Impacts of Color Space Transformations on Dysplastic Nuclei Segmentation Using CNN

Oral epithelial dysplasia is a common precancerous lesion type that can be graded as mild, moderate and severe. Although not all oral epithelial dysplasia become cancer over time, this premalignant condition has a significant rate of progressing to cancer and the early treatment has been shown to be considerably more successful. The diagnosis and distinctions between mild, moderate, and severe grades are made by pathologists through a complex and time-consuming process where some cytological features, including nuclear shape, are analysed. The use of computer-aided diagnosis can be applied as a tool to aid and enhance the pathologist decisions. Recently, deep learning based methods are earning more and more attention and have been successfully applied to nuclei segmentation problems in several scenarios. In this paper, we evaluated the impact of different color spaces transformations for automated nuclei segmentation on histological images of oral dysplastic tissues using fully convolutional neural networks (CNN). The CNN were trained using different color spaces from a dataset of tongue images from mice diagnosed with oral epithelial dysplasia. The CIE L*a*b* color space transformation achieved the best averaged accuracy over all analyzed color space configurations (88.2%). The results show that the chrominance information, or the color values, does not play the most significant role for nuclei segmentation purpose on a mice tongue histopathological images dataset.

Download Full-text

An Insight to Oral Epithelial Dysplasia

International Journal of Head and Neck Surgery ◽

10.5005/jp-journals-10001-1144 ◽

2013 ◽

Vol 4 (2) ◽

pp. 74-82 ◽

Cited By ~ 4

Author(s):

Varun Rastogi ◽

Naveen Puri ◽

Satyaranjan Mishra ◽

Swati Arora ◽

Geetpriya Kaur ◽

...

Keyword(s):

Histological Grade ◽

Precancerous Lesion ◽

Epithelial Dysplasia ◽

Oral Dysplasia ◽

Oral Epithelial Dysplasia ◽

Grading Systems ◽

Risk Of Progression ◽

Cytological Changes ◽

Oral Epithelial ◽

Head Neck

ABSTRACT Oral dysplasia is a potentially precancerous lesion diagnosed histologically. While the risk of progression is associated with histological grade, it is currently impossible to predict accurately which lesions will progress. Although most oral pathologists recognize and accept the criteria for grading epithelial dysplasia based on architectural and cytological changes, there can be considerable interexaminer and intraexaminer variation in the assessment of the presence or absence and the grade of oral epithelial dysplasia. This article reviews the alterations, criteria, different grading systems and the markers used for assessing the malignant transformation of epithelial dysplasia. How to cite this article Rastogi V, Puri N, Mishra S, Arora S, Kaur G, Yadav L. An Insight to Oral Epithelial Dysplasia. Int J Head Neck Surg 2013;4(2):74-82.

Download Full-text

AKURASI PENJEJAKAN OBJEK DALAM BERAGAM RUANG WARNA

Majalah Ilmiah Teknologi Elektro ◽

10.24843/mite.2015.v14i02p15 ◽

2015 ◽

Vol 14 (2) ◽

pp. 80

Author(s):

Gede Sukadarmika ◽

Dewa Made Wiharta ◽

Nyoman Putra Sastra

Keyword(s):

Color Space ◽

Color Histogram ◽

Numerical Comparison ◽

Color Spaces ◽

Object Position ◽

Actual Position ◽

Conducting Research ◽

Different Color ◽

The Impact

The object trace has been a problem in estimating an object position when the object is moving due to the heavy influence of the uncertainty. Many researcher claim that color histogram is reliable feature to represent this object. . Different investigators use different color spaces in conducting research on tracking the object. So, there is no numerical comparison of the impact of the use of different color spaces to the successful tracking. This study compare the performance of tracking an object by using a different color space i.e.: RGB, HSV, and CIELAB. The performance is shown numerically by comparing the actual position of the object with the results of the estimation.

Download Full-text

Color Space Transformation using Neural Networks

Color and Imaging Conference ◽

10.2352/issn.2169-2629.2019.27.29 ◽

2019 ◽

Vol 2019 (1) ◽

pp. 153-158

Author(s):

Lindsay MacDonald

Keyword(s):

Neural Network ◽

Neural Networks ◽

Color Space ◽

Reflectance Spectra ◽

Network Architectures ◽

Color Spaces ◽

Natural Materials ◽

Space Transformation ◽

Color Space Transformation

We investigated how well a multilayer neural network could implement the mapping between two trichromatic color spaces, specifically from camera R,G,B to tristimulus X,Y,Z. For training the network, a set of 800,000 synthetic reflectance spectra was generated. For testing the network, a set of 8,714 real reflectance spectra was collated from instrumental measurements on textiles, paints and natural materials. Various network architectures were tested, with both linear and sigmoidal activations. Results show that over 85% of all test samples had color errors of less than 1.0 ΔE2000 units, much more accurate than could be achieved by regression.

Download Full-text

SAVER: sodium valproate for the epigenetic reprogramming of high-risk oral epithelial dysplasia—a phase II randomised control trial study protocol

Trials ◽

10.1186/s13063-021-05373-8 ◽

2021 ◽

Vol 22 (1) ◽

Author(s):

Caroline McCarthy ◽

Joseph Sacco ◽

Stefano Fedele ◽

Michael Ho ◽

Stephen Porter ◽

...

Keyword(s):

High Risk ◽

Sodium Valproate ◽

Study Protocol ◽

Epigenetic Reprogramming ◽

Epithelial Dysplasia ◽

Phase Iii ◽

Cancer Development ◽

Randomised Control Trial ◽

Oral Epithelial Dysplasia ◽

Oral Epithelial

Abstract Background Sodium valproate (VPA) has been associated with a reduced risk of head and neck cancer development. The potential protective mechanism of action is believed to be via inhibition of histone deacetylase and subsequent epigenetic reprogramming. SAVER is a phase IIb open-label, randomised control trial of VPA as a chemopreventive agent in patients with high-risk oral epithelial dysplasia (OED). The aim of the trial is to gather preliminary evidence of the clinical and biological effects of VPA upon OED and assess the feasibility and acceptability of such a trial, with a view to inform a future definitive phase III study. Methods One hundred and ten patients with high-risk OED will be recruited from up to 10 secondary care sites in the UK and randomised into either VPA or observation only for 4 months. Women of childbearing potential will be excluded due to the teratogenic properties of VPA. Tissue and blood samples will be collected prior to randomisation and on the last day of the intervention/observation-only period (end of 4 months). Clinical measurement and additional safety bloods will be taken at multiple time points during the trial. The primary outcome will be a composite, surrogate endpoint of change in lesion size, change in grade of dysplasia and change in LOH profile at 8 key microsatellite regions. Feasibility outcomes will include recruitment targets, compliance with the study protocol and adverse effects. A qualitative sub-study will explore patient experience and perception of the trial. Discussion The current management options for patients with high-risk OED are limited and mostly include surgical resection and clinical surveillance. However, there remains little evidence whether surgery can effectively lead to a notable reduction in the risk of oral cancer development. Similarly, surveillance is associated with concerns regarding delayed diagnosis of OED progressing to malignancy. The SAVER trial provides an opportunity to investigate the effects of a repurposed, inexpensive and well-tolerated medication as a potential chemopreventive strategy for patients with high-risk OED. The clinical and biological findings of SAVER will inform the appropriateness, design and feasibility of a definitive phase III trial. Trial registration The trial is registered with the European Clinical Trials Database (Eudra-CT 2018-000197-30). (http://www.isrctn.com/ISRCTN12448611). The trial was prospectively registered on 24/04/2018.

Download Full-text