Early-Stage Ovarian Cancer Diagnosis Using Fuzzy Rough Sets with SVM Classification

2017 ◽  
pp. 43-60
Author(s):  
Nora Shoaip ◽  
Mohammed Mahfouz Elmogy ◽  
Alaa M. Riad ◽  
Hosam Zaghloul ◽  
Farid A. Badria
Keyword(s):  
Ovarian Cancer ◽  
Feature Selection ◽  
Cancer Diagnosis ◽  
Rough Set Theory ◽  
Early Stage ◽  
Classification Performance ◽  
Ca 125 ◽  
Processing Stage ◽  
Svm Classification ◽  
Early Stage Ovarian Cancer

Ovarian cancer is one of the most dangerous cancers among women which have a high rank of the cancers causing death. Ovarian cancer diagnoses are very difficult especially in early-stage because most symptoms associated with ovarian cancer such as Difficulty eating or feeling full quickly, Pelvic or abdominal pain, and Bloating are common and found in Women who do not have ovarian cancer. The CA-125 test is used as a tumor marker, high levels could be a sign of ovarian cancer, but sometimes it is not true because not all women with ovarian cancer have high CA-125 levels, particularly about 20% of ovarian cancers are found at an early stage. In this paper, we try to find the most important rules helping in Early-stage ovarian cancer Diagnosis by evaluating the significance of data between ovarian cancer and the amino acids. Therefore, we propose a Fuzzy Rough feature selection with Support Vector Machine (SVM) classification model. In the pre-processing stage, we use Fuzzy Rough set theory for feature selection. In post-processing stage, we use SVM classification which is a powerful method to get good classification performance. Finally, we compare the output results of the proposed system with other classification technique to guarantee the highest classification performance.

scholarly journals Development of a Multimarker Assay for Early Detection of Ovarian Cancer

2010 ◽  
Vol 28 (13) ◽  
pp. 2159-2166 ◽  
Author(s):  
Zoya Yurkovetsky ◽  
Steven Skates ◽  
Aleksey Lomakin ◽  
Brian Nolen ◽  
Trenton Pulsipher ◽  
...  
Keyword(s):  
Ovarian Cancer ◽  
Early Detection ◽  
Late Stage ◽  
Early Stage ◽  
Screening Strategy ◽  
Ca 125 ◽  
Great Promise ◽  
Biomarker Panel ◽  
Healthy Women ◽  
Early Stage Ovarian Cancer

PurposeEarly detection of ovarian cancer has great promise to improve clinical outcome.Patients and MethodsNinety-six serum biomarkers were analyzed in sera from healthy women and from patients with ovarian cancer, benign pelvic tumors, and breast, colorectal, and lung cancers, using multiplex xMAP bead-based immunoassays. A Metropolis algorithm with Monte Carlo simulation (MMC) was used for analysis of the data.ResultsA training set, including sera from 139 patients with early-stage ovarian cancer, 149 patients with late-stage ovarian cancer, and 1,102 healthy women, was analyzed with MMC algorithm and cross validation to identify an optimal biomarker panel discriminating early-stage cancer from healthy controls. The four-biomarker panel providing the highest diagnostic power of 86% sensitivity (SN) for early-stage and 93% SN for late-stage ovarian cancer at 98% specificity (SP) was comprised of CA-125, HE4, CEA, and VCAM-1. This model was applied to an independent blinded validation set consisting of sera from 44 patients with early-stage ovarian cancer, 124 patients with late-stage ovarian cancer, and 929 healthy women, providing unbiased estimates of 86% SN for stage I and II and 95% SN for stage III and IV disease at 98% SP. This panel was selective for ovarian cancer showing SN of 33% for benign pelvic disease, SN of 6% for breast cancer, SN of 0% for colorectal cancer, and SN of 36% for lung cancer.ConclusionA panel of CA-125, HE4, CEA, and VCAM-1, after additional validation, could serve as an initial stage in a screening strategy for epithelial ovarian cancer.

