scholarly journals Robust Whole Slide Image Analysis for Cervical Cancer Screening Using Deep Learning

2021 ◽  
Author(s):  
Shenghua Cheng ◽  
Sibo Liu ◽  
Jingya Yu ◽  
Gong Rao ◽  
Yuwei Xiao ◽  
...  
Keyword(s):  
Cervical Cancer ◽  
Cancer Screening ◽  
Cervical Cancer Screening ◽  
True Positive Rate ◽  
Classification Model ◽  
Computer Assisted ◽  
Slide Image ◽  
Positive Rate ◽  
Test Sets ◽  
Assisted Diagnosis

Abstract Computer-assisted diagnosis is key for popularizing cervical cancer screening. However, current recognition algorithms are insufficient in accuracy and generalization for cervical lesion cells, especially when facing diversity data in clinical applications. Inspired by manual reading slide under microscopes, we develop a progressive lesion cell recognition method combing low and high resolutions WSIs to recommend lesion cells and a recurrent neural network-based WSI classification model to evaluate the lesion degree of WSIs. After validating our system on 3,545 patient-wise WSIs with 79,218 annotations from multiple hospitals and several imaging instruments, on multi-center independent test sets of 1,170 patient-wise WSIs, we achieve 93.5% Specificity and 95.1% Sensitivity for classifying slides, closely equivalent to the average level of three independent cytopathologists, and obtain 88.5% TPR (true positive rate) for recommending top 10 lesion cells on 447 positive slides. After deploying, our system recognizes one giga-pixel WSI in about 1.5 minutes using one Nvidia 1080Ti GPU.

scholarly journals Robust whole slide image analysis for cervical cancer screening using deep learning

2021 ◽  
Vol 12 (1) ◽  
Author(s):  
Shenghua Cheng ◽  
Sibo Liu ◽  
Jingya Yu ◽  
Gong Rao ◽  
Yuwei Xiao ◽  
...  
Keyword(s):  
Cervical Cancer ◽  
Cancer Screening ◽  
Cervical Cancer Screening ◽  
True Positive Rate ◽  
Classification Model ◽  
Computer Assisted ◽  
Whole Slide Image ◽  
Slide Image ◽  
Positive Rate ◽  
Analysis System

AbstractComputer-assisted diagnosis is key for scaling up cervical cancer screening. However, current recognition algorithms perform poorly on whole slide image (WSI) analysis, fail to generalize for diverse staining and imaging, and show sub-optimal clinical-level verification. Here, we develop a progressive lesion cell recognition method combining low- and high-resolution WSIs to recommend lesion cells and a recurrent neural network-based WSI classification model to evaluate the lesion degree of WSIs. We train and validate our WSI analysis system on 3,545 patient-wise WSIs with 79,911 annotations from multiple hospitals and several imaging instruments. On multi-center independent test sets of 1,170 patient-wise WSIs, we achieve 93.5% Specificity and 95.1% Sensitivity for classifying slides, comparing favourably to the average performance of three independent cytopathologists, and obtain 88.5% true positive rate for highlighting the top 10 lesion cells on 447 positive slides. After deployment, our system recognizes a one giga-pixel WSI in about 1.5 min.

scholarly journals Computer-Assisted Screening for Cervical Cancer Using Digital Image Processing of Pap Smear Images

2020 ◽  
Vol 10 (5) ◽  
pp. 1800 ◽  
Author(s):  
Kyi Pyar Win ◽  
Yuttana Kitjaidure ◽  
Kazuhiko Hamamoto ◽  
Thet Myo Aung
Keyword(s):  
Image Processing ◽  
Cervical Cancer ◽  
Cancer Screening ◽  
Cervical Cancer Screening ◽  
Digital Image Processing ◽  
Digital Image ◽  
Pap Smear ◽  
Support Vector ◽  
Computer Assisted ◽  
Screening System

Cervical cancer can be prevented by having regular screenings to find any precancers and treat them. The Pap test looks for any abnormal or precancerous changes in the cells on the cervix. However, the manual screening of Pap smear in the microscope is subjective with poorly reproducible criteria. Therefore, the aim of this study was to develop a computer-assisted screening system for cervical cancer using digital image processing of Pap smear images. The analysis of Pap smear image is important in the cervical cancer screening system. There were four basic steps in our cervical cancer screening system. In cell segmentation, nuclei were detected using a shape-based iterative method, and the overlapping cytoplasm was separated using a marker-control watershed approach. In the features extraction step, three important features were extracted from the regions of segmented nuclei and cytoplasm. RF (random forest) algorithm was used as a feature selection method. In the classification stage, bagging ensemble classifier, which combined the results of five classifiers—LD (linear discriminant), SVM (support vector machine), KNN (k-nearest neighbor), boosted trees, and bagged trees—was applied. SIPaKMeD and Herlev datasets were used to prove the effectiveness of our proposed system. According to the experimental results, 98.27% accuracy in two-class classification and 94.09% accuracy in five-class classification was achieved using the SIPaKMeD dataset. When the results were compared with five classifiers, our proposed method was significantly better in two-class and five-class problems.

