Mark J. Gooding

Automatic segmentation of adherent biological cell boundaries and nuclei from brightfield microscopy images

The detection and segmentation of adherent eukaryotic cells from brightfield microscopy images represent challenging tasks in the image analysis field. This paper presents a free and open-source image analysis package which fully automates the tasks of cell detection, cell boundary segmentation, and nucleus segmentation in brightfield images. The package also performs image registration between brightfield and fluorescence images. The algorithms were evaluated on a variety of biological cell lines and compared against manual and fluorescence-based ground truths. When tested on HT1080 and HeLa cells, the cell detection step was able to correctly identify over 80% of cells, whilst the cell boundary segmentation step was able to segment over 75% of the cell body pixels, and the nucleus segmentation step was able to correctly identify nuclei in over 75% of the cells. The algorithms for cell detection and nucleus segmentation are novel to the field, whilst the cell boundary segmentation algorithm is contrast-invariant, which makes it more robust on these low-contrast images. Together, this suite of algorithms permit brightfield microscopy image processing without the need for additional fluorescence images. Finally our sephaCe application, which is available at http://www.sephace.com, provides a novel method for integrating these methods with any motorised microscope, thus facilitating the adoption of these techniques in biological research labs.

Advanced phase-based segmentation of multiple cells from brightfield microscopy images

Segmenting transparent phase objects, such as biological cells from brightfield microscope images, is a difficult problem due to the lack of observable intensity contrast and noise. Previous image analysis solutions have used excessive de- focusing or physical models to obtain the underlying phase properties. Here, an improved cell boundary detection algorithm is proposed to accurately segment multiple cells within the level set framework. This uses a novel speed term based on local phase and local orientation derived from the monogenic signal, which renders the algorithm invariant to intensity, making it ideal for these images. The new method can robustly handle noise and local minima, and distinguish touching cells. Validation is shown against manual expert segmentations.

PHASE-BASED SEGMENTATION OF CELLS FROM BRIGHTFIELD MICROSCOPY

Segmentation of transparent cells in brightfield microscopy images could facilitate the quantitative analysis of corresponding fluorescence images. However, this presents a challenge due to irregular morphology and weak intensity variation, particularly in ultra-thin regions. A boundary detection technique is applied to a series of variable focus images whereby a level set contour is initialised on a defocused image with improved intensity contrast, and subsequently evolved towards the correct boundary using images of improving focus. Local phase coherence is used to identify features within the images, driving contour evolution particularly in near-focus images which lack intensity contrast. Preliminary results demonstrate the effectiveness of this approach in segmenting the main cell body regions

A system for simultaneously measuring contact force, ultrasound, and position information for use in force-based correction of freehand scanning

During freehand ultrasound imaging, the sonographer places the ultrasound probe on the patients skin. This paper describes a system that simultaneously records the position of the probe, the contact force between the probe and skin, and the ultrasound image. The system consists of an ultrasound machine, a probe, a force sensor, an optical localizer, and a host computer. Two new calibration methods are demonstrated: a temporal calibration to determine the time delay between force and position measurements, and a gravitational calibration to remove the effect of gravity on the recorded force. Measurements made with the system showed good agreement with those obtained from a standard materials testing machine. The systems uses include three-dimensional (3-D) ultrasound imaging, force-based deformation correction of ultrasound images, and indentation testing.

Three-dimensional ultrasound imaging of mammary ducts in lactating women: a feasibility study.