scholarly journals Early-Stage Ovarian Malignancy Score versus Risk of Malignancy Indices: Accuracy and Clinical Utility for Preoperative Diagnosis of Women with Adnexal Masses

Medicina ◽  
2020 ◽  
Vol 56 (12) ◽  
pp. 702
Author(s):  
Phichayut Phinyo ◽  
Jayanton Patumanond ◽  
Panprapha Saenrungmuaeng ◽  
Watcharin Chirdchim ◽  
Tanyong Pipanmekaporn ◽  
...  
Keyword(s):  
Ovarian Cancer ◽  
Clinical Utility ◽  
Early Stage ◽  
Ca 125 ◽  
Lower Sensitivity ◽  
Ovarian Malignancy ◽  
Threshold Probability ◽  
Adnexal Masses ◽  
Risk Of Malignancy ◽  
Early Stage Ovarian Cancer

Background and objectives: To compare the diagnostic accuracy and clinical utility of the Early-stage Ovarian Malignancy (EOM) score with the Risk of Malignancy Index (RMI) in the presurgical assessment of women presenting with adnexal masses. Materials and Methods: A secondary analysis was carried out in a retrospective cohort of women who presented with an adnexal mass and were scheduled for surgery at Phrapokklao Hospital between September 2013 and December 2017. The clinical characteristics, ultrasonographic features of the masses, and preoperative CA-125 levels were recorded. The EOM and the RMI score were calculated and compared in terms of accuracy and clinical utility. Decision curve analysis (DCA), which examined the net benefit (NB) of applying the EOM and the RMI in practice at a range of threshold probabilities, was presented. Results: In this study, data from 270 patients were analyzed. Fifty-four (20.0%) women in the sample had early-stage ovarian cancer. All four RMI versions demonstrated a lower sensitivity for the detection of patients with early-stage ovarian cancer compared to an EOM score ≥ 15. An EOM ≥ 15 resulted in a higher proportion of net true positive or NB than all versions of the RMIs from a threshold probability of 5% to 30%. Conclusions: It also showed a higher capability to reduce the number of inappropriate referrals than the RMIs at a threshold probability between 5% and 30%. The EOM score showed higher diagnostic sensitivity and has the potential to be clinically more useful than the RMIs to triage women who present with adnexal masses for referral to oncologic gynecologists. Further external validation is required to support our findings.

Surface Functionalization of an Ovarian Cancer Diagnostic Biosensor

2011 ◽  
Author(s):  
Asad Ahmad ◽  
Nathan Gallant ◽  
Rasim Guldiken ◽  
Onursal Onen
Keyword(s):  
Ovarian Cancer ◽  
Early Stage ◽  
South Florida ◽  
Ca 125 ◽  
Blood Levels ◽  
Short Median Survival Time ◽  
Urinary Levels ◽  
Risk Patients ◽  
The University ◽  
Early Stage Ovarian Cancer

Ovarian cancer is the fifth leading cause of death among women in United States and the disease has 1.4% (1 in 71) lifetime risk. Patients with ovarian cancer have a short median survival time after diagnosis with their 5-year survival rate being less than 40%. Early stage ovarian cancer represents an important target for screening since it is lethal in most late stage cases (1). Currently the primary screening procedure for ovarian cancer are blood levels of cancer antigen (CA) 125, however CA 125 levels can also be elevated due to other disorders and do not provide conclusive results (2). Utilizing the research done at the Cell and Molecular Biology department at the University of South Florida which conclusively revealed that urinary levels of bcl-2 are elevated in ovarian cancer patients (3), this research it the first of its kind looking to assess the capture of an analyte protein on a series of potential bioconjugated surfaces for use in a novel acoustic biosensor. Therefore, this research addresses the need for a reliable and economic testing platform to detect ovarian cancer.

