1
|
Wang J, Du J, Tao C, Qi M, Yan J, Hu B, Zhang Z. Classification of Benign-Malignant Thyroid Nodules Based on Hyperspectral Technology. SENSORS (BASEL, SWITZERLAND) 2024; 24:3197. [PMID: 38794051 PMCID: PMC11126106 DOI: 10.3390/s24103197] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Received: 02/15/2024] [Revised: 05/14/2024] [Accepted: 05/15/2024] [Indexed: 05/26/2024]
Abstract
In recent years, the incidence of thyroid cancer has rapidly increased. To address the issue of the inefficient diagnosis of thyroid cancer during surgery, we propose a rapid method for the diagnosis of benign and malignant thyroid nodules based on hyperspectral technology. Firstly, using our self-developed thyroid nodule hyperspectral acquisition system, data for a large number of diverse thyroid nodule samples were obtained, providing a foundation for subsequent diagnosis. Secondly, to better meet clinical practical needs, we address the current situation of medical hyperspectral image classification research being mainly focused on pixel-based region segmentation, by proposing a method for nodule classification as benign or malignant based on thyroid nodule hyperspectral data blocks. Using 3D CNN and VGG16 networks as a basis, we designed a neural network algorithm (V3Dnet) for classification based on three-dimensional hyperspectral data blocks. In the case of a dataset with a block size of 50 × 50 × 196, the classification accuracy for benign and malignant samples reaches 84.63%. We also investigated the impact of data block size on the classification performance and constructed a classification model that includes thyroid nodule sample acquisition, hyperspectral data preprocessing, and an algorithm for thyroid nodule classification as benign and malignant based on hyperspectral data blocks. The proposed model for thyroid nodule classification is expected to be applied in thyroid surgery, thereby improving surgical accuracy and providing strong support for scientific research in related fields.
Collapse
Affiliation(s)
- Junjie Wang
- Xi’an Institute of Optics and Precision Mechanics, Chinese Academy of Sciences, Xi’an 710119, China; (J.W.); (J.D.); (C.T.); (M.Q.); (J.Y.)
- University of Chinese Academy of Sciences, Beijing 100049, China
- Key Laboratory of Biomedical Spectroscopy of Xi’an, Xi’an 710119, China
| | - Jian Du
- Xi’an Institute of Optics and Precision Mechanics, Chinese Academy of Sciences, Xi’an 710119, China; (J.W.); (J.D.); (C.T.); (M.Q.); (J.Y.)
- Key Laboratory of Biomedical Spectroscopy of Xi’an, Xi’an 710119, China
| | - Chenglong Tao
- Xi’an Institute of Optics and Precision Mechanics, Chinese Academy of Sciences, Xi’an 710119, China; (J.W.); (J.D.); (C.T.); (M.Q.); (J.Y.)
- Key Laboratory of Biomedical Spectroscopy of Xi’an, Xi’an 710119, China
| | - Meijie Qi
- Xi’an Institute of Optics and Precision Mechanics, Chinese Academy of Sciences, Xi’an 710119, China; (J.W.); (J.D.); (C.T.); (M.Q.); (J.Y.)
- Key Laboratory of Biomedical Spectroscopy of Xi’an, Xi’an 710119, China
| | - Jiayue Yan
- Xi’an Institute of Optics and Precision Mechanics, Chinese Academy of Sciences, Xi’an 710119, China; (J.W.); (J.D.); (C.T.); (M.Q.); (J.Y.)
- University of Chinese Academy of Sciences, Beijing 100049, China
- Key Laboratory of Biomedical Spectroscopy of Xi’an, Xi’an 710119, China
| | - Bingliang Hu
- Xi’an Institute of Optics and Precision Mechanics, Chinese Academy of Sciences, Xi’an 710119, China; (J.W.); (J.D.); (C.T.); (M.Q.); (J.Y.)
