1
|
Liu W, Jin X, Li J, Xue Y, Li Y, Qian Z, Li W, Yan X. Study of cervical precancerous lesions detection by spectroscopy and support vector machine. MINIM INVASIV THER 2020; 30:208-214. [PMID: 32347137 DOI: 10.1080/13645706.2020.1723111] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/24/2022]
Abstract
BACKGROUND AND OBJECTIVE Diffuse reflectance spectroscopy (DRS) offers a fast, non-invasive, and low-cost alternative for cervical cancer diagnosis. We aim to develop a method for screening precancerous lesions based on DRS. MATERIAL AND METHODS Characteristic parameters of cervical tissue were extracted from spectra, including optical characteristic parameters such as absorption and scattering coefficients, and some slope and area parameters of the spectrum. Data were randomly divided into training (60%) and test (40%) sets. Of the 210 included patients, 166 were healthy, 22 had erosion of the cervix, and 31 had cervical intraepithelial neoplasia (CIN). The support vector machine (SVM) algorithm was used to classify normal and abnormal cervical tissue based on 11 characteristic parameters. RESULTS The SVM with linear kernel function, applied on the training data, could distinguish tissue with lesions from healthy tissue with an accuracy of 1.00. When the classifiers were applied to the test set, erosion of cervix and CIN could be discriminated from healthy tissue with an accuracy of 0.95 (±0.03). CONCLUSIONS This research shows that the diagnostic algorithm can be valuable for non-invasive diagnosis of cervical cancer. This is a significant step toward the development of a tool for tissue assessment of cervical cancer.
Collapse
Affiliation(s)
- Wenwen Liu
- Department of Biomedical Engineering, College of Automation Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing, China
| | - Xiaofei Jin
- Department of Biomedical Engineering, College of Automation Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing, China
| | - Junjun Li
- Department of Biomedical Engineering, College of Automation Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing, China
| | - Yanbai Xue
- Department of Biomedical Engineering, College of Automation Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing, China
| | - Yiran Li
- Department of Biomedical Engineering, College of Automation Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing, China
| | - Zhiyu Qian
- Department of Biomedical Engineering, College of Automation Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing, China
| | - Weitao Li
- Department of Biomedical Engineering, College of Automation Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing, China
| | - Xuemei Yan
- Department of Gynecology, Nanjing BenQ Hospital Co Ltd, Nanjing, China
| |
Collapse
|
2
|
Johansen TH, Møllersen K, Ortega S, Fabelo H, Garcia A, Callico GM, Godtliebsen F. Recent advances in hyperspectral imaging for melanoma detection. WIRES COMPUTATIONAL STATISTICS 2019. [DOI: 10.1002/wics.1465] [Citation(s) in RCA: 19] [Impact Index Per Article: 3.8] [Reference Citation Analysis] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 11/11/2022]
Affiliation(s)
| | - Kajsa Møllersen
- Department of Community Medicine UiT The Arctic University of Norway Tromsø Norway
| | - Samuel Ortega
- Institute for Applied Microelectronics University of Las Palmas de Gran Canaria Las Palmas Spain
| | - Himar Fabelo
- Institute for Applied Microelectronics University of Las Palmas de Gran Canaria Las Palmas Spain
| | - Aday Garcia
- Institute for Applied Microelectronics University of Las Palmas de Gran Canaria Las Palmas Spain
| | - Gustavo M. Callico
- Institute for Applied Microelectronics University of Las Palmas de Gran Canaria Las Palmas Spain
| | - Fred Godtliebsen
- Department of Mathematics and Statistics UiT The Arctic University of Norway Tromsø Norway
| |
Collapse
|
3
|
Zhu H, Morris JS, Wei F, Cox DD. Multivariate functional response regression, with application to fluorescence spectroscopy in a cervical pre-cancer study. Comput Stat Data Anal 2017; 111:88-101. [PMID: 29051679 PMCID: PMC5642121 DOI: 10.1016/j.csda.2017.02.004] [Citation(s) in RCA: 9] [Impact Index Per Article: 1.3] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/20/2022]
Abstract
Many scientific studies measure different types of high-dimensional signals or images from the same subject, producing multivariate functional data. These functional measurements carry different types of information about the scientific process, and a joint analysis that integrates information across them may provide new insights into the underlying mechanism for the phenomenon under study. Motivated by fluorescence spectroscopy data in a cervical pre-cancer study, a multivariate functional response regression model is proposed, which treats multivariate functional observations as responses and a common set of covariates as predictors. This novel modeling framework simultaneously accounts for correlations between functional variables and potential multi-level structures in data that are induced by experimental design. The model is fitted by performing a two-stage linear transformation-a basis expansion to each functional variable followed by principal component analysis for the concatenated basis coefficients. This transformation effectively reduces the intra-and inter-function correlations and facilitates fast and convenient calculation. A fully Bayesian approach is adopted to sample the model parameters in the transformed space, and posterior inference is performed after inverse-transforming the regression coefficients back to the original data domain. The proposed approach produces functional tests that flag local regions on the functional effects, while controlling the overall experiment-wise error rate or false discovery rate. It also enables functional discriminant analysis through posterior predictive calculation. Analysis of the fluorescence spectroscopy data reveals local regions with differential expressions across the pre-cancer and normal samples. These regions may serve as biomarkers for prognosis and disease assessment.