scholarly journals Validation of Candidate Serum Ovarian Cancer Biomarkers for Early Detection

2007 ◽  
Vol 2 ◽  
pp. 117727190700200 ◽  
Author(s):  
Feng Su ◽  
Jennifer Lang ◽  
Ashutosh Kumar ◽  
Carey Ng ◽  
Brian Hsieh ◽  
...  
Keyword(s):  
Ovarian Cancer ◽  
Early Detection ◽  
Regression Models ◽  
Serum Proteins ◽  
Early Stage ◽  
Ovarian Tumors ◽  
Ca 125 ◽  
Endometrioid Adenocarcinoma ◽  
Serum Samples ◽  
Early Stage Ovarian Cancer

Objective We have previously analyzed protein profiles using Surface Enhanced Laser Desorption and Ionization Time-Of-Flight Mass Spectroscopy (SELDI-TOF-MS) [Kozak et al. 2003, Proc. Natl. Acad. Sci. U.S.A. 100:12343–8] and identified 3 differentially expressed serum proteins for the diagnosis of ovarian cancer (OC) [Kozak et al. 2005, Proteomics, 5:4589–96], namely, apolipoprotein A-I (apoA-I), transthyretin (TTR) and transferin (TF). The objective of the present study is to determine the efficacy of the three OC biomarkers for the detection of early stage (ES) OC, in direct comparison to CA125. Methods The levels of CA125, apoA-I, TTR and TF were measured in 392 serum samples [82 women with normal ovaries (N), 24 women with benign ovarian tumors (B), 85 women with ovarian tumors of low malignant potential (LMP), 126 women with early stage ovarian cancer (ESOC), and 75 women with late stage ovarian cancer (LSOC)], obtained through the GOG and Cooperative Human Tissue Network. Following statistical analysis, multivariate regression models were built to evaluate the utility of the three OC markers in early detection. Results Multiple logistic regression models (MLRM) utilizing all biomarker values (CA125, TTR, TF and apoA-I) from all histological subtypes (serous, mucinous, and endometrioid adenocarcinoma) distinguished normal samples from LMP with 91% sensitivity (specificity 92%), and normal samples from ESOC with a sensitivity of 89% (specificity 92%). MLRM, utilizing values of all four markers from only the mucinous histological subtype showed that collectively, CA125, TTR, TF and apoA-I, were able to distinguish normal samples from mucinous LMP with 90% sensitivity, and further distinguished normal samples from early stage mucinous ovarian cancer with a sensitivity of 95%. In contrast, in serum samples from patients with mucinous tumors, CA125 alone was able to distinguish normal samples from LMP and early stage ovarian cancer with a sensitivity of only 46% and 47%, respectively. Furthermore, collectively, apoA-I, TTR and TF (excluding CA-125) distinguished i) normal samples from samples representing all histopathologic subtypes of LMP, with a sensitivity of 73%, ii) normal samples from ESOC with a sensitivity of 84% and iii) normal samples from LSOC with a sensitivity of 97%. More strikingly, the sensitivity in distinguishing normal versus mucinous ESOC, utilizing apoA-I, TF and TTR (CA-125 excluded), was 95% (specificity 86%; AUC 95%). Conclusions These results suggest that the biomarker panel consisting of apoA-I, TTR and TF may significantly improve early detection of OC.

scholarly journals Automated Assay of a Four-Protein Biomarker Panel for Improved Detection of Ovarian Cancer

Cancers ◽  
2021 ◽  
Vol 13 (2) ◽  
pp. 325
Author(s):  
Christopher Walker ◽  
Tuan-Minh Nguyen ◽  
Shlomit Jessel ◽  
Ayesha B. Alvero ◽  
Dan-Arin Silasi ◽  
...  
Keyword(s):  
Ovarian Cancer ◽  
Early Detection ◽  
Predictive Value ◽  
Early Stage ◽  
Ca 125 ◽  
Healthy Controls ◽  
Classification Models ◽  
Serum Samples ◽  
Protein Biomarker ◽  
Biomarker Panel