- Key Laboratory of Biomedical Spectroscopy of Xi’an, Xi’an 710119, China
| | - Zhoufeng Zhang
- Xi’an Institute of Optics and Precision Mechanics, Chinese Academy of Sciences, Xi’an 710119, China; (J.W.); (J.D.); (C.T.); (M.Q.); (J.Y.)
- Key Laboratory of Biomedical Spectroscopy of Xi’an, Xi’an 710119, China
| |
Collapse
|
2
|
Nikolaev AV, Fang Y, Essers J, Panth KM, Ambagtsheer G, Clahsen-van Groningen MC, Minnee RC, van Soest G, de Bruin RW. Pre-transplant kidney quality evaluation using photoacoustic imaging during normothermic machine perfusion. PHOTOACOUSTICS 2024; 36:100596. [PMID: 38379853 PMCID: PMC10877941 DOI: 10.1016/j.pacs.2024.100596] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Subscribe] [Scholar Register] [Received: 09/20/2023] [Revised: 02/03/2024] [Accepted: 02/08/2024] [Indexed: 02/22/2024]
Abstract
Due to the shortage of kidneys donated for transplantation, surgeons are forced to use the organs with an elevated risk of poor function or even failure. Although the existing methods for pre-transplant quality evaluation have been validated over decades in population cohort studies across the world, new methods are needed as long as delayed graft function or failure in a kidney transplant occurs. In this study, we explored the potential of utilizing photoacoustic (PA) imaging during normothermic machine perfusion (NMP) as a means of evaluating kidney quality. We closely monitored twenty-two porcine kidneys using 3D PA imaging during a two-hour NMP session. Based on biochemical analyses of perfusate and produced urine, the kidneys were categorized into 'non-functional' and 'functional' groups. Our primary focus was to quantify oxygenation (sO2) within the kidney cortical layer of depths 2 mm, 4 mm, and 6 mm using two-wavelength PA imaging. Next, receiver operating characteristic (ROC) analysis was performed to determine an optimal cortical layer depth and time point for the quantification of sO2 to discriminate between functional and non-functional organs. Finally, for each depth, we assessed the correlation between sO2 and creatinine clearance (CrCl), oxygen consumption (VO2), and renal blood flow (RBF). We found that hypoxia of the renal cortex is associated with poor renal function. In addition, the determination of sO2 within the 2 mm depth of the renal cortex after 30 min of NMP effectively distinguishes between functional and non-functional kidneys. The non-functional kidneys can be detected with the sensitivity and specificity of 80% and 85% respectively, using the cut-off point of sO2 < 39%. Oxygenation significantly correlates with RBF and VO2 in all kidneys. In functional kidneys, sO2 correlated with CrCl, which is not the case for non-functional kidneys. We conclude that the presented technique has a high potential for supporting organ selection for kidney transplantation.