Collapse
Affiliation(s)
- Hongxiao Zhu
- Department of Statistics, Virginia Tech, Blacksburg, VA 24061
| | - Jeffrey S Morris
- The University of Texas MD Anderson Cancer Center, Houston, TX 77230
| | - Fengrong Wei
- Department of Mathematics, University of West Georgia, Carrollton, GA 30118
| | - Dennis D Cox
- Department of Statistics, Rice University, Houston, TX 77005
| |
Collapse
|
4
|
Abstract
OBJECTIVE To assess the diagnostic value of alternative (digital) colposcopy techniques for detection of cervical intraepithelial neoplasia (CIN) 2 or worse in a colposcopy population. DATA SOURCES MEDLINE, EMBASE, ClinicalTrials.gov, and the Cochrane Library were searched from inception up to January 11, 2016, for studies that evaluated the diagnostic value of alternative (digital) colposcopy techniques. METHODS OF STUDY SELECTION Inclusion criteria were: 1) an alternative (digital) colposcopy technique was used in a colposcopy population; 2) a histologic outcome was reported, classified as CIN, differentiating between mild dysplasia or less (CIN 1 or less), and moderate dysplasia or worse (CIN 2 or greater); 3) the entire cervix was scanned at once or a per-woman analysis was performed; 4) no other topical application than acetic acid and Lugol's solution was used; 5) at least three eligible studies had to be available within a single technique; and 6) studies obtained research ethics approval. Language was restricted to English. TABULATION, INTEGRATION, AND RESULTS Two reviewers assessed the eligibility of the identified articles. Disagreements were resolved by a third reviewer. Thirteen studies met the inclusion criteria. We found six studies on fluorescence and reflectance spectroscopy, including 2,530 women, with a pooled sensitivity of 93% (95% confidence interval [CI] 89-95%) and specificity of 62% (95% CI 47-76%). Four studies on dynamic spectral imaging were found including 1,173 women with a pooled sensitivity of 69% (95% CI 48-85%) and specificity of 83% (95% CI 76-88%). We found three studies on optical coherence tomography including 693 women with a pooled sensitivity of 48% (95% CI 32-64%) and specificity of 77% (95% CI 52-91%). Previously published conventional colposcopy results showed a sensitivity of 61% (95% CI 58-63%) and a specificity of 85% (95% CI 83-86%). CONCLUSION Alternative (digital) colposcopy techniques may result in increased sensitivity and specificity, but no recommendation for introduction in clinical practice can be made yet.
Collapse
|
5
|
Multimodal Hyperspectroscopic Imaging for Detection of High-Grade Cervical Intraepithelial Neoplasia. J Low Genit Tract Dis 2017; 21:166-170. [PMID: 28403024 DOI: 10.1097/lgt.0000000000000309] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.1] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/26/2022]
Abstract
OBJECTIVE Numerous new alternative digital colposcopy techniques have been developed, of which multimodal hyperspectroscopy (MHS) showed a high sensitivity in previous studies. The objective of this prospective single-center cohort study was to evaluate the clinical value of MHS for detecting high-grade cervical intraepithelial neoplasia in a colposcopy referral population and colposcopy follow-up population, to assess whether MHS could be safely used to improve care for women at risk for high-grade cervical intraepithelial neoplasia. MATERIALS AND METHODS A total of 125 women from a colposcopy referral population and colposcopy follow-up population were evaluated with MHS and tested for the presence of high-risk human papillomavirus (HPV) with HPV-16/18 genotyping. Spectroscopic measurements of the cervix were taken and compared with an end point based on histology, high-risk HPV, and cytology. Evaluable data for analysis were collected from 102 of the subjects. Sensitivity, specificity, and predictive values were calculated for MHS and colposcopic impression based on conventional colposcopic examination. RESULTS From the total study population of the 102 patients, 47 were enrolled in the colposcopy referral group and 55 in the colposcopy follow-up group. The MHS yielded a sensitivity of 93.6% (95% CI = 78.6-99.2), with a corresponding specificity of 42.3% (95% CI = 30.6-54.6) in the group with a composite end point. No adverse effects occurred, and patient acceptability was high. CONCLUSIONS Multimodal hyperspectroscopy is a digital colposcopy technique that offers an easy, rapid, well-tolerated point-of-care assessment with a high sensitivity for the presence of high-grade cervical intraepithelial lesions, however, with a low specificity, resulting in limited clinical value.
