• Nenhum resultado encontrado

Novel Method for Detection of Gastric Cancer by Hyperspectral Image Technique

N/A
N/A
Protected

Academic year: 2017

Share "Novel Method for Detection of Gastric Cancer by Hyperspectral Image Technique"

Copied!
5
0
0

Texto

(1)

Copyright © 2014 IJECCE, All right reserved

Novel Method for Detection of Gastric Cancer by

Hyperspectral Image Technique

Preetha Nair, S. Ranjitha, Dr. H. N. Suresh

AbstractDuring the last few decades, many studies have been performed on the early detection of cancer.Gastric cancer causes about 8,00,00 deaths worldwide per year. The objective of the work is to develop a novel hyperspectral imaging algorithm for gastric cancer detection by the medical community, however, due to the amounts of data and the highly computational nature of the algorithms, running these complex algorithms can result computationally intensive. This paper, for the early detection of gastric cancer, proposes the analysis system of an endoscopic image of the stomach, which detects the abnormal region by using the change of color in the image and by providing the surface tissue information to the detector. While advanced inflammation and cancer may be easily detected, early inflammation and cancer are difficult to detect and requires more attention to be detected. This research proposes the implementation of the data bank that will provide the medical community with the necessary tools to develop novel algorithms for the early and accurate gastric cancer patents. Graphical processing unit provides the necessary computing power for the acceleration of the proposed algorithms in the data bank to their capability of running data parallel algorithms. The sensitivity, specificity and accuracy rates of the algorithm’s

diagnostic capability will be enhanced by image processing.

Keywords Biomedical Optical Imaging, Hyperspectral, Multispectral, Spectral Imaging, Spectroscopy.

I. I

NTRODUCTION

Imaging forms an essential part of cancer clinical protocols and is able to furnish morphological,structural, metabolic and functional information. Integration with other diagnostic tools such as in vitro tissue and fluids analysis assists in clinical decision-making. Hybrid imaging techniques are able to supply complementary information for improved staging and therapy planning. Image guided and targeted minimally invasive therapy has the promise to improve outcome and reduce collateral effects. Early detection of cancer through screening based on imaging is probably the major contributor to a reduction in mortality for certain cancers. Targeted imaging of receptors, gene therapy expression and cancer stem cells are research activities that will translate into clinical use in the next decade. Technological developments will increase imaging speed to match that of physiological processes. Targeted imaging and therapeutic agents will be developed in tandem through close collaboration between academia and biotechnology, information technology and pharmaceutical industries.

Several imaging modalities such as magnetic resonance imaging, computed tomography, ultrasconography, Doppler scanning and nuclear imaging have completely expanded medical imaging field. Modern research works showed that hyperspectral imaging (HI), has emerged as a new member of the family of the medical imaging modalities. HI provides a powerful tool for non-invasive

tissue analyses. This technology is able to capture both the spatial and spectral data of an organ or tissue in one snapshot. In other words, the imaging system produces many narrow band images at different wavelengths. Not similar to conventional three channel color cameras and other filter-based imaging systems, this system capture full neighboring spectral data with spectral and spatial information. Present work, HI will be used for detection of gastric cancer by using developed novel algorithm. Multiple biomedical imaging techniques are used in all phases of cancer management.

The concept of only using tumour volume as a measure of disease progression has been shown to be inadequate as it only can show a delayed response to therapy and no indication of metabolism and other parameters. This has led to the use of multiple imaging techniques in cancer management. The development of a hybrid imaging system such as PET/CT (Beyer et al., 2002) that combines the metabolic sensitivity of PET and the temporal and spatial resolution of CT.As a result there has been an increased use of imaging of biomarkers to demonstrate metabolism, cell proliferation, cell migration, receptor expression, gene expression, signal transduction, hypoxia, apoptosis, angiogenesis and vascular function. Measurements of these parameters can be used to plan therapy, to give early indications of treatment response and to detect drug resistance and disease recurrence. Imaging biomarkers are being developed for the selection of cancer patients most likely to respond to specific drugs and for the early detection of response to treatment with the aim of accelerating the measurement of endpoints. Examples are the replacement of patient survival and clinical endpoints with early measurement of responses such as glucose metabolism or DNA synthesis. With combined imaging systems such as PET/CT, SPECT/CT and in the future the combination of systems using for example PET and MR and ultrasound and MR, there will be a need to have standardization in order to follow longitudinal studies of response to therapy. Cancer is a multi-factorial disease and imaging needs to be able to demonstrate the various mechanisms and phases of pathogenesis.