Collapse
Affiliation(s)
- Anton V. Nikolaev
- Erasmus MC, Cardiovascular Institute, Thorax Center, Department of Cardiology, Dr. Molewaterplein 40, 3015 GD, Rotterdam, the Netherlands
| | - Yitian Fang
- Erasmus MC Transplant Institute, Department of Surgery, Division of HPB and Transplant Surgery, Erasmus Medical Center, Rotterdam, Dr. Molewaterplein 40, 3015 GD, Rotterdam, the Netherlands
| | - Jeroen Essers
- Department of Molecular Genetics, Erasmus Medical Center, Dr. Molewaterplein 40, 3015 GD, Rotterdam, the Netherlands
- Department of Radiotherapy, Erasmus Medical Center, Dr. Molewaterplein 40, 3015 GD, Rotterdam, the Netherlands
- Department of Vascular Surgery, Erasmus Medical Center, Dr. Molewaterplein 40, 3015 GD, Rotterdam, the Netherlands
| | - Kranthi M. Panth
- Department of Molecular Genetics, Erasmus Medical Center, Dr. Molewaterplein 40, 3015 GD, Rotterdam, the Netherlands
| | - Gisela Ambagtsheer
- Erasmus MC Transplant Institute, Department of Surgery, Division of HPB and Transplant Surgery, Erasmus Medical Center, Rotterdam, Dr. Molewaterplein 40, 3015 GD, Rotterdam, the Netherlands
| | - Marian C. Clahsen-van Groningen
- Department of Pathology and Clinical Bioinformatics, Erasmus Medical Center, Dr. Molewaterplein 40, 3015 GD, Rotterdam, the Netherlands
| | - Robert C. Minnee
- Erasmus MC Transplant Institute, Department of Surgery, Division of HPB and Transplant Surgery, Erasmus Medical Center, Rotterdam, Dr. Molewaterplein 40, 3015 GD, Rotterdam, the Netherlands
| | - Gijs van Soest
- Erasmus MC, Cardiovascular Institute, Thorax Center, Department of Cardiology, Dr. Molewaterplein 40, 3015 GD, Rotterdam, the Netherlands
- Department of Precision and Microsystems Engineering, Faculty of Mechanical Engineering, Delft University of Technology, Van Mourilk Broekmanweg 6, 2628 XE, Delft, the Netherlands
- Wellman Center for Photomedicine, Massachusetts General Hospital, 40 Blossom Street, Boston, MA 02114, USA
| | - Ron W.F. de Bruin
- Erasmus MC Transplant Institute, Department of Surgery, Division of HPB and Transplant Surgery, Erasmus Medical Center, Rotterdam, Dr. Molewaterplein 40, 3015 GD, Rotterdam, the Netherlands
| |
Collapse
|
3
|
Cui R, Yu H, Xu T, Xing X, Cao X, Yan K, Chen J. Deep Learning in Medical Hyperspectral Images: A Review. SENSORS (BASEL, SWITZERLAND) 2022; 22:s22249790. [PMID: 36560157 PMCID: PMC9784550 DOI: 10.3390/s22249790] [Citation(s) in RCA: 9] [Impact Index Per Article: 4.5] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Received: 11/07/2022] [Revised: 12/11/2022] [Accepted: 12/12/2022] [Indexed: 06/13/2023]
Abstract
With the continuous progress of development, deep learning has made good progress in the analysis and recognition of images, which has also triggered some researchers to explore the area of combining deep learning with hyperspectral medical images and achieve some progress. This paper introduces the principles and techniques of hyperspectral imaging systems, summarizes the common medical hyperspectral imaging systems, and summarizes the progress of some emerging spectral imaging systems through analyzing the literature. In particular, this article introduces the more frequently used medical hyperspectral images and the pre-processing techniques of the spectra, and in other sections, it discusses the main developments of medical hyperspectral combined with deep learning for disease diagnosis. On the basis of the previous review, tne limited factors in the study on the application of deep learning to hyperspectral medical images are outlined, promising research directions are summarized, and the future research prospects are provided for subsequent scholars.
