Growing up in Ghana, Mercy Nyamewaa Asiedu became increasingly aware of the subpar health system and associated high mortality rates that affected her family and friends. She soon realized that a contributing factor was the lack of effective healthcare technologies due to equipment that did not work, inadequate knowledge to operate machinery, and machines that could not be fixed after a breakdown due to lack of availability of spare parts or technicians. Funds from the Prize will be split between two projects.
The plan for this project is to use mobile technologies and artificial intelligence (AI) to bridge gaps in healthcare access and data insights in sub-Saharan Africa. Given the high penetration of smart mobile phones and mobile data in the region, Mercy’s team seeks to develop a mobile application that (1) provides patients with chronic diseases the tools to store and track their medical activities, (2) connects verified providers to patients on the app via SMS, phone, video calls, and in-person visits, with simple country-specific payment methods such as mobile-money and, (3) uses patient data to provide AI-generated personalized health recommendations.
With user approval, the team hopes to use the deidentified patient health data stored, to better understand chronic health manifestations and trends in the context of sub-Saharan Africa. The team will pilot the mobile application with a focus on diabetes but plans to expand to other chronic health conditions including heart disease, cancer, and sickle cell disease.
Learn more about GAPhealth.
Meeting the Challenge
Over the past two decades, the drastic increase in noncommunicable diseases in Africa has been driven by: (1) the great strides made in the areas of communicable diseases leading to lowered risk of infectious diseases with increased life expectancy; and, (2) increased westernized diet and lifestyle including unhealthy fast, processed and sugary food diets, reduced physical activity and air pollution, leading to higher risks for diabetes, cardiovascular and respiratory diseases.
The World Health Organization has projected non-communicable diseases (NCDs) to become the leading cause of death in sub- Saharan Africa by 2030. Unless actions are taken, this growth will overburden health systems in the region. Diabetes, specifically, is one of the leading causes of NCD-related deaths and disability-adjusted life years in the area. Nineteen million people in sub-Saharan Africa have been diagnosed with and are living with diabetes, with greater than 60% remaining undiagnosed. This number is expected to increase by 147% to 49 million by 2045, leading both the WHO and Lancet to issue urgent calls to action in the area.
Two significant limitations in diabetes management are: (1) The broken care pipeline/linkages between screening/prevention, and specialist care. Limited diabetes specialists leads to difficulty for patients to connect with physicians, for consultations, prescriptions, etc. This has been further exacerbated by the lack of telehealth and the COVID pandemic, which disproportionately causes adverse outcomes in people with diabetes, creating challenges for in-person visits; and, (2) Lack of quality data to understand manifestations of diabetes among people in sub-Saharan Africa in order to predict trends and create contextualized management guidelines for the region. This lack of data is partially due to the lack of a consolidated, widespread electronic health record system. In this era of machine learning and big data, having a well-consolidated database can truly make a difference in understanding and predicting disease trends.
The vision for this project is to develop high quality, low cost, interventions specifically for low-resource health facilities to address the shortcomings of current technological solutions for cervical cancer prevention and commercialize it in markets where it is needed the most. The proposal is to integrate innovative diagnostic algorithms into a new mobile software and perform clinical validation studies in Durham, NC, Peru, and Kenya.
Mercy’s team has developed a low-cost, patient-centered, speculum-free colposcope, called the Callascope. It has the form factor of a tampon and is comparable to a standard digital colposcope. The team has also developed a very simple, ergonomic introducer for the Callascope, expanding vaginal walls, and centering the cervix for accurate visualization without the speculum. A smartphone attached to the Callascope provides navigation with live video and image capture.
An algorithm based on cervix image features and machine learning techniques achieved a sensitivity and specificity, of 81.3% and 78.6% respectively, when used to distinguish cervical intraepithelial neoplasia (CIN+) relative to normal and benign tissues. The sensitivity and specificity of expert physicians reviewing the same data were 77% and 51%, respectively.
Learn more about the Calla Health Foundation.
Meeting the Challenge
Invasive cervical cancer (ICC) affects the lives of half a million women worldwide each year, and over half these women die. The majority of these deaths occur in low and middle-income countries (LMICs). By contrast, in the US, cervical cancer incidence and mortality have decreased by 70% over the last 70 years due to early wide-spread screening, but the diagnosis and treatment model is not practical to implement in medically underserved regions due to the lack of resources to procure, implement, and maintain the technologies and the need for multiple hospital visits.
To address these issues, the recommendation for cervical cancer screening in low-resource settings is to use visual inspection with acetic acid (VIA), followed by treatment with low-cost
options like cryotherapy or thermocoagulation for pre-cancer. Making screening affordable, accessible, and accurate generates the biggest return on investment, as cervical cancer is an extremely slow-growing disease (15+ years); if caught early, it can be treated well before it progresses to invasive disease with a 95% survival rate. However, there are several barriers associated with VIA. The first is that VIA is a low-tech version of colposcopy; the naked eye is used to visualize the cervix, typically by non-experts. This leads to a high level of diagnostic variability. The absence of image capture further exacerbates this problem as there is minimal quality control to improve provider skills over time. Additionally, the absence of
magnification diminishes accuracy, particularly for microscopic lesions. But perhaps the most significant barrier to screening is the duckbill speculum. This device has been shown to be a significant reason for women avoiding cervical cancer screening, due to anxiety, fear, discomfort, pain, embarrassment, and/or vulnerability during the procedure.