(2)

Copyright © 2014 IJECCE, All right reserved

II. I

MAGE

C

ONTRAST

Imaging systems produce images that have differences in contrast. The differences in contrast can be due to changes in physical properties caused by the endogenous nature of the tissue or by the use of exogenous agents.

Endogenous mechanisms include:

a. radiation absorption, reflection and transmission, b. magnetic relaxivity

c. magnetic susceptibility d.water molecule diffusion e. magnetic spin tagging f. oxygenation

g. spectral distribution h. temperature i. electrical impedance j.acoustic frequency shifts k. mechanical elasticity

Exogenous mechanisms include:

a. radiation absorption, reflection and emission b. spin hyperpolarization

c. magnetic relaxivity d. magnetic susceptibility e. magnetization transfer f. saturation transfer g. isotope spectra h.fluorescence i. bioluminescence j. perfusion k.extracellular pH l. hypoxia

Diagnostic imaging agents introduced intravenously, intraarterially or via natural orifices will play an increasing role in cancer imaging. In particular new tracers for PET (Machulla et al., 2000; Eriksson et al., 2002) and nuclear medicine (Pappo et al., 2000; Xiaobing Tian et al., 2004) are leading the development of molecular imaging.C-based PET tracers can also be exogenous substances found in the human body. On the other hand,F-based PET tracers are often analogues of substances found in the human body. Nanotechnology-based agents will be developed during the next decade for MRI (Neuwalt et al., 2004; Schellenberger et al.,2002; Harishingani et al., 2003; Li et al., 2004; Kircher et al.

III. L

ITERATURE

S

URVEY

Despite both the incidence and mortality rates of gastric cancer showing decreasing trends, gastric cancer remains one of the most common causes of death by cancer worldwide [1, 2]. There are significant regional differences in gastric cancer onset, with Asian countries, including India, Japan and Korea, known to have a particularly high incidence rate compared to the Western countries. In some countries, following the introduction of a mass screening program that utilizes double-contrast barium radiography for early the detection of gastric cancer and alongside developments in endoscopic equipment and improved diagnostic capability, gastric cancer is now being detected more often in the

asymptomatic stages [3]. As a result, approximately 50% of the cases of gastric cancer currently treated in Asian countries are early stage disease [4]. In contrast, in Western countries, gastric cancer is often detected at an advanced stage and prognosis remains poor. Prognosis depends on the stage at which it is detected, and complete excision of the cancer is the only curative option. The excellent postoperative results for early gastric cancer, with a 5-year survival rate of over 90% in both Western and Asian countries, indicate just how important it is to detect the cancer at the earliest possible stage [5, 6]. Moreover, with the improved detection rate of early gastric cancer in Asian, more minimally invasive treatments have been investigated, and the use of endoscopic mucosal resection (EMR) has become widespread. This technique has the support of many endoscopists, including those in Western countries [7]. In addition, a new modality of endoscopic treatment, endoscopic submucosal dissection (ESD), has become commonly performed in facilities across India, helping to dramatically increase the number of early gastric cancer cases treated endoscopically [8].

In other hand hyper-spectral imaging (HI) is an emerging modality for detection of cancer and other applications. HI can be extended for UV rays to infrared wavelength regions. In fact, this imaging system produces several, narrow-band images at different wavelengths. Compared to using conventional colour cameras and other filter-based imaging systems, this system produces full neighboring spectral data with both spectral and spatial information [9]. In medical applications, HI data provide a powerful tool for non-invasive tissue analysis. HI has also been used to provide quantitative data regarding tissue oxygen saturation in patients with peripheral vascular disease [10] to detect ischemic regions of the intestine during surgery [11] to predict and follow healing in foot ulcers of diabetic patients [12] and to diagnose hemorrhagic shock [13]. Although HI has been used to discriminate between cancerous and normal tissue [14] this method was limited to the visible wavelength range and required the injection of fluorescent material. HI has also been evaluated for use in the cytologic diagnosis of cancer cells [15]. High-resolution HI microscopy was also evaluated for its ability to detect abnormalities in skin tissue using hematoxylin and eosin (H&E)-stained preparations of normal and abnormal skin, benign nevi, and melanomas [16]. However, both of these methods are limited to use at the cytological level and in the visible wavelength range, and they require additional procedures for sample preparation. It has been reported that HI has also been used to detect gastric tumors in human subjects [17]. The spectral signatures of the gastric cancer and noncancerous stomach tissue were created in infrared wavelengths.