Background: Mortality from ovarian cancer remains high due to the lack of methods for early detection. The difficulty lies in the low prevalence of the disease necessitating a significantly high specificity and positive-predictive value (PPV) to avoid unneeded and invasive intervention. Currently, cancer antigen- 125 (CA-125) is the most commonly used biomarker for the early detection of ovarian cancer. In this study we determine the value of combining macrophage migration inhibitory factor (MIF), osteopontin (OPN), and prolactin (PROL) with CA-125 in the detection of ovarian cancer serum samples from healthy controls. Materials and Methods: A total of 432 serum samples were included in this study. 153 samples were from ovarian cancer patients and 279 samples were from age-matched healthy controls. The four proteins were quantified using a fully automated, multi-analyte immunoassay. The serum samples were divided into training and testing datasets and analyzed using four classification models to calculate accuracy, sensitivity, specificity, PPV, negative predictive value (NPV), and area under the receiver operating characteristic curve (AUC). Results: The four-protein biomarker panel yielded an average accuracy of 91% compared to 85% using CA-125 alone across four classification models (p = 3.224 × 10−9). Further, in our cohort, the four-protein biomarker panel demonstrated a higher sensitivity (median of 76%), specificity (median of 98%), PPV (median of 91.5%), and NPV (median of 92%), compared to CA-125 alone. The performance of the four-protein biomarker remained better than CA-125 alone even in experiments comparing early stage (Stage I and Stage II) ovarian cancer to healthy controls. Conclusions: Combining MIF, OPN, PROL, and CA-125 can better differentiate ovarian cancer from healthy controls compared to CA-125 alone.

scholarly journals Comparison of the survival outcomes of laparoscopy versus laparotomy in treatment of early-stage ovarian cancer: a systematic review and meta-analysis

2021 ◽  
Vol 14 (1) ◽  
Author(s):  
Qingduo Kong ◽  
Hongyi Wei ◽  
Jing Zhang ◽  
Yilin Li ◽  
Yongjun Wang
Keyword(s):  
Ovarian Cancer ◽  
Overall Survival ◽  
Early Stage ◽  
Meta Analysis ◽  
Progression Free Survival ◽  
Cochrane Library ◽  
Survival Outcomes ◽  
Free Survival ◽  
Review Manager ◽  
Early Stage Ovarian Cancer

Abstract Background Laparoscopy has been widely used for patients with early-stage epithelial ovarian cancer (eEOC). However, there is limited evidence regarding whether survival outcomes of laparoscopy are equivalent to those of laparotomy among patients with eEOC. The result of survival outcomes of laparoscopy is still controversial. The aim of this meta-analysis is to analyze the survival outcomes of laparoscopy versus laparotomy in the treatment of eEOC. Methods According to the keywords, Pubmed, Embase, Cochrane Library and Clinicaltrials.gov were searched for studies from January 1994 to January 2021. Studies comparing the efficacy and safety of laparoscopy versus laparotomy for patients with eEOC were assessed for eligibility. Only studies including outcomes of overall survival (OS) were enrolled. The meta-analysis was performed using Stata software (Version 12.0) and Review Manager (Version 5.2). Results A total of 6 retrospective non-random studies were included in this meta-analysis. The pooled results indicated that there was no difference between two approaches for patients with eEOC in OS (HR = 0.6, P = 0.446), progression-free survival (PFS) (HR = 0.6, P = 0.137) and upstaging rate (OR = 1.18, P = 0.54). But the recurrence rate of laparoscopic surgery was lower than that of laparotomic surgery (OR = 0.48, P = 0.008). Conclusions Laparoscopy and laparotomy appear to provide comparable overall survival and progression-free survival outcomes for patients with eEOC. Further high-quality studies are needed to enhance this statement.

Sign in / Sign up

Export Citation Format

Share Document