Research from Sweden shows that using AI in breast cancer diagnosis can improve early detection and ease the workload for radiologists. The study, published in Lancet, screened over 100,000 women as part of Sweden’s national program. This has special relevance for India, where breast cancer makes up about 26.6% of all cancer cases in women, and many areas lack trained radiologists.
Here’s what the study found: Women aged 40 to 74 were screened, with their mammograms reviewed by at least two radiologists. AI scored results on a scale of 1 to 10, with scores of 1-7 classified as low risk, 8-9 as intermediate, and 10 as high risk. Those classified as low to intermediate risk were reviewed by one radiologist, while high-risk patients had their images examined by two radiologists.
AI detection proved effective, identifying 6.4 cases per 1,000 women compared to 5 per 1,000 using standard methods. It identified 270 cases of invasive cancer, compared to 217 with traditional screening. Moreover, AI did not significantly increase false positives and reduced the workload of radiologists by about 44.2%.
Why is this important? Breast cancer is one of the most common cancers in women globally, and regular mammography has been shown to lower deaths from the disease. Although researchers still need to confirm if AI can further reduce mortality rates, their findings suggest it can help detect more cancers early.
In India, AI could help address the challenges of human error and inconsistencies in skills across different healthcare facilities. Dr. Abhishek Shankar, an oncologist at AIIMS-Delhi, points out that using AI could standardize cancer diagnosis, offering support to doctors with varying levels of expertise.
However, training AI models requires data specific to India, as breast cancer types can differ significantly from those in Western countries. There are several challenges to address:
- Breast cancer causes about 98,000 deaths annually in India, with women developing it on average 10 years earlier than in the West. The typical age for diagnosis in the West is around 50, while in India, it’s more common between 40 and 50.
- Detecting breast cancer early in younger women remains an issue, as mammography is most effective for those over 50 due to breast tissue differences.
- India sees higher occurrences of aggressive cancer forms like triple-negative breast cancer and HER2-positive cancer.
Are such technologies available in India? Yes! An Indian-developed AI screening test called Thermalytix was recently tested in Punjab. This innovative approach screened over 15,000 women without using traditional mammography, which can be costly and less accessible. Instead, Thermalytix detects slight temperature changes in the breast to assess cancer risk. The portable device requires minimal infrastructure. In this study, 460 women were identified as high risk, with 27 confirmed to have cancer through follow-up imaging.
These advancements highlight the potential of AI in improving breast cancer diagnosis and making screenings more accessible, especially in countries with high cancer burdens.
Source link
breast cancer, breast cancer screening, ai and breast cancer, ai to screen breast cancer, breast cancer news, caner, cancer screening, health news, lancet study, indian express