IV. L

ITERATURE

G

AP

(3)

Copyright © 2014 IJECCE, All right reserved stages of disease. However, some bottlenecks have limited

its use for in vivo screening applications, most notably their huge temporal cost and poor spatial resolution. In addition, their sensitivity and accuracy need to be improved. For gastric cancer, because of the instinctive squirming of the human stomach and the noise caused by gastric juice, detecting the tumor range accurately is difficult under HI. To address this problem, this work will present an HI system based on hyper spectral and the corresponding classification algorithm based on sparse representation.

IV. P

ROJECT

F

LOW

C

HART

V. O

BJECTIVES

The proposed research objectives of this research work are:

 To conduct a thorough literature survey on gastric cancer and method of detections

 To conduct a previous research on cancer detection using hyperspectral imaging

 To collect the data about gastric cancer from the well known hospital

 To design an experiment to acquire the hyperspectral images of the gastric cancer (tumors) from the patients who suffering from gastric cancer.

 To find the suitable normalization and preprocessing techniques to normalized the spectra

 To build a library of algorithms used for processing hyuperspectral images used for cancer detection techniques.

VI. M

ETHODOLOGY

A. Fundamental of Tissue Optics:

In this task will be investigated the various interaction mechanisms between light and tissue. The work tries to

identify the following major categories of interaction that lead to alterations of it’s sue structureor composition: 

Photochemical:

Absorption of light by molecules

presents in or added to tissue. Photochemical interaction from the basis for photodynamic therapy

 Thermal: biological effect due to deposition of thermal energy in tissue.

Photoablative

: in the ultra – violet wavelength range photons posses sufficient energy to cause photo – dissociation of biopolymers and subsequent desorption of the fragments.

Electromechanical

: occurs at very high fluence rates where dieletric breakdown of tissue is induced which can lead to the formation of plasma. The rapid expansion of such plasma generates a shock wave which can rupture the tissue.

B. Experimental studies:

The spectral imaging endoscopic system introduces in this work will consist of a linear variable filter, a xenon light source, the light transfer cable, the hysteroscope and the CCD camera. A stepper motor, an encoder and linear motion system provide the tunability to the linear variable filter and a microcontroller, a USB to TTL interface controller and a stepper motor driver control the spectral imaging endoscopic system from a personal computer through dedicated firmware / software that will be developed.

C. Tasks 3. Hardware design:

The PCB design will be determined by the functionality of the endoscopic system and the future improvements. There will be no space limitations inside the device so the size is not an issue.

D. Development of algorithm:

The proposed algorithm (K-means Clustering) will be included the following tasks

 Identify the proportion or abundance of eachmaterial present in each pixel based on the spectral signature of material

 Be able to differentiate between healthy and cancerous tissue in ahyperspectral image.

 Serve to process these images.

 Classification and dimensionality reduction of hyperspectral images,

 Incorporate commonly used algorithms into robust library for Cancer Detectiontechniques.

E. Software development

Having completed the firmware, dedicated software will be developed to control the hyperspectraland enable the clinical evaluation of the system. The software will be developed using Visual C++ and it performs the following tasks:

 Controls the position of the filter and thus, the wavelength of the light source.

 Controls the camera exposure and the acquisition settings.

 Displays the images captured from the camera in two monitors.

 Keeps a simple database of the patients that will be examined.

(4)

Copyright © 2014 IJECCE, All right reserved  Performs automated time sequence image capture.

 Performs automated spectral cube acquisition over a specified time sequence.

