speckle noise.     

Adaptive optics with pupil tracking for high resolution retinal imaging

Adaptive optics, when integrated into retinal imaging systems, compensates for rapidly changing ocular aberrations in real time and results in improved high resolution images that reveal the photoreceptor mosaic. Imaging the retina at high resolution has numerous potential medical applications, and yet for the development of commercial products that can be used in the clinic, the complexity and high cost of the present research systems have to be addressed. We present a new method to control the deformable mirror in real time based on pupil tracking measurements which uses the default camera for the alignment of the eye in the retinal imaging system and requires no extra cost or hardware. We also present the first experiments done with a compact adaptive optics flood illumination fundus camera where it was possible to compensate for the higher order aberrations of a moving model eye and in vivo in real time based on pupil tracking measurements, without the real time contribution of a wavefront sensor. As an outcome of this research, we showed that pupil tracking can be effectively used as a low cost and practical adaptive optics tool for high resolution retinal imaging because eye movements constitute an important part of the ocular wavefront dynamics.OCIS codes: (110.1080) Active or adaptive optics, (100.4999) Pattern recognition, target tracking, (170.4460) Ophthalmic optics and devices, (170.3890) Medical optics instrumentation     

Quantitative Jones matrix imaging using vectorial Fourier ptychography

This paper presents a microscopic imaging technique that uses variable-angle illumination to recover the complex polarimetric properties of a specimen at high resolution and over a large field-of-view. The approach extends Fourier ptychography, which is a synthetic aperture-based imaging approach to improve resolution with phaseless measurements, to additionally account for the vectorial nature of light. After images are acquired using a standard microscope outfitted with an LED illumination array and two polarizers, our vectorial Fourier ptychography (vFP) algorithm solves for the complex 2x2 Jones matrix of the anisotropic specimen of interest at each resolved spatial location. We introduce a new sequential Gauss-Newton-based solver that additionally jointly estimates and removes polarization-dependent imaging system aberrations. We demonstrate effective vFP performance by generating large-area (29 mm²), high-resolution (1.24 μm full-pitch) reconstructions of sample absorption, phase, orientation, diattenuation, and retardance for a variety of calibration samples and biological specimens.     

医学图像有损压缩技术研究进展

医学成像已经成为医学诊断与处理的重要依据。随着医学成像设备分辨率的不断提高,所获取的医学图像数据量也在持续增长,这给医学图像的存储与实时传输带来了巨大压力,并严重制约了其后续应用。有损压缩方式能够在满足一定图像质量的条件下,实现医学图像较大程度的压缩,目前已成为国内外研究的热点。本文对医学图像有损压缩技术研究进展进行总结,并介绍有损压缩条件下的质量评价方法,最后对医学图像有损压缩技术的发展趋势进行展望。     

Cardiac motion and strain detection using 4D CT images: comparison with tagged MRI,and echocardiography

Ischemic heart disease is a leading cause of death in the modern world. Coronary obstruction is the basis for ischemic heart disease and leads to decreased cardiac supply and decreased contractility of the myocardium. Recently, high quality 4D computed tomography (CT) has become available for cardiac imaging and provides the clinician with high quality anatomical images. In this article, a new method is proposed to detect 3D motion and strain from 4D cardiac CT images by constraining intensity constancy, myocardial volume changes and motion smoothness assumptions. The proposed method is validated by using manual tracking of the cardiac CT landmarks. The average error for the manual tracking, by an expert, was 2.9 ± 0.9 mm. As an additional validation, the cardiac CT strain values were compared to the cardiac tagged magnetic resonance imaging (MRI) and 2D B-mode echocardiography strain values of the same patients. The correlation ratio was significantly high for CT and tagged MRI radial strain values (r = 0.76, 95 % confidence interval, P < 0.001). The correlation ratio was meaningful for CT and echocardiography radial strain values as well (r = 0.67, 95 % confidence interval, P < 0.001). The correlation ratio for CT and tagged MRI circumferential strain values was acceptable (r = 0.73, 95 % confidence interval, P < 0.001), while the correlation ratio for CT and echocardiography circumferential strain values was good as well (r = 0.61, 95 % confidence interval, P < 0.001). In general, motion and strain values computed from cardiac CT images agree with motion and strain values computed from tagged MRI and echocardiography images.     

Image recognition and neuronal networks: Intelligent systems for the improvement of imaging information

Summary

Intelligent computerised systems can provide useful assistance to the physician in the rapid identification of tissue abnormalities and accurate diagnosis in real-time. This paper reviews basic issues in medical imaging and neural network-based systems for medical image interpretation. In the framework of intelligent systems, a simple scheme that has been implemented is presented as an example of the use of intelligent systems to discriminate between normal and cancerous regions in colonoscopic images. Preliminary results indicate that this scheme is capable of high accuracy detection of abnormalities within the image. It can also be successfully applied to different types of images, to detect abnormalities that belong to different cancer types.     

