Figure 1 - uploaded by Kyohei Okubo
Content may be subject to copyright.
Endoscopic view and photographs of excised specimens of GIST. (A) (a-k) GIST can be seen as SMT protruding into the gastric lumen. (A) (l) SMT cannot be seen with an endoscope. (B) (a-l) Visible light photographs of specimens captured by a digital camera. GIST gastrointestinal tumor, SMT submucosal tumor.
Source publication
The diagnosis of gastrointestinal stromal tumor (GIST) using conventional endoscopy is difficult because submucosal tumor (SMT) lesions like GIST are covered by a mucosal layer. Near-infrared hyperspectral imaging (NIR-HSI) can obtain optical information from deep inside tissues. However, far less progress has been made in the development of techni...
Context in source publication
Context 1
... patients were screened, among whom 12 patients (10 men and 2 women) were enrolled and 2 were excluded (NIR-HSI images were acquired from the serosa side only). Their median age was 68 years (range 41-81 years). The median tumor size was 41 mm (range 24-80 mm). Endoscopic images of the lesions and pictures of excised 12 GIST specimens are shown in Fig. 1A,B. Figure 2A show pictures of 12 GIST specimens, captured by NIR-HSI. It can be seen that 7 specimens (a-d,f,g,j in Fig. 2A) were completely covered with mucosa, and 3 specimens (h,I,k in Fig On the basis of training data, analysis of the spectra in HSI images was performed using the SVM algorithm, and GIST and normal regions were ...
Similar publications
Context and objective. Gastrointestinal stromal tumors are rare and constitute a model of targetedtherapy in oncology. The objective of the present work was to describe the characteristics of gastrointestinal stromal tumors (GISTs). Methods. This was a descriptive retrospective case serial study on GISTs diagnosed in the Pathology Department of the...
Preclinical forms of Gastrointestinal stromal tumor (GIST), small asymptomatic lesions, called micro-GIST, are detected in ~30% of the general population. GIST driver mutation can be already detected in microGISTs, even if they do not progress into malignant cancer; these mutations are necessary, but insufficient events to foster tumor progression....
Introduction
Survival for gastrointestinal stromal tumor (GIST) has been increasing over the years after the introduction of tyrosine kinase inhibitors. However, the role of metastasectomy for GIST is still controversial. Patients are currently treated with imatinib or sunitinib in case of imatinib failures as optimal medical therapy for metastatic...
Background
In the phase 3 GRID trial, regorafenib improved progression-free survival (PFS) independent of KIT mutations in exons 9 and 11. In this retrospective, exploratory analysis of the GRID trial, we investigated whether a more comprehensive KIT mutation analysis could identify mutations that impact treatment outcome with regorafenib and a reg...
PurposeTo investigate the impact of delayed adjuvant imatinib on GIST patients with high risk of recurrence.Method
Adult GIST patients were retrospectively collected from our hospital between 2011 and 2018, and patients having high risk of recurrence were included for subsequent analyses. The primary endpoint was recurrence-free survival (RFS).Resu...
Citations
... Moreover, NIR light exhibits higher transparency through the living body than ultraviolet and visible light [8]. Because spectral information can be acquired from the deep parts of the body, NIR-HSI is expected to be applied in the medical field for the visualization of lesions hidden in normal tissue [9,10]. Therefore, many types of devices that can perform HSI according to the imaging target and situation have been developed. ...
In this study, we developed a rigid-scope system that can perform hyperspectral imaging (HSI) between visible and 1600 nm wavelengths using a supercontinuum light source and an acousto-optic tunable filter to emit specific wavelengths. The system optical performance was verified, and the classification ability was investigated. Consequently, it was demonstrated that HSI (490–1600 nm) could be performed. In addition, seven different targets could be classified by the neural network with an accuracy of 99.6%, recall of 93.7%, and specificity of 99.1% when the wavelength range of over 1000 nm (OTN) was extracted from HSI data as train data.
... The mesentery, representing fatty tissue, intentionally covers the nerve tissue, simulating challenging clinical observation conditions. For the data acquisition, an imaging system with a high-speed NIR hyperspectral camera (Compovision, CV-N800HS; Sumitomo Electric Industries, Ltd., Osaka, Japan) was used to obtain the NIR image with a wavelength range of 1000-2350 nm and a wavelength resolution of 6.3 nm [87]. ...
The Contrastive Learning for blind Hyperspectral Unmixing (CLHU) is a self-supervised deep learning approach for blind hyperspectral unmixing by jointly estimating the endmembers and fractional abundances. Unlike existing deep learning methods that rely on reconstruction capabilities, CLHU leverages the input-endmembers relationship for abundance estimation. Extensive experiments demonstrate CLHU’s effectiveness, achieving state-of-the-art performance in hyperspectral unmixing. Furthermore, the experimental findings indicate that endmember interactions are considered during the estimation process by the proposed method, particularly in non-linear problem contexts. This novel approach offers a promising perspective for the field and holds potential for further enhancements in hyperspectral unmixing tasks.
