Big Breakthrough in Cervical Cancer Market: A New Method for Detection of Precursors to Cervical Cancer to Help Local Pathologists
Posted On April 13, 2021
In third world countries, cervical cancer cases have considerably dropped, and appreciation for this should be given to national screening programs. They help in the detection of cell abnormalities and human papillomavirus (HPV) in cervical samples. Even with the statistics looking positive, experts in the medical field believe that total cervical cancer cases would increase globally in the coming years. The main reason for this lies mainly in the shortages of screening resources and HPV vaccines in poor countries. Hence, the most recent need is to develop Innovative Diagnostic Solutions which analyze the local conditions and constraints. This is important if the number of diagnoses has to be increased around the globe.
Researchers might have found a solution in this regard. They recently stated that with the usage of Artificial Intelligence (AI) and mobile digital microscopy, it might be possible to create screening tools. These sorts of tools have the ability to detect precursors to cervical cancer in the settings with limited resources. This would be a significant breakthrough in Cervical Cancer Market as this method would facilitate effective discovery and treatment of precursors to cervical cancer. This would be more relevant for low-income countries, and wherein there is a presence of less advanced laboratory equipment and less number of pathologists.
The team trained AI systems enough to identify cell abnormalities in the cervix region; early detection of the same can be easily treated. Between September 2018 and September 2019, about 740 smear samples were taken from women at a rural clinic, Kenya. The received samples were digitalized through a portable scanner and uploaded to a cloud-based deep-learning system (DLS) by using mobile networks. Half of the smears were used by researchers to train the AI to identify distinct pre-cancerous lesions, and the other half was used to judge its accuracy.
The received digital assessment by AI was then compared with the physical one made by two independent pathologists. The study concluded that both the assessments were similar in nature. The newly developed DLS was noted to have a sensitivity of 96-100% regarding the identification of patients with pre-cancerous lesions. None of the patients that had high-grade lesions were wrongly assessed to be falsely positive. In terms of identifying smears that did not have lesions, a similar assessment was made by DLS as the two pathologists in about 78-85% of cases.
The team is optimistic that this method would be helpful in excluding a significant amount of smears. This would greatly help free up local experts' time for examining only those patients that stick out. Nonetheless, before this method can be confidently used, there needs to be further research with a sample base that is larger and more diverse.