Computed optical interferometric tomography for high-speed volumetric cellular imaging

Three-dimensional high-resolution imaging methods are important for cellular-level research. Optical coherence microscopy (OCM) is a low-coherence-based interferometry technology for cellular imaging with both high axial and lateral resolution. Using a high-numerical-aperture objective, OCM normally has a shallow depth of field and requires scanning the focus through the entire region of interest to perform volumetric imaging. With a higher-numerical-aperture objective, the image quality of OCM is affected by and more sensitive to aberrations. Interferometric synthetic aperture microscopy (ISAM) and computational adaptive optics (CAO) are computed imaging techniques that overcome the depth-of-field limitation and the effect of optical aberrations in optical coherence tomography (OCT), respectively. In this work we combine OCM with ISAM and CAO to achieve high-speed volumetric cellular imaging. Experimental imaging results of ex vivo human breast tissue, ex vivo mouse brain tissue, in vitro fibroblast cells in 3D scaffolds, and in vivo human skin demonstrate the significant potential of this technique for high-speed volumetric cellular imaging.OCIS codes: (170.4500) Optical coherence tomography, (090.1000) Aberration compensation, (180.3170) Interference microscopy, (170.6900) Three-dimensional microscopy, (110.1758) Computational imaging, (100.3200) Inverse scattering     

衍射增强技术应用于医学成像的实验研究

目的本实验通过各类样品的DEI成像，获得衬度及分辨率，并与常规吸收成像比较，对衍射增强成像（DEI）方法进行分析和评价。方法在北京高能物理研究所同步辐射装置（BSRF）形貌站（4W1A束线）上，取人及动物脏器在摇摆曲线上进行扫描，并与常规吸收像进行比较，通过显微放大法获得分辨率。结果不同组织之间存在不同的衍射增强，DEI均显示较好的衬度和分辨率，图像分辨率达到微米级，并且扫描位置对图像显示具有较大影响。结论DEI从相位衬度角度进行成像，极大地改进了成像质量，理论上来说是安全的，有望运用于临床。     

Achromatic miniature lens system for coherent Raman scattering microscopy

We discuss the design and performance of a miniature objective lens optimized for coherent Raman scattering microscopy. The packaged lens assembly has a numerical aperture of 0.51 in water and an outer diameter of 8 mm. The lens system exhibits minimum chromatic aberrations, and produces coherent Raman scattering images with sub-micrometer lateral resolution (0.648 μm) using near-infrared excitation pulses. We demonstrate that despite the small dimensions of the miniature objective, the performance of this lens system is comparable to standard microscope objective lenses, offering opportunities for miniaturizing coherent Raman scattering imaging probes without sacrificing the image quality.OCIS codes: (180.4315) Nonlinear microscopy, (350.3950) Micro-optics, (220.3630) Lenses     

3D resolved mapping of optical aberrations in thick tissues

We demonstrate a simple method for mapping optical aberrations with 3D resolution within thick samples. The method relies on the local measurement of the variation in image quality with externally applied aberrations. We discuss the accuracy of the method as a function of the signal strength and of the aberration amplitude and we derive the achievable resolution for the resulting measurements. We then report on measured 3D aberration maps in human skin biopsies and mouse brain slices. From these data, we analyse the consequences of tissue structure and refractive index distribution on aberrations and imaging depth in normal and cleared tissue samples. The aberration maps allow the estimation of the typical aplanetism region size over which aberrations can be uniformly corrected. This method and data pave the way towards efficient correction strategies for tissue imaging applications.     

Models Genesis

Transfer learning from natural images to medical images has been established as one of the most practical paradigms in deep learning for medical image analysis. To fit this paradigm, however, 3D imaging tasks in the most prominent imaging modalities (e.g., CT and MRI) have to be reformulated and solved in 2D, losing rich 3D anatomical information, thereby inevitably compromising its performance. To overcome this limitation, we have built a set of models, called Generic Autodidactic Models, nicknamed Models Genesis, because they are created ex nihilo (with no manual labeling), self-taught (learnt by self-supervision), and generic (served as source models for generating application-specific target models). Our extensive experiments demonstrate that our Models Genesis significantly outperform learning from scratch and existing pre-trained 3D models in all five target 3D applications covering both segmentation and classification. More importantly, learning a model from scratch simply in 3D may not necessarily yield performance better than transfer learning from ImageNet in 2D, but our Models Genesis consistently top any 2D/2.5D approaches including fine-tuning the models pre-trained from ImageNet as well as fine-tuning the 2D versions of our Models Genesis, confirming the importance of 3D anatomical information and significance of Models Genesis for 3D medical imaging. This performance is attributed to our unified self-supervised learning framework, built on a simple yet powerful observation: the sophisticated and recurrent anatomy in medical images can serve as strong yet free supervision signals for deep models to learn common anatomical representation automatically via self-supervision. As open science, all codes and pre-trained Models Genesis are available at https://github.com/MrGiovanni/ModelsGenesis.     