D. Characterization and clinical validation (Spectral

band calibration)

Random patients will be examined with hyperspectral image in collaboration with the well known hospital. The patients will be referred for hysteroscopy after previous ultrasonic check and / or prior doctor advice. During the examination, only doctors and medically trained personnel will be present. The users of the hyperspectral will be trained previously ex-vivo. In all cases spectral cubes will be acquired from various areas of interest according to the physician’s opinion. Spectral imaging will provide in principle information for a much larger tissue area which is essential in the clinical practice. Integrated to endoscopy, this imaging modality will provides enhanced visualization of features of diagnostic importance such as neovascularization. It holds also the promise to provide spectral signatures of tissue lesions for enabling their in vivo identification and spectral mapping and grading for guiding treatment. The aforementioned technological deficiencies comprise a barrier for expanding spectral imaging and analysis to a number of medical specialties. Referring particularly to hysteroscopy employing a thin tip size endoscope, imaging spectroscopy would provide a valuable tool for assisting the diagnosis of endometrial neoplasias and cancers for intensifying tissue abnormalities that would be causal factors for infertility. From the acquired spectral images a full spectrum will be calculated for every image pixel, spanning both visible and HI spectral ranges. The developed system will be used for performing spectral analysis and spectral mapping of the endometrium and first clinical findings will be investigated.

VII. R

ESEARCH

O

UTCOME

Hyperspectral imaging allows surgeons to less invasively examine a vast area without actually touching or removing tissue. A merit of this techniques is the ability to both spatially and spectrally determine the difference among variant tissues or organs in surgery. The image-processing algorithms can incorporate detailed classification procedures that would be used for region extraction and identification of organs or tissues. Utilizing this technology in surgery will allow a novel exploration of anatomy and pathology and may offer hope as new tool for detection of tissue abnormalities.

IX. C

ONCLUSION

Multiple biomedical imaging techniques are used in all phases of cancer management. Imaging forms an essential part of cancer clinical protocols and is able to furnish morphological, structural, metabolic and functional information. Integration with other diagnostic tools such as in vitro tissue and fluids analysis assists in clinical decision-making. Hybrid imaging techniques are able to supply complementary information for improved staging

and therapy planning. Image guided and targeted minimally invasive therapy has the promise to improve outcome and reduce collateral effects. Early detection of cancer through screening based on imaging is probably the major contributor to a reduction in mortality for certain cancers. Targeted imaging of receptors, gene therapy expression and cancer stem cells are research activities that will translate into clinical use in the next decade. Technological developments will increase imaging speed to match that of physiological processes. Targeted imaging and therapeutic agents will be developed in tandem through close collaboration between academia and biotechnology, information technology and pharmaceutical industries.

R

EFERENCES

[1] Jemal, R. Siegel, J. Xu, and E. Ward, “Cancer statistics, 2010,” CA Cancer Journal for Clinicians, vol. 60, no. 5, pp. 277–300, 2010.

[2] J. Ferlay, H. R. Shin, F. Bray, D. Forman, C. Mathers, and D. M. Parkin, “Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008,” International Journal of Cancer, vol. 127, no. 12, pp. 2893–2917, 2010.

[3] C. Hamashima, D. Shibuya, H. Yamazaki et al., “The Japanese guidelines for gastric cancer screening,” Japanese Journal of Clinical Oncology, vol. 38, no. 4, pp. 259–267, 2008.

[4] K. Maruyama, M. Kaminishi, K. I. Hayashi et al., “Gastric cancer treated in 1991 in Japan: data analysis of nationwide registry,” Gastric Cancer, vol. 9, no. 2, pp. 51–66, 2006. [5] H. Ono, H. Kondo, T. Gotoda, et al., “Endoscopic mucosal

resection for treatment of early gastric cancer,” Gut, vol. 48, no. 2, pp. 225–229, 2001.

[6] F. J. Oliveira, H. Ferrao, E. Furtado, H. Batista, and L. Conceicao, “Early gastric cancer: report of 58 cases,” Gastric Cancer, vol. 1, no. 1, pp. 51–56, 1998.

[7] R. Soetikno, T. Kaltenbach, R. Yeh, and T. Gotoda, “Endoscopic mucosal resection for early cancers of the upper gastrointestinal tract,” Journal of Clinical Oncology, vol. 23, no. 20, pp. 4490– 4498, 2005.

[8] K. Hotta, T. Oyama, T. Akamatsu et al., “A comparison of outcomes of Endoscopic Submucosal Dissection (ESD) for early gastric neoplasms between high-volume and low-volume centers: multi-center retrospective questionnaire study conducted by the Nagano ESD Study Group,” Internal Medicine, vol. 49, no. 4, pp. 253–259, 2010.