Exploring an Appropriate Method of Cervical Cancer Screening in Rural China

2019 ◽  
Vol 31 (7) ◽  
pp. 652-658 ◽  
Author(s):  
Xiaosong Zhang ◽  
Gengli Zhao ◽  
Hui Bi ◽  
Min Zhou ◽  
Xueyin Wang ◽  
...  
Keyword(s):  
Cervical Cancer ◽  
Cancer Screening ◽  
Cervical Cancer Screening ◽  
Rural Areas ◽  
Detection Rate ◽  
Rural China ◽  
Intraepithelial Neoplasia ◽  
Screening Strategy ◽  
Significant Difference ◽  
Positive Rate

Background. To explore the feasibility of careHPV (human papillomavirus) with cytology triage as a cervical cancer screening in rural areas of China. Methods. A total of 7138 women aged 35 to 64 years were divided into 2 groups. Women in careHPV group (n = 3536) underwent careHPV and 288 positive subjects underwent cytology, of which 65 women were ≥ASC-US (atypical squamous cells of undetermined significance). Women in the cytology group (n = 3602) underwent cytology and 111 women were ≥ASC-US. All subjects with ≥ASC-US were referred to colposcopy and biopsy. Results. The average age of subjects was 48.2 ± 7.8 years. In the careHPV group, the HPV-positive rate was 8.1%. The detection rate of ≥ASC-US was 1.8% in the careHPV group and 3.1% in the cytology group ( P = .001). There was no significant difference in detection rate of ≥CINII (cervical intraepithelial neoplasia) in the careHPV group (0.7%) and the cytology group (0.6%; P = .416). In addition, to identify 1 case ≥CINII, an average of 2.6 colposcopies were needed in the careHPV group, and 5.3 colposcopies were performed to diagnose 1 case ≥CINII in the cytology group. Conclusions. careHPV with cytology triage offered similar efficiency in identifying abnormalities of CINII and above compared with cytology screening. With the reduced requirement for cytology testing and colposcopy, careHPV may be a more favorable cervical cancer screening strategy in areas of China where there is a lack of cytology services.

scholarly journals Cervical screening in foreign domestic workers in Singapore

2021 ◽  
Vol 50 (2) ◽  
pp. 135-140
Author(s):  
Julia CL Eng ◽  
Joyce BT Er ◽  
Carrie SY Wan ◽  
YK Lim ◽  
Ida Ismail-Pratt ◽  
...  
Keyword(s):  
Cervical Cancer ◽  
Cancer Screening ◽  
Cervical Cancer Screening ◽  
Domestic Workers ◽  
Pilot Program ◽  
High Grade ◽  
Pilot Programme ◽  
Positive Rate ◽  
Foreign Domestic Workers ◽  
Hpv Screening

Introduction: Globally, cervical cancer is the fourth most common cancer in women, with about 85% occurring in low-middle income countries (LMIC) and an age-standardised incidence rate of more than 15 per 100,000. It is largely preventable through HPV vaccination and cervical cancer screening. In Singapore, 18% of the foreign domestic workforce hail from Indonesia, the Philippines, Myanmar, and India. However, there is no data on preinvasive cervical disease and cervical cancer in foreign domestic workers (FDWs) and the aim of this pilot programme is to determine the baseline screen positive rate of high-grade intraepithelial in this population. Methodology: A total of 322 FDWs were offered HPV screening through the Helping Our Helper (HOH) pilot programme. Data from this pilot program was analysed and reported using simple descriptive statistics. Results: Out of the 322 FDWs who registered for HPV screening, 68.6% participated. There was a 22.2% screen-positive rate; 10% of those who screened positive for high-risk HPV had histologically confirmed high-grade cervical intraepithelial neoplasia. This result is similar to other data on cervical cancer screening in Singaporeans. This pilot project screened less than 1% of the eligible FDWs in Singapore. Discussion: The findings of this pilot program suggest that there is public health value in providing cervical cancer screening to FDWs. Improving cervical cancer screening by increasing awareness and including routine cervical cancer screening as part of the employment medical examination should be studied. Keywords: Cervical cancer, CIN 2, colposcopy, HPV, HSIL, LSIL

scholarly journals STUDI KOMPARASI METODE SEGMENTASI PARU-PARU PADA CITRA CT-SCAN AKSIAL

2019 ◽  
Vol 7 (3) ◽  
pp. 283
Author(s):  
Made Satria Wibawa
Keyword(s):  
Ground Truth ◽  
True Positive Rate ◽  
Computer Assisted ◽  
True Positive ◽  
Ct Diagnosis ◽  
True Negative ◽  
Fuzzy C Means ◽  
True Negative Rate ◽  
Positive Rate ◽  
Assisted Diagnosis

Kanker paru dapat diobati jika diagnosis dini dilakukan. Diagnosis dapat dilakukan menggunakan modalitas citra Computed Tomography (CT). Diagnosis kanker paru melalui citra CT dilakukan oleh tenaga medis. Untuk membantu diagnosis kanker, tenaga medis dapat dibantu dengan Computer Assisted Diagnosis (CAD). Dalam CAD, tahapan pertama yang paling penting adalah segmentasi citra paru-paru. Penelitian ini melakukan studi komparasi metode segmentasi citra CT paru-paru. Terdapat tiga metode segmentasi yang digunakan, yaitu Otsu, K-Means dan Fuzzy C-Means. Proses evaluasi menggunakan metrik akurasi, true negative rate dan true positive rate. Berdasarkan nilai yang diperoleh dari ketiga parameter evaluasi tersebut, ketiga metode segmentasi dapat memberikan hasil segmentasi yang mendekati citra ground truth. Namun, dilihat dari sebaran hasil nilai ketiga parameter evaluasi yang didapatkan dari seluruh citra, metode Otsu sedikit lebih unggul dibandingkan metode K-Means dan Fuzzy C-Means.

Sign in / Sign up

Export Citation Format

Share Document