Breast cancer screening is a crucial tool in early detection, enabling timely intervention and improving survival rates. While women at high risk due to genetic mutations or previous chest irradiation are recommended for regular MRI screening alongside mammography, those with an intermediate risk often receive only mammography. This group, which includes women with a family history of breast cancer but no confirmed genetic mutation, dense breast tissue or prior high-risk lesions, may still benefit from MRI screening. However, the limited availability and high costs of MRI mean that it is not widely accessible for this category of patients.
Artificial intelligence is emerging as a potential solution to improve screening efficiency by identifying which intermediate-risk women are most likely to benefit from MRI. By applying AI to mammography, researchers aim to refine patient selection for MRI, ensuring optimal resource allocation while maintaining high cancer detection rates. This approach has the potential to bridge the gap in screening strategies, offering a cost-effective way to enhance early detection in intermediate-risk populations.
The Challenge of Screening Women with Intermediate Risk
Screening strategies for women with intermediate breast cancer risk remain inconsistent worldwide. While high-risk women are prioritised for MRI, those with an intermediate risk are generally offered mammography alone, despite its limitations in detecting certain cancers, particularly in dense breast tissue. Evidence suggests that MRI screening can be effective for this group, as it detects cancers that mammography may miss. Many of these cancers are smaller and less likely to have spread to lymph nodes at the time of diagnosis, making early intervention more effective.
Despite these advantages, widespread MRI screening for intermediate-risk women remains impractical due to resource constraints. MRI is expensive, time-consuming and not universally accessible. Many healthcare systems lack the capacity to accommodate additional MRI screenings beyond those recommended for high-risk groups. Furthermore, the clinical utility of MRI for intermediate-risk women has been debated, leading to variations in screening guidelines across countries. As a result, decisions regarding MRI screening often depend on individual physician recommendations or patient preferences rather than a standardised approach.
AI in Mammography: A New Screening Paradigm
AI is increasingly being used to enhance breast cancer detection by improving the accuracy of mammography assessments. AI-driven models can analyse mammograms and assign a case-based cancer suspicion score, which can help determine whether a woman should undergo supplemental MRI. This triage approach is designed to ensure that MRI is reserved for those who are most likely to benefit from additional screening, reducing unnecessary procedures while maintaining a high detection rate.
Recent studies have demonstrated the effectiveness of AI in identifying women with a higher probability of developing breast cancer within the intermediate-risk population. AI models have achieved a strong predictive performance, particularly among women with a personal history of breast cancer. In research trials, AI successfully detected 84% of breast cancers in intermediate-risk women when used as a selection tool for MRI screening. Notably, it identified 68% of mammographically occult cancers—cancers that would not have been detected through mammography alone.
By refining the selection process, AI can reduce the overall number of MRIs performed while still identifying the majority of cancers. This approach offers a potential solution to the current dilemma faced by healthcare providers: balancing the need for improved early detection with the limited availability of MRI. Additionally, AI-based selection may help standardise screening decisions, reducing reliance on physician discretion and ensuring that MRI is used efficiently and equitably.
Balancing Costs, Efficiency and Early Detection
One of the major challenges in implementing MRI for intermediate-risk screening is the associated financial and logistical burden. While MRI is a powerful diagnostic tool, its widespread use is constrained by cost and resource limitations. AI-based triaging offers a way to optimise these resources by ensuring that MRI is prioritised for women with a higher likelihood of developing breast cancer.
Current screening protocols often result in unnecessary MRI examinations for women who may not require them while leaving others without access to MRI despite potential benefits. AI-driven selection has the potential to improve efficiency by focusing on those who are most at risk, thereby reducing costs and improving healthcare outcomes. By selecting only 50% of intermediate-risk women for MRI, AI can still identify a substantial majority of cancers, offering a practical solution to the problem of overburdened screening services.
Despite its potential, AI implementation in breast cancer screening is not without challenges. Further validation is required to refine AI models and ensure their reliability across diverse populations. Additionally, healthcare providers must address potential concerns regarding AI’s role in decision-making and ensure that AI-based triaging is used as a tool to support, rather than replace, clinical judgement.
AI is set to transform breast cancer screening for women with intermediate risk by improving detection rates while optimising the allocation of MRI resources. By using AI to analyse mammograms and identify women most likely to benefit from MRI, healthcare providers can enhance early detection without overwhelming MRI capacity. This approach addresses both the clinical need for improved screening and the financial and logistical barriers that currently limit MRI availability.
While challenges remain in refining AI models and ensuring their integration into clinical practice, the evidence suggests that AI-driven triage could be a crucial step towards more effective and equitable breast cancer screening. By reducing unnecessary imaging and improving the detection of mammographically occult cancers, AI could play a vital role in the future of breast cancer care, offering a practical and cost-effective strategy for optimising screening programmes worldwide.
Source: Radiology
Image Credit: iStock