Collapse
|
6
|
Gallwas J, Jalilova A, Ladurner R, Kolben TM, Kolben T, Ditsch N, Homann C, Lankenau E, Dannecker C. Detection of cervical intraepithelial neoplasia by using optical coherence tomography in combination with microscopy. JOURNAL OF BIOMEDICAL OPTICS 2017; 22:16013. [PMID: 28118427 DOI: 10.1117/1.jbo.22.1.016013] [Citation(s) in RCA: 13] [Impact Index Per Article: 1.9] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Received: 08/29/2016] [Accepted: 01/03/2017] [Indexed: 05/09/2023]
Abstract
Optical coherence tomography (OCT) is a noninvasive high-resolution imaging technique that permits the detection of cancerous and precancerous lesions of the uterine cervix. The purpose of this study was to evaluate a new system that integrates an OCT device into a microscope. OCT images were taken from loop electrosurgical excision procedure (LEEP) specimens under microscopic guidance. The images were blinded with respect to their origin within the microscopic image and analyzed independently by two investigators using initially defined criteria and later compared to the corresponding histology. Sensitivity and specificity were calculated with respect to the correct identification of high-grade squamous intraepithelial lesions (HSIL). The interinvestigator agreement was assessed by using Cohen’s kappa statistics. About 160 OCT images were obtained from 20 LEEP specimens. Sixty randomly chosen images were used to define reproducible criteria for evaluation. The assessment of the remaining 100 images showed a sensitivity of 88% (second investigator 84%) and a specificity of 69% (65%) in detecting HSIL. Surgical microscopy-guided OCT appears to be a promising technique for immediate assessment of microanatomical changes. In the gynecological setting, the combination of OCT with a colposcope may improve the detection of high-grade squamous intraepithelial lesions.
Collapse
Affiliation(s)
- Julia Gallwas
- Ludwig Maximilians University Munich, Grosshadern Medical Campus, Department of Obstetrics and Gynecology, Marchioninistrasse 15, 81377 Munich, Germany
| | - Aydan Jalilova
- Ludwig Maximilians University Munich, Grosshadern Medical Campus, Department of Obstetrics and Gynecology, Marchioninistrasse 15, 81377 Munich, Germany
| | - Roland Ladurner
- Ludwig-Maximilians University Munich, Innenstadt Medical Campus, Department of Surgery, Nussbaumstrasse 20, 80336 Munich, Germany
| | - Theresa Maria Kolben
- Ludwig Maximilians University Munich, Grosshadern Medical Campus, Department of Obstetrics and Gynecology, Marchioninistrasse 15, 81377 Munich, Germany
| | - Thomas Kolben
- Ludwig Maximilians University Munich, Grosshadern Medical Campus, Department of Obstetrics and Gynecology, Marchioninistrasse 15, 81377 Munich, Germany
| | - Nina Ditsch
- Ludwig Maximilians University Munich, Grosshadern Medical Campus, Department of Obstetrics and Gynecology, Marchioninistrasse 15, 81377 Munich, Germany
| | - Christian Homann
- cLudwig Maximilians University Munich, Grosshadern Medical Campus, Laser-Research Laboratory, LIFE Center, Feodor-Lynen-Strasse 19, 81377 Munich, Germany
| | - Eva Lankenau
- OptoMedical Technologies GmbH, Maria Goeppert Strasse 9, 23562 Luebeck, Germany
| | - Christian Dannecker
- Ludwig Maximilians University Munich, Grosshadern Medical Campus, Department of Obstetrics and Gynecology, Marchioninistrasse 15, 81377 Munich, Germany
| |
Collapse
|
7
|
Yang J, Zhu H, Choi T, Cox DD. Smoothing and Mean-Covariance Estimation of Functional Data with a Bayesian Hierarchical Model. BAYESIAN ANALYSIS 2016; 11:649-670. [PMID: 34457106 PMCID: PMC8387981 DOI: 10.1214/15-ba967] [Citation(s) in RCA: 12] [Impact Index Per Article: 1.5] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 06/13/2023]
Abstract