Collapse
Affiliation(s)
- Rong Cui
- College of Electronic and Information Engineering, Changchun University, Changchun 130022, China
| | - He Yu
- College of Electronic and Information Engineering, Changchun University, Changchun 130022, China
- Jilin Provincial Key Laboratory of Human Health Status Identification and Function Enhancement, Changchun University, Changchun 130022, China
| | - Tingfa Xu
- Image Engineering & Video Technology Lab, School of Optics and Photonics, Beijing Institute of Technology, Beijing 100081, China
- Beijing Institute of Technology Chongqing Innovation Center, Chongqing 401120, China
| | - Xiaoxue Xing
- College of Electronic and Information Engineering, Changchun University, Changchun 130022, China
- Jilin Provincial Key Laboratory of Human Health Status Identification and Function Enhancement, Changchun University, Changchun 130022, China
| | - Xiaorui Cao
- College of Electronic and Information Engineering, Changchun University, Changchun 130022, China
| | - Kang Yan
- College of Electronic and Information Engineering, Changchun University, Changchun 130022, China
| | - Jiexi Chen
- College of Electronic and Information Engineering, Changchun University, Changchun 130022, China
| |
Collapse
|
4
|
Preoperative Function Assessment of Ex Vivo Kidneys with Supervised Machine Learning Based on Blood and Urine Markers Measured during Normothermic Machine Perfusion. Biomedicines 2022; 10:biomedicines10123055. [PMID: 36551812 PMCID: PMC9776285 DOI: 10.3390/biomedicines10123055] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/15/2022] [Revised: 11/13/2022] [Accepted: 11/23/2022] [Indexed: 11/29/2022] Open
Abstract
Establishing an objective quality assessment of an organ prior to transplantation can help prevent unnecessary discard of the organ and reduce the probability of functional failure. In this regard, normothermic machine perfusion (NMP) offers new possibilities for organ evaluation. However, to date, few studies have addressed the identification of markers and analytical tools to determine graft quality. In this study, function and injury markers were measured in blood and urine during NMP of 26 porcine kidneys and correlated with ex vivo inulin clearance behavior. Significant differentiation of kidneys according to their function could be achieved by oxygen consumption, oxygen delivery, renal blood flow, arterial pressure, intrarenal resistance, kidney temperature, relative urea concentration, and urine production. In addition, classifications were accomplished with supervised learning methods and histological analysis to predict renal function ex vivo. Classificators (support vector machines, k-nearest-neighbor, logistic regression and naive bayes) based on relevant markers in urine and blood achieved 75% and 83% accuracy in the validation and test set, respectively. A correlation between histological damage and function could not be detected. The measurement of blood and urine markers provides information of preoperative renal quality, which can used in future to establish an objective quality assessment.
Collapse
|
5
|
Martinez-Vega B, Tkachenko M, Matkabi M, Ortega S, Fabelo H, Balea-Fernandez F, La Salvia M, Torti E, Leporati F, Callico GM, Chalopin C. Evaluation of Preprocessing Methods on Independent Medical Hyperspectral Databases to Improve Analysis. SENSORS (BASEL, SWITZERLAND) 2022; 22:8917. [PMID: 36433516 PMCID: PMC9693077 DOI: 10.3390/s22228917] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Figures] [Subscribe] [Scholar Register] [Received: 10/11/2022] [Revised: 11/10/2022] [Accepted: 11/15/2022] [Indexed: 06/16/2023]
Abstract
Currently, one of the most common causes of death worldwide is cancer. The development of innovative methods to support the early and accurate detection of cancers is required to increase the recovery rate of patients. Several studies have shown that medical Hyperspectral Imaging (HSI) combined with artificial intelligence algorithms is a powerful tool for cancer detection. Various preprocessing methods are commonly applied to hyperspectral data to improve the performance of the algorithms. However, there is currently no standard for these methods, and no studies have compared them so far in the medical field. In this work, we evaluated different combinations of preprocessing steps, including spatial and spectral smoothing, Min-Max scaling, Standard Normal Variate normalization, and a median spatial smoothing technique, with the goal of improving tumor detection in three different HSI databases concerning colorectal, esophagogastric, and brain cancers. Two machine learning and deep learning models were used to perform the pixel-wise classification. The results showed that the choice of preprocessing method affects the performance of tumor identification. The method that showed slightly better results with respect to identifing colorectal tumors was Median Filter preprocessing (0.94 of area under the curve). On the other hand, esophagogastric and brain tumors were more accurately identified using Min-Max scaling preprocessing (0.93 and 0.92 of area under the curve, respectively). However, it is observed that the Median Filter method smooths sharp spectral features, resulting in high variability in the classification performance. Therefore, based on these results, obtained with different databases acquired by different HSI instrumentation, the most relevant preprocessing technique identified in this work is Min-Max scaling.