Correction of phase aberrations for sectored annular array ultrasound transducers

Two methods for correction of unknown phase aberrations induced by inhomogeneous acoustic velocities in tissues are explored for the two dimensional geometry of a sectored annular array system. The methods employed are adaptations of a cross correlation technique and a speckle brightness maximization technique. The methods correct phase distortions via the introduction of phase shifts in the timing sequence at the beamformer stage of a sectored annular array transducer. The techniques are investigated employing software models and a computer controlled automated scanning system. A 65-element sectored annular array is modelled via a rotating 5 element transducer. Tissue equivalent materials were moulded into a double layer aberrating medium to simulate phase distortions encountered in the rectus abdominis muscle in vivo. A comparison of the effectiveness of the two correction methods is presented. Contrast of an anechoic region is increased from 0.34 ± 0.08 to 0.48 ± 0.06 for the cross correlation technique, and up to 0.62 ± 0.05 for the speckle brightness maximization method. The performance of these correction techniques on target phantoms suggests that considerable improvements in image quality should be possible for clinical systems.     

Validity of Extended-Field-of-View Ultrasound Imaging to Evaluate Quantity and Quality of Trunk Skeletal Muscles

This study examined the validity of extended-field-of-view (EFOV) ultrasound imaging for evaluating the quantity (cross-sectional area [CSA]) and quality (accumulation of intramuscular fat) of trunk skeletal muscles (rectus abdominis, abdominal oblique and erector spinae) using magnetic resonance imaging (MRI) as a reference. Thirty healthy young men participated in this study. Cross-sectional images of the trunk at the height of the third lumbar vertebra were acquired and compared by EFOV ultrasound imaging and MRI. No significant differences were observed in CSAs between the two methods (0.74 ≤ R² ≤ 0.85). Echo intensities significantly correlated with MRI-derived accumulation of intramuscular fat in each skeletal muscle group. However, the correlation coefficients were relatively low (0.37 ≤ r ≤0.47; p < 0.05). These results indicate that EFOV ultrasound imaging is a reliable method for assessing trunk skeletal muscle CSA. Further research is warranted to find the optimal ultrasound setting for evaluating muscle quality.     

Automating Spine Curvature Measurement in Volumetric Ultrasound via Adaptive Phase Features

Ultrasound volume projection imaging (VPI) has been recently suggested. This novel imaging method allows a non-radiation assessment of spine deformity with free-hand 3-D ultrasound imaging. This paper presents a fully automatic method to evaluate the spine curve in VPI images corresponding to different projection depth of the volumetric ultrasound, thus making it possible to analyze 3-D spine deformity. The new automatic method is based on prior knowledge about the geometric arrangement of the spinous processes. The frequency bandwidth of log-Gabor filters is adaptively adjusted to calculate the oriented phase congruency, facilitating the segmentation of the spinous column profile. And the spine curvature angle is finally calculated according to the inflection points of the curve over the segmented spinous column profile. The performance of the automatic method is evaluated on spine VPI images among patients with different scoliotic angles. The curvature angles obtained using the proposed method have a high linear correlation with those by the manual method (r = 0.90, p < 0.001) and X-ray Cobb's method (r = 0.87, p < 0.001). The feasibility of 3-D spine deformity assessment is also demonstrated using VPI images corresponding to various projection depth. The results suggest that this method can substantially improve the recognition of the spinous column profile, especially facilitating the applications of 3-D spine deformity assessment.     

Mini gamma cameras for intra-operative nuclear tomographic reconstruction

Nuclear imaging modalities like PET or SPECT are in extensive use in medical diagnostics. In a move towards personalized therapy, we present a flexible nuclear tomographic imaging system to enable intra-operative SPECT-like 3D imaging. The system consists of a miniaturized gamma camera mounted on a robot arm for flexible positioning, while spatio-temporal localization is provided by an optical tracking system. To facilitate statistical tomographic reconstruction of the radiotracer distribution using a maximum likelihood approach, a precise model of the mini gamma camera is generated by measurements. The entire system is evaluated in a series of experiments using a hot spot phantom, with a focus on criteria relevant for the intra-operative workflow, namely the number of required imaging positions as well as the required imaging time. The results show that high quality reconstructed images of simple hot spot configurations with positional errors of less than one millimeter are possible within acquisition times as short as 15 s.     

Mini gamma cameras for intra-operative nuclear tomographic reconstruction

Nuclear imaging modalities like PET or SPECT are in extensive use in medical diagnostics. In a move towards personalized therapy, we present a flexible nuclear tomographic imaging system to enable intra-operative SPECT-like 3D imaging. The system consists of a miniaturized gamma camera mounted on a robot arm for flexible positioning, while spatio-temporal localization is provided by an optical tracking system. To facilitate statistical tomographic reconstruction of the radiotracer distribution using a maximum likelihood approach, a precise model of the mini gamma camera is generated by measurements. The entire system is evaluated in a series of experiments using a hot spot phantom, with a focus on criteria relevant for the intra-operative workflow, namely the number of required imaging positions as well as the required imaging time. The results show that high quality reconstructed images of simple hot spot configurations with positional errors of less than one millimeter are possible within acquisition times as short as 15 s.