CERVICAL CANCER PREVENTION
The World Health Organization (WHO) recommends a screen-and-treat approach for cervical precursor lesions in low- and middle-income countries (LMICs). Currently, visual inspection with acetic acid (VIA) is used for screening, followed by ablation if a lesion is identified. There are significant challenges with current approaches to screen-and-treat strategies for cervical cancer. First, VIA performance has poor specificity and depends on provider training and experience. Second, speculum use is a significant factor in women’s resistance to undergoing cervical cancer screening, largely due to anxiety, pain, and/or vulnerability during the procedure. Third, the lack of experts in the field makes interpretation of cervical images challenging. Our vision is to create a new strategy for wide scale cervical cancer surveillance in LMICs. Our ultimate goal is to bring low-cost, high quality interventions into primary care settings, enabling providers to implement an effective screen-and-treat model that is on par with the conventional three-visit model currently used in high-resource settings.
Colposcopy directed cervical biopsy is the standard confirmatory test for women with positive screening results. Colposcopes are expensive and require referral to a facility that can house this machine, a trained colposcopist who can interpret the images, and a pathologist and laboratory that can process and read the biopsy. These factors make colposcopy and biopsy inaccessible to the many women in LMICs who are at greatest risk for developing cervical cancer. By reconfiguring the colposcope into the form factor of a tampon, it can be inserted into the vaginal canal to provide a close-up view of the cervix at the price point of an SLR digital camera. Hand-held devices, such as the intraoral scope used to image teeth, nicely demonstrate this idea of bringing the camera close to the surface it is imaging. This conceptual idea is what led to the development of the Pocket Colposcope which brings the capabilities of specialists to the primary care setting and doubles as both a screening (in lieu of VIA) and as a diagnostic colposcope.
(A) Pocket Colposcope. (B) Representative acetic acid images acquired with a standard-of-care and Pocket Colposcope. (C) The Pocket Colposcope has been validated against standard of care on more than 1,000 women at hospitals in various partner sites. Numbers represent the number of women screened with the Pocket colposcope at each site.
When the high-quality images obtained with the Pocket colposcope are coupled with machine learning algorithms, expert colposcopy can be implemented at the point-of-care setting, enabling community health providers in the front lines to be able to make effective decisions that maximize value and minimize cost. We have demonstrated how effective this strategy can be using a machine-learning framework trained on images of the cervix obtained with the Pocket Colposcope. With increases in computational power, complex machine learning algorithms can now be easily run on a variety of portable devices, including smart phones and tablets, making it easy to envision how a smart phone coupled with the Pocket Colposcope would truly bring cervical cancer screening and/or diagnosis into the hands of community health providers.
High-quality images of the cervix obtained with the Pocket Colposcope. These images are used in our machine learning algorithms in make effective diagnoses. This is superior to the average values achieved by three expert physicians on the same data set.
Another barrier to screening is the duckbill speculum. This is a significant factor in women avoiding cervical cancer screening, largely due to anxiety, fear, discomfort, pain, embarrassment, and/or vulnerability during the procedure. The speculum is also a cause of discomfort for women with vaginismus, which involves involuntary tightening of the vagina often due to sexual abuse. LMICs typically have the highest sexual violence rates worldwide and also have the highest rate of cervical cancer incidence and mortality. Rendering the screening process speculum-free when women FIRST enter into the care cascade would allow the majority of women who do not have abnormalities to be screened without a speculum and would provide an opportunity to educate women who require referral about the need for speculum-based treatment once they are in the health system.
We have recently developed a very simple mechanical inserter without any moving parts that resembles the shape of a Calla Lily and has a stem for introducing the Pocket colposcope. It has a port for acetic acid or Lugol’s iodine application. Further, the Calla Lily tip is designed to gently nudge the cervix, which is often titled away from the vaginal canal, to the center of the field of view for image capture. The inserter maintains the tip of the Pocket colposcope about 35 mm away from the ectocervix, so that its full surface can be imaged in a single snapshot, but at a much smaller distance from the ectocervix than would be the case if a conventional colposcope and speculum were used. A smart phone attached to the colposcope provides navigation with live video and can capture and images of the ectocervix.
The Callascope device and an image of the cervix displayed on a tablet. Today more than 75 women have had pelvic exams with the Callscope and not only can women do this themselves but the Callascope removes the discomfort associated with the conventional speculum exam.
Asiedu MN, Agudogo JS, Dotson ME, Krieger MS, Schmitt JW, Huchko M, Suneja G, Proeschold-Bell RJ, Smith JS, Jenson D, Hogan W. A Novel, Versatile Speculum-free Callascope for Clinical Examination and Self-Visualization of the Cervix. bioRxiv. 2019 Jan 1:618348.
Asiedu MN, Simhal A, Chaudhary U, Mueller JL, Lam CT, Schmitt JW, Venegas G, Sapiro G, Ramanujam N. Development of algorithms for automated detection of cervical pre-cancers with a low-cost, point-of-care, Pocket Colposcope. IEEE Transactions on Biomedical Engineering. 2018 Dec 18.
Mueller JL, Lam CT, Dahl D, Asiedu MN, Krieger MS, Bellido‐Fuentes Y, Kellish M, Peters J, Erkanli A, Ortiz EJ, Muasher LC. Portable Pocket colposcopy performs comparably to standard‐of‐care clinical colposcopy using acetic acid and Lugol's iodine as contrast mediators: an investigational study in Peru. BJOG: An International Journal of Obstetrics & Gynaecology. 2018 Sep;125(10):1321-9.
Asiedu MN, Simhal A, Lam CT, Mueller J, Chaudhary U, Schmitt JW, Sapiro G, Ramanujam N. Image processing and machine learning techniques to automate diagnosis of Lugol's iodine cervigrams for a low-cost point-of-care digital colposcope. InOptics and Biophotonics in Low-Resource Settings IV 2018 Feb 13 (Vol. 10485, p. 1048508). International Society for Optics and Photonics.
Lam CT, Mueller J, Asma B, Asiedu M, Krieger MS, Chitalia R, Dahl D, Taylor P, Schmitt JW, Ramanujam N. An integrated strategy for improving contrast, durability, and portability of a Pocket Colposcope for cervical cancer screening and diagnosis. PloS one. 2018 Feb 9;13(2):e0192530.
Mueller JL, Lam CT, Kellish M, Peters J, Asiedu M, Krieger MS, Gallagher JE, Erkanli A, Ortiz EJ, Muasher LC, Taylor PT. Clinical evaluation of a portable pocket colposcope for cervical cancer screening in the United States, Perú, and Tanzania. In2017 IEEE Healthcare Innovations and Point of Care Technologies (HI-POCT) 2017 Nov 6 (pp. 117-120). IEEE.
Asiedu MN, Agudogo J, Krieger MS, Miros R, Proeschold-Bell RJ, Schmitt JW, Ramanujam N. Design and preliminary analysis of a vaginal inserter for speculum-free cervical cancer screening. PloS one. 2017 May 31;12(5):e0177782.
Mueller JL, Asma E, Lam CT, Krieger MS, Gallagher JE, Erkanli A, Hariprasad R, Malliga JS, Muasher LC, Mchome B, Oneko O. International Image Concordance Study to Compare a Point of Care Tampon Colposcope to a Standard-of-Care Colposcope. Journal of lower genital tract disease. 2017 Apr;21(2):112.
Lam CT, Krieger MS, Gallagher JE, Asma B, Muasher LC, Schmitt JW, Ramanujam N. Design of a novel low cost point of care tampon (POCkeT) colposcope for use in resource limited settings. PloS one. 2015 Sep 2;10(9):e0135869