Objective. The main function of the breast is to produce milk for offspring. As such, the ductal system, which carries milk from the milk‐secreting glands (alveoli) to the nipple, is central to the natural function of the breast. The ductal system is also the region in which many malignancies originate and spread. In this study, we aimed to assess the feasibility of manual mapping of ductal systems from 3‐dimensional (3D) ultrasound data and to evaluate the structures found with respect to conventional understanding of breast anatomy and physiology. Methods. Three‐dimensional ultrasound data of the breast were acquired using a mechanical system, which captures data in a conical shape covering most of the breast without excessive compression. Manual mapping of the ductal system was performed using custom software for data from 4 lactating volunteers. Results. Observational results are presented for ultrasound data from the 4 lactating volunteers. For all volunteers, only a small number of ductal structures were engorged with milk, suggesting that the lactiferous activity of the breast may be localized. These enlarged ducts were predominantly found in the inferior lateral quadrant of each breast. The observation was also made that the enlarged, milk‐storing parts of the duct were spread throughout the ductal system and not directly below the nipple as conventional anatomy suggests. Conclusions. Ultrasound visualization of the 3D structure of milk‐laden ducts in an uncompressed breast has been shown. Using manual tracing, it was possible to track milk‐laden ducts of diameters less than 1 mm.

Driver behavioural classification from trajectory data

In recent years, traffic video surveillance has increased significantly. However, most of the footage is reviewed by humans or not at all. Tools capable of analysing traffic video sequences and autonomously extracting information are required. This paper presents an analysis of two automatic methods for classifying driver behaviour using only data provided by vehicle trackers. The algorithms are tested on several simulated traffic situations and their performance is compared to human observers. Factor analysis is shown to outperform human observers. We believe this is the first time automatic behavioural clustering of drivers using trajectory information has been successfully demonstrated.

Volume Reconstruction from Sparse 3D Ultrasonography

3D freehand ultrasound has extensive application for organ volume measurement and has been shown to have better reproducibility than estimates of volume made from 2D measurement followed by interpolation to 3D. One key advantage of free-hand ultrasound is that of image compounding, but this advantage is lost in many automated reconstruction systems. A novel method is presented for the automated segmentation and surface reconstruction of organs from sparse 3D ultrasound data. Preliminary results are demonstrated for simulated data, and two cases of in-vivo data; breast ultrasound and imaging of ovarian follicles.

On the Use of Low-Pass Filters for Image Processing with Inverse Laplacian Models

A novel signal processing-oriented approach to solving problems involving inverse Laplacians is introduced. The Monogenic Signal is a powerful method of computing the phase of discrete signals in image data, however it is typically used with band-pass filters in the capacity of a feature detector. Substituting low-pass filters allows the Monogenic Signal to produce approximate solutions to the inverse Laplacian, with the added benefit of tunability and the generation of three equivariant properties (namely local energy, local phase and local orientation), which allow the development of powerful numerical solutions for a new set of problems. These principles are applied here in the context of biological cell segmentation from brightfield microscopy image data. The Monogenic Signal approach is used to generate reduced noise solutions to the Transport of Intensity Equation for optical phase recovery, and the resulting local phase and local orientation terms are combined in an iterative level set approach to accurately segment cell boundaries. Potential applications of this approach are discussed with respect to other fields.

Automatic mammary duct detection in 3d ultrasound

This paper presents a method for the initial detection of ductal structures within 3D ultrasound images using second-order shape measurements. The desire to detect ducts is motivated in a number of way, principally as step in the detection and assessment of ductal carcinoma in-situ. Detection is performed by measuring the variation of the local second-order shape from a prototype shape corresponding to a perfect tube. We believe this work is the first demonstration of the ability to detect sections of duct automatically in ultrasound images. The detection is performed with a view to employing vessel tracking method to delineate the full ductal structure.

OP33.07: Dependence of three‐dimensional ultrasound fetal cardiac volume measurement on orientation of probe with respect to the heart

The average examination lasted 5min (range 2–9min), regardless of the presence of a CHD. Conclusions: An examiner with mid-level experience can detect most relevant normal cardiac structures in optimally recorded fetal cardiac volumes. All 6 anomalies were detected and 4/6 diagnosed correctly, but too many healthy fetuses would have been referred for a repeat exam. The method used for this pilot study will now be applied to the scenario ‘volumes recorded by a mid-level operator, but evaluated by an expert’ to decide if STIC can be used for ‘second opinion’.