... In addition, NIR light is characterized by low-light loss owing to the absorption of water and light scattering in the living body and higher transparency through the living body than those achieved by ultraviolet and visible light [8]. Because spectral information can be acquired from the deep parts of the body, it is expected to be applied in the medical field for the visualization of lesions hidden in normal tissue [9,10]. ...
We propose a rigid-scope system that can perform hyperspectral imaging (HSI) between visible to over 1000 nm (OTN) using a supercontinuum light source and an acousto-optic tunable filter to emit specific wavelengths. In this study, the system was developed, and its optical performance verification and classification ability were investigated. Consequently, it was demonstrated that HSI (490–1600 nm) could be performed. In addition, seven different targets could be classified by the neural network with an accuracy of 99.6%, recall of 93.7%, and specificity of 99.1%, when the wavelength range of OTN was extracted from HSI data as train data.
... Although the NIR-HSI method with machine learning presented here has thus been applied only to ex vivo organs, the most significant advantage of this method is its potential for mapping the distribution of fatty acids with their characteristic chemical structures without slicing, homogenizing, or chemically staining organ samples (Fig. 3). HSI techniques reportedly have the potential for food quality screening 11 and detecting submucosal tumors 12 . In this study, we demonstrated that the distribution of molecules related to metabolism and pathophysiology in the liver could also be visualized in situ based on the chemical structure using NIR-HSI and machine learning with well-prepared training data. ...
Fatty acids play various physiological roles owing to their diverse structural characteristics, such as hydrocarbon chain length (HCL) and degree of saturation (DS). Although the distribution of fatty acids in biological tissues is associated with lipid metabolism, in situ imaging tools are still lacking for HCL and DS. Here, we introduce a framework of near-infrared (1000–1400 nm) hyperspectral label-free imaging with machine learning analysis of the fatty acid HCL and DS distribution in the liver at each pixel, in addition to the previously reported total lipid content. The training data of 16 typical fatty acids were obtained by gas chromatography from liver samples of mice fed with various diets. A two-dimensional mapping of these two parameters was successfully performed. Furthermore, the HCL/DS plot exhibited characteristic clustering among the different diet groups. Visualization of fatty acid distribution would provide insights for revealing the pathophysiological conditions of liver diseases and metabolism.
... The mesentery, representing fatty tissue, intentionally covers the nerve tissue, simulating challenging clinical observation conditions. For the data acquisition, an imaging system with a high-speed NIR hyperspectral camera (Compovision, CV-N800HS; Sumitomo Electric Industries, Ltd., Osaka, Japan) was used to obtain the NIR image with a wavelength range of 1000-2350 nm and a wavelength resolution of 6.3 nm [80]. By addressing this challenging test scenario, we aim to demonstrate the robustness and effectiveness of the CLHU framework in discriminating between these similar tissue types, offering potential benefits for medical applications. ...
p>The Contrastive Learning for blind Hyperspectral Unmixing (CLHU) is a self-supervised deep learning approach for blind hyperspectral unmixing. Unlike existing deep learning methods that rely on reconstruction capabilities, CLHU leverages the input-endmembers relationship for abundance estimation. The endmembers are dynamically updated during training, controlled by two regularization factors. Extensive experiments demonstrate CLHU's effectiveness, achieving state-of-the-art performance in hyperspectral unmixing. This novel approach offers a promising perspective for the field and holds potential for further enhancements in hyperspectral unmixing tasks.</p
... We previously reported that near-infrared (NIR) hyperspectral imaging (HSI) could be used to identify gastrointestinal stromal tumors in the muscular layer that was not exposed to the mucosa. 11,12 Light in the NIR region, ranging from 800 to 2500 nm, is useful for probing deep parts of the tissues owing to its low absorption and scattering. The absorption spectra in the NIR region convey fingerprint data owing to the overtone or combined vibrations of the chemical bonds. ...
Significance:
Determining the extent of gastric cancer (GC) is necessary for evaluating the gastrectomy margin for GC. Additionally, determining the extent of the GC that is not exposed to the mucosal surface remains difficult. However, near-infrared (NIR) can penetrate mucosal tissues highly efficiently.
Aim:
We investigated the ability of near-infrared hyperspectral imaging (NIR-HSI) to identify GC areas, including exposed and unexposed using surgical specimens, and explored the identifiable characteristics of the GC.
Approach:
Our study examined 10 patients with diagnosed GC who underwent surgery between 2020 and 2021. Specimen images were captured using NIR-HSI. For the specimens, the exposed area was defined as an area wherein the cancer was exposed on the surface, the unexposed area as an area wherein the cancer was present although the surface was covered by normal tissue, and the normal area as an area wherein the cancer was absent. We estimated the GC (including the exposed and unexposed areas) and normal areas using a support vector machine, which is a machine-learning method for classification. The prediction accuracy of the GC region in every area and normal region was evaluated. Additionally, the tumor thicknesses of the GC were pathologically measured, and their differences in identifiable and unidentifiable areas were compared using NIR-HSI.
