Jiyun Byun

Evaluation and benchmark for biological image segmentation

This paper describes ongoing work on creating a benchmarking and validation dataset for biological image segmentation. While the primary target is biological images, we believe that the dataset would be of help to researchers working in image segmentation and tracking in general. The motivation for creating this resource comes from the observation that while there are a large number of effective segmentation methods available in the research literature, it is difficult for the application scientists to make an informed choice as to what methods would work for her particular problem. No one single tool exists that is effective on a diverse set of application contexts and different methods have their own strengths and limitations. We describe below three different classes of data, ranging in scale from subcellular to cellular to tissue level images, each of which pose their own set of challenges to image analysis. Of particular value to the image processing researchers is that the data comes with associated ground truth information that can be used to evaluate the effectiveness of different methods. The analysis and evaluation are also integrated into a database framework that is available online at http://dough.ece.ucsb.edu.

Muller cell reactivity and photoreceptor cell death are reduced after experimental retinal detachment using an inhibitor of the Akt/mTOR pathway.

PURPOSE To test the effect of Palomid 529, an inhibitor of the Akt/mTOR pathway, on Müller cell proliferation, subretinal glial scar formation, and photoreceptor survival after experimental retinal detachment (RD). METHODS Palomid 529 (600 microg) in balanced salt solution or balanced salt solution alone was injected intravitreally immediately after RD into the right eyes of 12 rabbits. Ten micrograms of BrdU was injected intravitreally on day 3. Animals were killed on day 3 or 7, at which time retinal sections were labeled with anti-BrdU to detect dividing cells, with anti-vimentin to identify Müller cells, and with the isolectin B4 to identify microglia and macrophages. Outer nuclear layer (ONL) thickness was measured from fluorescence-labeled nuclear-stained sections. Labeling was imaged using confocal microscopy. Six additional animals received either drug or balanced salt solution injections into normal eyes, and paraffin sections were stained with hematoxylin and eosin. RESULTS In the drug-treated eyes there was a significant decrease in the number of anti-BrdU-labeled Müller cells, the number and size of subretinal scars, and the number of isolectin B4-labeled cells. The ONL was also significantly thicker, and there was no evidence of toxic effects. CONCLUSIONS Palomid 529 is an effective suppressor of Müller cell proliferation, glial scar formation, and photoreceptor cell death in a rabbit model of RD. This suggests that inhibiting the Akt/mTOR signal transduction pathway may be an effective strategy to decrease proliferation and photoreceptor cell death induced by detachment and perhaps represents a novel therapy for related human diseases such as proliferative vitreoretinopathy.

The effects of transient retinal detachment on cavity size and glial and neural remodeling in a mouse model of X-linked retinoschisis.

PURPOSE To determine the cellular consequences of retinal detachment in retinoschisin knockout (Rs1-KO) mice, a model for retinoschisin in humans. METHODS Experimental retinal detachments (RDs) were induced in the right eyes of both Rs1-KO and wild-type (wt) control mice. Immunocytochemistry was performed on retinal tissue at 1, 7, or 28 days after RD with antibodies to anti-GFAP, -neurofilament, and -rod opsin to examine cellular changes after detachment. Images of the immunostained tissue were captured by laser scanning confocal microscopy. Quantitative analysis was performed to measure the number of Hoechst-stained photoreceptor nuclei and their density, number, and size of inner retinal cavities, as well as the number of subretinal glial scars. RESULTS Since detachments were created with balanced salt solution, by examination, all retinas had spontaneously reattached by 1 day. Cellular responses common to many photoreceptor degenerations occurred in the nondetached retinas of Rs1-KO mice, and, of importance, RD did not appear to significantly accentuate these responses. The number of schisis cavities was not changed after detachment, but their size was reduced. CONCLUSIONS These data indicate that large short-term RD in Rs1-KO mice, followed by a period of reattachment may cause a slight increase in photoreceptor cell death, but detachments do not accentuate the gliosis and neurite sprouting already present and may in fact reduce the size of existing retinal cavities. This finding suggests that performing subretinal injections to deliver therapeutic agents may be a viable option in the treatment of patients with retinoschisis without causing significant cellular damage to the retina.

Quantitative analysis of immunofluorescent retinal images

We present a novel method to quantitatively analyze confocal microscope images of retinas. We automatically detect nuclei within the outer nuclear layer (ONL) in a retinal image. Based on nuclei detection results, we also automatically measure the thickness of the ONL and the local cell density within the ONL. These measurements provide the first thorough quantitative analysis of retinal images. Our results not only verify previous conclusions about retinal restructuring during detachment, but also provide biologists with significant information about the regional responses in the ONL

A Variational Approach to Exploit Prior Information in Object-Background Segregation: Application to Retinal Images

One of the main challenges in image segmentation is to adapt prior knowledge about the objects/regions that are likely to be present in an image, in order to obtain more precise detection and recognition. Typical applications of such knowledge-based segmentation include partitioning satellite images and microscopy images, where the context is generally well defined. In particular, we present an approach that exploits the knowledge about foreground and background information given in a reference image, in segmenting images containing similar objects or regions. This problem is presented within a variational framework, where cost functions based on pair-wise pixel similarities are minimized. This is perhaps one of the first attempts in using non-shape based prior information within a segmentation framework. We validate the proposed method to segment the outer nuclear layer (ONL) in retinal images. This approach successfully segments the ONL within an image and enables further quantitative analysis.

Towards automated bioimage analysis: from features to semantics

Recent advances in bio-molecular imaging have afforded biologists a more thorough understanding of cellular functions in complex tissue structures. For example, high resolution fluorescence images of the retina reveal details about tissue restructuring during detachment experiments. Time sequence imagery of microtubules provides insight into subcellular dynamics in response to cancer treatment drugs. However, technological progress is accompanied by a rapid proliferation of image data. Traditional analysis methods, namely manual measurements and qualitative assessments, become time consuming and are often nonreproducible. Computer vision tools can efficiently analyze these vast amounts of data with promising results. This paper provides an overview of several challenges faced in bioimage processing and our recent progress in addressing these issues