Collapse
Affiliation(s)
- Beatriz Martinez-Vega
- Research Institute for Applied Microelectronics (IUMA), University of Las Palmas de Gran Canaria, 35017 Las Palmas de Gran Canaria, Spain
| | - Mariia Tkachenko
- Innovation Center Computer-Assisted Surgery (ICCAS), University of Leipzig, 04103 Leipzig, Germany
- Center for Scalable Data Analytics and Artificial Intelligence (ScaDS.AI), University of Leipzig, 04105 Leipzig, Germany
| | - Marianne Matkabi
- Innovation Center Computer-Assisted Surgery (ICCAS), University of Leipzig, 04103 Leipzig, Germany
- Department of Electrical Engineering, Mechanical Engineering and Industrial Engineering, Anhalt University of Applied Science Anhalt, 06366 Köthen, Germany
| | - Samuel Ortega
- Research Institute for Applied Microelectronics (IUMA), University of Las Palmas de Gran Canaria, 35017 Las Palmas de Gran Canaria, Spain
- Nofima, Norwegian Institute of Food Fisheries and Aquaculture Research, NO-9291 Tromsø, Norway
| | - Himar Fabelo
- Research Institute for Applied Microelectronics (IUMA), University of Las Palmas de Gran Canaria, 35017 Las Palmas de Gran Canaria, Spain
- Fundacion Canaria Instituto de Investigación Sanitaria de Canarias (FIISC), 35019 Las Palmas de Gran Canaria, Spain
| | - Francisco Balea-Fernandez
- Research Institute for Applied Microelectronics (IUMA), University of Las Palmas de Gran Canaria, 35017 Las Palmas de Gran Canaria, Spain
- Department of Psychology, Sociology and Social Work, University of Las Palmas de Gran Canaria, 35017 Las Palmas de Gran Canaria, Spain
| | - Marco La Salvia
- Department of Electrical, Computer and Biomedical Engineering, University of Pavia, I-27100 Pavia, Italy
| | - Emanuele Torti
- Department of Electrical, Computer and Biomedical Engineering, University of Pavia, I-27100 Pavia, Italy
| | - Francesco Leporati
- Department of Electrical, Computer and Biomedical Engineering, University of Pavia, I-27100 Pavia, Italy
| | - Gustavo M. Callico
- Research Institute for Applied Microelectronics (IUMA), University of Las Palmas de Gran Canaria, 35017 Las Palmas de Gran Canaria, Spain
| | - Claire Chalopin
- Innovation Center Computer-Assisted Surgery (ICCAS), University of Leipzig, 04103 Leipzig, Germany
| |
Collapse
|
6
|
Editorial Paper for the Special Issue “Algorithms in Hyperspectral Data Analysis”. ALGORITHMS 2022. [DOI: 10.3390/a15040112] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 12/04/2022]
Abstract
This Special Issue contains four papers focused on hyperspectral data analysis [...]
Collapse
|
7
|
Sommer F, Sun B, Fischer J, Goldammer M, Thiele C, Malberg H, Markgraf W. Hyperspectral Imaging during Normothermic Machine Perfusion—A Functional Classification of Ex Vivo Kidneys Based on Convolutional Neural Networks. Biomedicines 2022; 10:biomedicines10020397. [PMID: 35203605 PMCID: PMC8962340 DOI: 10.3390/biomedicines10020397] [Citation(s) in RCA: 5] [Impact Index Per Article: 2.5] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/27/2021] [Revised: 01/28/2022] [Accepted: 01/30/2022] [Indexed: 12/18/2022] Open
Abstract
Facing an ongoing organ shortage in transplant medicine, strategies to increase the use of organs from marginal donors by objective organ assessment are being fostered. In this context, normothermic machine perfusion provides a platform for ex vivo organ evaluation during preservation. Consequently, analytical tools are emerging to determine organ quality. In this study, hyperspectral imaging (HSI) in the wavelength range of 550–995 nm was applied. Classification of 26 kidneys based on HSI was established using KidneyResNet, a convolutional neural network (CNN) based on the ResNet-18 architecture, to predict inulin clearance behavior. HSI preprocessing steps were implemented, including automated region of interest (ROI) selection, before executing the KidneyResNet algorithm. Training parameters and augmentation methods were investigated concerning their influence on the prediction. When classifying individual ROIs, the optimized KidneyResNet model achieved 84% and 62% accuracy in the validation and test set, respectively. With a majority decision on all ROIs of a kidney, the accuracy increased to 96% (validation set) and 100% (test set). These results demonstrate the feasibility of HSI in combination with KidneyResNet for non-invasive prediction of ex vivo kidney function. This knowledge of preoperative renal quality may support the organ acceptance decision.