[9] H. Akbari et al., “Hyperspectral image segmentation and its application in abdominal surgery,” Int. J. Funct. Inform. Personal. Med. 2(2), 201–216 (2009).

[10] D. C. Kellicut et al., “Emerging technology: hyperspectral imaging,” Perspect. Vasc. Surg. Endovasc. Ther. 16(1), 53–57 (2004).

[11] H. Akbari et al., “Detection and analysis of the intestinal ischemia using visible and invisible hyperspectral imaging,” IEEE Trans. Biomed. Eng. 57(8), 2011–2017 (2010).

[12] L. Khaodhiar et al., “The use of medical hyperspectral technology to evaluate microcirculatory chaves in diabetic foot ulcers and to predict clinical outcomes,” Diabetes Care 30(4), 903–910 (2007).

[13] L. C. Cancio et al., “Hyperspectral imaging: a new approach to the diagnosis of hemorrhagic shock,” J. Trauma 60(5), 1087– 1095 (2006).

[14] A. Jemal et al., “Cancer statistics, 2010,” CA Cancer J. Clin. 60(5), 277–300 (2010).

[15] M. E. Martin et al., “Development of an advanced hyperspectral imaging (HSI) system with applications for cancer detection,” Ann. Biomed. Eng. 34(6), 1061–1068 (2006).

(5)

Copyright © 2014 IJECCE, All right reserved [17] D. T. Dicker et al., “Differentiation of normal skin and

melanoma using high resolution hyperspectral imaging,” Cancer Biol. Ther. 5(8),1033–1038 (2006).

A

UTHOR

S

P

ROFILE

Preetha Nair

at present She is working as a Research scholor, Dept. Of ECE, Jain University,Bangalore, Karnataka,. She is good exposure in the field of signal processing, mage processing, etc. She got more 5 years of Teaching experience in various field. She is a member of IEEE, ISTE, MAPS.

Emil: prnvdr@yahoo.co.in

Dr. H.N. Suresh, Professor

Research coordinator of BIT, Bangalore Institute of Technology, Dept. of electronics and Instrumentation Technology, VV Puram, Bangalore–04, Karnataka, He received his BE (E&C) from P.E.S College of Engineering, Mysore University, Karnataka, India, in the year 1989 and completed his M.Tech (Bio Medical Instrumentation) from SJCE Mysore affiliated to University of Mysore., in the year of 1996 and since then he is actively involved in teaching and research and has Twenty six years of experience in teaching. He obtained his PhD (ECE) from Anna university of Technology. He worked at various capacities in affiliated University Engineering Colleges. For Visveswaraya Technical University and Bangalore University he worked as a Chairman for Board of Examiners, and member of Board of Studies etc. At present he is working as Professor in Bangalore Institute of Technology, Bangalore Affiliated to Visveswaraya Technical University. He has good exposure in the field of signal processing, Wavelet Transforms, Neural Networks, Pattern recognition, Bio Medical Signal Processing, Networking and Adaptive Neural network systems. He has published more 25 research papers in the refereed international journals and presented contributed research papers in refereed international and national conferences. He is a member of IEEE, Bio Medical Society of India, ISTE, IMAPS & Fellow member of IETE.

Referências

Documentos relacionados

Methods : This paper presents a novel approach (avoiding pre-processing, e.g. segmentation and iltering) for reticular pattern detection in dermoscopic images, using texture

This paper presents a set of measurements for spectral, radiometric and spatial camera characterization to be applied to imaging systems operating in the thermal infrared.. A

Using image processing techniques suitable for hyperspectral image analysis (Spectral Angle Mapper, MNF, End-member matching), Hyperion data over Udaipur, Dungarpur

We examined soil surface colour change to green and hydrotaxis following addition of water to biological soil crusts using pigment extraction, hyperspectral imaging, microsensors and

Several spectral, altimetric, and planimetric spatial data sets were used for feature extraction and quality assessment, namely: a pansharpened QuickBird image, its multispectral

The objective of this study was to investigate the prevalence of ‘Candidatus Mycoplasma haemobos’ by using a PCR-based detection method and correlate this with dairy herd

Imaging spectroscopy (IS), or hyperspectral remote sensing (HRS), is an advanced tool that provides high spectral resolution data in an image, with the aim of

Purpose: To develop a new three-dimensional (3D) spectral-domain optical coherence tomography (SD-OCT) data analysis method using a machine learning technique based on