Results:
The average prediction accuracy of the GC regions combined with both areas was 77.2%; with exposed and unexposed areas was 79.7% and 68.5%, respectively; and with normal regions was 79.7%. Additionally, the areas identified as cancerous had a tumor thickness of >2 mm.
Conclusions:
NIR-HSI identified the GC regions with high rates. As a feature, the exposed and unexposed areas with tumor thicknesses of >2 mm were identified using NIR-HSI.
... The mesentery, representing fatty tissue, intentionally covers the nerve tissue, simulating challenging clinical observation conditions. For the data acquisition, an imaging system with a high-speed NIR hyperspectral camera (Compovision, CV-N800HS; Sumitomo Electric Industries, Ltd., Osaka, Japan) was used to obtain the NIR image with a wavelength range of 1000-2350 nm and a wavelength resolution of 6.3 nm [80]. By addressing this challenging test scenario, we aim to demonstrate the robustness and effectiveness of the CLHU framework in discriminating between these similar tissue types, offering potential benefits for medical applications. ...
p>The Contrastive Learning for blind Hyperspectral Unmixing (CLHU) is a self-supervised deep learning approach for blind hyperspectral unmixing. Unlike existing deep learning methods that rely on reconstruction capabilities, CLHU leverages the input-endmembers relationship for abundance estimation. The endmembers are dynamically updated during training, controlled by two regularization factors. Extensive experiments demonstrate CLHU's effectiveness, achieving state-of-the-art performance in hyperspectral unmixing. This novel approach offers a promising perspective for the field and holds potential for further enhancements in hyperspectral unmixing tasks.</p
... The mesentery, representing fatty tissue, intentionally covers the nerve tissue, simulating challenging clinical observation conditions. For the data acquisition, an imaging system with a high-speed NIR hyperspectral camera (Compovision, CV-N800HS; Sumitomo Electric Industries, Ltd., Osaka, Japan) was used to obtain the NIR image with a wavelength range of 1000-2350 nm and a wavelength resolution of 6.3 nm [80]. By addressing this challenging test scenario, we aim to demonstrate the robustness and effectiveness of the CLHU framework in discriminating between these similar tissue types, offering potential benefits for medical applications. ...
p>The Contrastive Learning for blind Hyperspectral Unmixing (CLHU) is a self-supervised deep learning approach for blind hyperspectral unmixing. Unlike existing deep learning methods that rely on reconstruction capabilities, CLHU leverages the input-endmembers relationship for abundance estimation. The endmembers are dynamically updated during training, controlled by two regularization factors. Extensive experiments demonstrate CLHU's effectiveness, achieving state-of-the-art performance in hyperspectral unmixing. This novel approach offers a promising perspective for the field and holds potential for further enhancements in hyperspectral unmixing tasks.</p
... Meanwhile, Sato et al. employed a pushbroom NIR HS camera in the range 1000-2350 nm with 256 bands (Sato et al., 2020). During their study, 12 ex-vivo gastrointestinal stromal tumors were imaged. ...
... Liu et al., 2020). Sato et al., 2020 in their study proposed a novel approach utilizing NIR-HSI (near-infrared HSI) in detecting gastrointestinal stromal tumour (GIST). Hu et al., 2019 apply hyperspectral technology to the pathological diagnosis of gastric cancer tumours, developing a novel spectral-spatial CNN model reaching an accuracy of almost 98%. ...
... Therefore, calibration is critical and one of the first steps in pre-processing HS images. Generally, the acquired raw images are calibrated using white and dark reference images (Sato et al., 2020). Ravi et al., 2017, Manni et al., 2020, Q. Wang et al., 2020 ...
... (Camps-Valls & Bruzzone, 2005). Sato et al., 2020 propose an SVM (radial basis function [RBF] kernel) algorithm for gastrointestinal tumour HSIs, giving an accuracy of more than 85%. RBF SVM is also compared with other SVM classifiers in (Fabelo, Melián, et al., 2019), giving the best results for skin lesions diagnosis. ...
Hyperspectral Imaging (HSI) has grown to be one of the most crucial optical imaging modalities with applications in numerous industries. The non‐invasive nature of HSI has led to widening its horizon to the medical domain, especially in areas like early diagnosis of various diseases. HSI combines both imaging and spectroscopy properties, thereby exploiting spectral and spatial dimensions of images captured, providing quick and accurate interpretation of data. The current study aims to give an exhaustive overview of HSI's applications pertaining to the medical industry for fast detection of diseases and aiding in surgical procedures. The survey focuses on hyperspectral imaging combined with various approaches ‐ machine learning, deep learning, genetic algorithms, and anomaly detection for the treatment of disorders. In addition, the survey highlights accompanying pre‐processing approaches, performance metrics, inferences, and future prospects of HSI in the medical domain. The current study can gauge computer vision specialists, researchers in machine and deep learning domain, doctors, and scientists by giving them a platform for improving existing treatment methods for the betterment of society.