Functional data, with basic observational units being functions (e.g., curves, surfaces) varying over a continuum, are frequently encountered in various applications. While many statistical tools have been developed for functional data analysis, the issue of smoothing all functional observations simultaneously is less studied. Existing methods often focus on smoothing each individual function separately, at the risk of removing important systematic patterns common across functions. We propose a nonparametric Bayesian approach to smooth all functional observations simultaneously and nonparametrically. In the proposed approach, we assume that the functional observations are independent Gaussian processes subject to a common level of measurement errors, enabling the borrowing of strength across all observations. Unlike most Gaussian process regression models that rely on pre-specified structures for the covariance kernel, we adopt a hierarchical framework by assuming a Gaussian process prior for the mean function and an Inverse-Wishart process prior for the covariance function. These prior assumptions induce an automatic mean-covariance estimation in the posterior inference in addition to the simultaneous smoothing of all observations. Such a hierarchical framework is flexible enough to incorporate functional data with different characteristics, including data measured on either common or uncommon grids, and data with either stationary or nonstationary covariance structures. Simulations and real data analysis demonstrate that, in comparison with alternative methods, the proposed Bayesian approach achieves better smoothing accuracy and comparable mean-covariance estimation results. Furthermore, it can successfully retain the systematic patterns in the functional observations that are usually neglected by the existing functional data analyses based on individual-curve smoothing.
Collapse
Affiliation(s)
- Jingjing Yang
- Department of Biostatistics, University of Michigan, Ann Arbor, MI 48109, USA
| | - Hongxiao Zhu
- Department of Statistics, Virginia Tech, Blacksburg, VA 24061, USA
| | - Taeryon Choi
- Department of Statistics, Korea University, Seoul 136-701, Republic of Korea
| | - Dennis D Cox
- Department of Statistics, Rice University, Houston, TX 77005, USA
| |
Collapse
|
8
|
Yamal JM, Guillaud M, Atkinson EN, Follen M, MacAulay C, Cantor SB, Cox DD. Prediction using hierarchical data: Applications for automated detection of cervical cancer. Stat Anal Data Min 2015; 8:65-74. [PMID: 26617681 DOI: 10.1002/sam.11261] [Citation(s) in RCA: 5] [Impact Index Per Article: 0.6] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/10/2022]
Abstract
Although the Papanicolaou smear has been successful in decreasing cervical cancer incidence in the developed world, there exist many challenges for implementation in the developing world. Quantitative cytology, a semi-automated method that quantifies cellular image features, is a promising screening test candidate. The nested structure of its data (measurements of multiple cells within a patient) provides challenges to the usual classification problem. Here we perform a comparative study of three main approaches for problems with this general data structure: a) extract patient-level features from the cell-level data; b) use a statistical model that accounts for the hierarchical data structure; and c) classify at the cellular level and use an ad hoc approach to classify at the patient level. We apply these methods to a dataset of 1,728 patients, with an average of 2,600 cells collected per patient and 133 features measured per cell, predicting whether a patient had a positive biopsy result. The best approach we found was to classify at the cellular level and count the number of cells that had a posterior probability greater than a threshold value, with estimated 61% sensitivity and 89% specificity on independent data. Recent statistical learning developments allowed us to achieve high accuracy.