Collapse
|
8
|
Mühle R, Markgraf W, Hilsmann A, Malberg H, Eisert P, Wisotzky EL. Comparison of different spectral cameras for image-guided organ transplantation. JOURNAL OF BIOMEDICAL OPTICS 2021; 26:JBO-210076RR. [PMID: 34304399 PMCID: PMC8305772 DOI: 10.1117/1.jbo.26.7.076007] [Citation(s) in RCA: 4] [Impact Index Per Article: 1.3] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Figures] [Subscribe] [Scholar Register] [Received: 03/08/2021] [Accepted: 06/24/2021] [Indexed: 06/13/2023]
Abstract
SIGNIFICANCE Hyperspectral and multispectral imaging (HMSI) in medical applications provides information about the physiology, morphology, and composition of tissues and organs. The use of these technologies enables the evaluation of biological objects and can potentially be applied as an objective assessment tool for medical professionals. AIM Our study investigates HMSI systems for their usability in medical applications. APPROACH Four HMSI systems (one hyperspectral pushbroom camera and three multispectral snapshot cameras) were examined and a spectrometer was used as a reference system, which was initially validated with a standardized color chart. The spectral accuracy of the cameras reproducing chemical properties of different biological objects (porcine blood, physiological porcine tissue, and pathological porcine tissue) was analyzed using the Pearson correlation coefficient. RESULTS All the HMSI cameras examined were able to provide the characteristic spectral properties of blood and tissues. A pushbroom camera and two snapshot systems achieve Pearson coefficients of at least 0.97 compared to the ground truth, indicating a very high positive correlation. Only one snapshot camera performs moderately to high positive correlation (0.59 to 0.85). CONCLUSION The knowledge of the suitability of HMSI cameras for accurate measurement of chemical properties of biological objects offers a good opportunity for the selection of the optimal imaging tool for specific medical applications, such as organ transplantation.
Collapse
Affiliation(s)
- Richard Mühle
- Technische Universität Dresden, Institute of Biomedical Engineering, Dresden, Germany
- Technische Universität Dresden, Department of Neurosurgery, Faculty of Medicine Carl Gustav Carus, Dresden, Germany
| | - Wenke Markgraf
- Technische Universität Dresden, Institute of Biomedical Engineering, Dresden, Germany
| | - Anna Hilsmann
- Fraunhofer Heinrich-Hertz-Institute, Department of Vision and Imaging Technologies, Berlin, Germany
| | - Hagen Malberg
- Technische Universität Dresden, Institute of Biomedical Engineering, Dresden, Germany
| | - Peter Eisert
- Fraunhofer Heinrich-Hertz-Institute, Department of Vision and Imaging Technologies, Berlin, Germany
- Humboldt Universität zu Berlin, Department of Visual Computing, Berlin, Germany
| | - Eric L. Wisotzky
- Fraunhofer Heinrich-Hertz-Institute, Department of Vision and Imaging Technologies, Berlin, Germany
- Humboldt Universität zu Berlin, Department of Visual Computing, Berlin, Germany
| |
Collapse
|