Clinical evaluation of a portable pocket colposcope for cervical cancer screening in the United States, Perú, and Tanzania

Mueller, Jenna L.; Lam, Christopher T.; Kellish, Mark; Peters, Jenna; Asiedu, Mercy Nyamewaa; Krieger, Marlee; Gallagher, Jennifer E.; Erkanli, Alaattin; Ortiz, Ernesto; Muasher, Lisa C.; Taylor, Peyton T.; Oneko, Olola; Venegas, Gino; Schmitt, John W.; Ramanujam, Nimmi

doi:10.1109/hic.2017.8227598

Cited by 5 publications

(5 citation statements)

References 2 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The performance of the algorithm was comparable to physicians. In recent high-quality studies, the sensitivity and specificity to detect HSIL varied greatly, ranging between 33%-93% and 53%-95%, respectively (36)(37)(38)(39)(40), which was comparable to the result of our study.…”

Section: Discussionsupporting

confidence: 88%

Computer-aided diagnosis of cervical dysplasia using colposcopic images

You

Xue

et al. 2022

Front. Oncol.

View full text Add to dashboard Cite

Backgroundcomputer-aided diagnosis of medical images is becoming more significant in intelligent medicine. Colposcopy-guided biopsy with pathological diagnosis is the gold standard in diagnosing CIN and invasive cervical cancer. However, it struggles with its low sensitivity in differentiating cancer/HSIL from LSIL/normal, particularly in areas with a lack of skilled colposcopists and access to adequate medical resources.Methodsthe model used the auto-segmented colposcopic images to extract color and texture features using the T-test method. It then augmented minority data using the SMOTE method to balance the skewed class distribution. Finally, it used an RBF-SVM to generate a preliminary output. The results, integrating the TCT, HPV tests, and age, were combined into a naïve Bayes classifier for cervical lesion diagnosis.Resultsthe multimodal machine learning model achieved physician-level performance (sensitivity: 51.2%, specificity: 86.9%, accuracy: 81.8%), and it could be interpreted by feature extraction and visualization. With the aid of the model, colposcopists improved the sensitivity from 53.7% to 70.7% with an acceptable specificity of 81.1% and accuracy of 79.6%.Conclusionusing a computer-aided diagnosis system, physicians could identify cancer/HSIL with greater sensitivity, which guided biopsy to take timely treatment.

show abstract

Section: Discussionsupporting

confidence: 88%

Computer-aided diagnosis of cervical dysplasia using colposcopic images

You

Xue

et al. 2022

Front. Oncol.

View full text Add to dashboard Cite

show abstract

“…Pathology labels were retrospectively obtained from previous clinical studies from all patients with a visually appearing abnormal area [ 14 , 36 ]. Standard of care clinical diagnoses were made from images obtained with either a standard clinical colposcope or a digital camera and not with the investigational device.…”

Section: Methodsmentioning

confidence: 99%

Multicontrast Pocket Colposcopy Cervical Cancer Diagnostic Algorithm for Referral Populations

et al. 2022

View full text Add to dashboard Cite

Objective and Impact Statement. We use deep learning models to classify cervix images—collected with a low-cost, portable Pocket colposcope—with biopsy-confirmed high-grade precancer and cancer. We boost classification performance on a screened-positive population by using a class-balanced loss and incorporating green-light colposcopy image pairs, which come at no additional cost to the provider. Introduction. Because the majority of the 300,000 annual deaths due to cervical cancer occur in countries with low- or middle-Human Development Indices, an automated classification algorithm could overcome limitations caused by the low prevalence of trained professionals and diagnostic variability in provider visual interpretations. Methods. Our dataset consists of cervical images (n=1,760) from 880 patient visits. After optimizing the network architecture and incorporating a weighted loss function, we explore two methods of incorporating green light image pairs into the network to boost the classification performance and sensitivity of our model on a test set. Results. We achieve an area under the receiver-operator characteristic curve, sensitivity, and specificity of 0.87, 75%, and 88%, respectively. The addition of the class-balanced loss and green light cervical contrast to a Resnet-18 backbone results in a 2.5 times improvement in sensitivity. Conclusion. Our methodology, which has already been tested on a prescreened population, can boost classification performance and, in the future, be coupled with Pap smear or HPV triaging, thereby broadening access to early detection of precursor lesions before they advance to cancer.

show abstract

“…Six publications were ineligible because test accuracy data were missing [19][20][21][22][23][24]. Seven publications were based on study populations already included in our analysis [17,[25][26][27][28][29]. We have presented a complete list of excluded full-text assessments and the reasons for their exclusion in Additional file 3.…”

Section: Literature Search Overviewmentioning

confidence: 99%

Screening test accuracy of portable devices that can be used to perform colposcopy for detecting CIN2+ in low- and middle-income countries: a systematic review and meta-analysis

et al. 2020

View full text Add to dashboard Cite

Background Portable devices that can be used to perform colposcopy may improve cervical cancer screening in low- and middle-income countries (LMIC) where access to colposcopy is limited. The objective of this study was to systematically review the diagnostic test accuracy (DTA) of these devices for the detection of cervical intraepithelial neoplasia grade 2 or higher (CIN2+). Methods In accordance with our protocol (Prospero CRD42018104286), we searched Embase, Medline and the Cochrane Controlled Register of Trials up to 9/2019. We included DTA studies, which investigated portable devices with moderate-to-high optical magnification (≥ 6×) for colposcopy, as described in the manual for Colposcopy and Treatment by the International Agency for Research on Cancer, with a histopathological reference standard. We used the QUADAS-2 tool to assess study quality. We examined results for sensitivity and specificity in paired forest plots, stratified by stages in the clinical pathway. We pooled estimates of test accuracy for the index test, used as an add-on to other tests, using a bivariate random-effect model. Results We screened 1737 references and assessed 239 full-text articles for eligibility. Five single-gate DTA studies, including 2693 women, met the inclusion criteria. Studies evaluated two devices (Gynocular™ and Pocket) at different stages of the screening pathway. In three studies, which used the index test in an add-on capacity in 1273 women, we found a pooled sensitivity of 0.79 (95% CI 0.55–0.92) and specificity of 0.83 (95% CI 0.59–0.94). The main sources of bias were partial verification, incorporation and classification bias. Conclusion Few studies have evaluated portable devices able to perform colposcopy, so their accuracy for the detection of CIN2+ remains uncertain. Future studies should include patient-relevant and long-term outcomes, including missed cases, overtreatment, residual and recurrent disease. To meet the challenge of eliminating cervical cancer in LMIC, methods for visual assessment of the cervix need urgent redress.

show abstract

Clinical evaluation of a portable pocket colposcope for cervical cancer screening in the United States, Perú, and Tanzania

Cited by 5 publications

References 2 publications

Computer-aided diagnosis of cervical dysplasia using colposcopic images

Computer-aided diagnosis of cervical dysplasia using colposcopic images

Multicontrast Pocket Colposcopy Cervical Cancer Diagnostic Algorithm for Referral Populations

Screening test accuracy of portable devices that can be used to perform colposcopy for detecting CIN2+ in low- and middle-income countries: a systematic review and meta-analysis

Contact Info

Product

Resources

About