Collapse
Affiliation(s)
- Jose-Miguel Yamal
- Department of Biostatistics, The University of Texas School of Public Health, 1200 Herman Pressler, Suite W-928, Houston, TX 77030, USA
| | - Martial Guillaud
- Department of Integrative Oncology, British Columbia Cancer Research Centre, 675 West 10th Ave, Vancouver, BC, V5Z 1L3, Canada
| | - E Neely Atkinson
- Department of Statistics, Rice University, 6100 Main St., Houston, TX 77005, USA
| | - Michele Follen
- Department of Obstetrics and Gynecology, Brookdale Hospital and Medical Center, 555 Rockaway Pkwy, Brooklyn, NY 11212, USA
| | - Calum MacAulay
- Department of Integrative Oncology, British Columbia Cancer Research Centre, 675 West 10th Ave, Vancouver, BC, V5Z 1L3, Canada
| | - Scott B Cantor
- Department of Health Services Research, The University of Texas MD Anderson Cancer Center, P.O. Box 301402, Unit 1444, Houston, TX 77230-1402, USA
| | - Dennis D Cox
- Department of Statistics, Rice University, 6100 Main St., Houston, TX 77005, USA
| |
Collapse
|
9
|
Novel advancements in colposcopy: historical perspectives and a systematic review of future developments. J Low Genit Tract Dis 2015; 18:246-60. [PMID: 24633164 DOI: 10.1097/lgt.0b013e3182a72170] [Citation(s) in RCA: 6] [Impact Index Per Article: 0.7] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 12/13/2022]
Abstract
OBJECTIVE To describe novel innovations and techniques for the detection of high-grade dysplasia. MATERIALS AND METHODS Studies were identified through the PubMed database, spanning the last 10 years. The key words (["computerized colposcopy" or "digital colposcopy" or "spectroscopy" or "multispectral digital colposcopy" or "dynamic spectral imaging", or "electrical impedance spectroscopy" or "confocal endomicroscopy" or "confocal microscopy"or "optical coherence tomography"] and ["cervical dysplasia" or cervical precancer" or "cervix" or "cervical"]) were used. The inclusion criteria were published articles of original research referring to noncolposcopic evaluation of the cervix for the detection of cervical dysplasia. Only English-language articles from the past 10 years were included, in which the technologies were used in vivo, and sensitivities and specificities could be calculated. RESULTS The single author reviewed the articles for inclusion. Primary search of the database yielded 59 articles, and secondary cross-reference yielded 12 articles. Thirty-two articles met the inclusion criteria. CONCLUSIONS An instrument that globally assesses the cervix, such as computer-assisted colposcopy, optical spectroscopy, and dynamic spectral imaging, would provided the most comprehensive estimate of disease and is therefore best suited when treatment is preferred. Electrical impedance spectroscopy, confocal microscopy, and optical coherence tomography provide information at the cellular level to estimate histology and are therefore best suited when deferment of treatment is preferred. If a device is to eventually replace the colposcope, it will likely combine technologies to best meet the needs of the target population, and as such, no single instrument may prove to be universally appropriate. Analyses of false-positive rates, additional colposcopies and biopsies, cost, and absolute life-savings will be important when considering these technologies and are limited thus far.
Collapse
|
10
|
Wang L, Lee JS, Lane P, Atkinson EN, Zuluaga A, Follen M, MacAulay C, Cox DD. A statistical model for removing inter-device differences in spectroscopy. OPTICS EXPRESS 2014; 22:7617-7624. [PMID: 24718136 PMCID: PMC4083050 DOI: 10.1364/oe.22.007617] [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: 01/23/2014] [Revised: 03/13/2014] [Accepted: 03/16/2014] [Indexed: 06/03/2023]
Abstract
We are investigating spectroscopic devices designed to make in vivo cervical tissue measurements to detect pre-cancerous and cancerous lesions. All devices have the same design and ideally should record identical measurements. However, we observed consistent differences among them. An experiment was designed to study the sources of variation in the measurements recorded. Here we present a log additive statistical model that incorporates the sources of variability we identified. Based on this model, we estimated correction factors from the experimental data needed to eliminate the inter-device variability and other sources of variation. These correction factors are intended to improve the accuracy and repeatability of such devices when making future measurements on patient tissue.
Collapse
Affiliation(s)
- Lu Wang
- Department of Statistics, Rice University, 6100 Main St., Houston, TX 77005 USA
| | - Jong Soo Lee
- Department of Applied Economics and Statistics, University of Delaware, 206 Townsend Hall, Newark, DE 19716 USA
| | - Pierre Lane
- Department of Cancer Imaging, British Columbia Cancer Research Centre, 600 West 10th Ave., Vancouver, British Columbia V5Z4E6 Canada
| | - E. Neely Atkinson
- Department of Statistics, Rice University, 6100 Main St., Houston, TX 77005 USA
| | - Andres Zuluaga
- Clearview App, Inc. 3900 Essex Ln Ste 250 Houston, Texas 77027-5181 USA
| | - Michele Follen
- Department of Obstetrics and Gynecology, Paul L. Foster School of Medicine, Texas Tech University Health Sciences Center, El Paso, TX 79905 USA
| | - Calum MacAulay
- Department of Cancer Imaging, British Columbia Cancer Research Centre, 600 West 10th Ave., Vancouver, British Columbia V5Z4E6 Canada
| | - Dennis D. Cox
- Department of Statistics, Rice University, 6100 Main St., Houston, TX 77005 USA
| |
Collapse
|