How Artificial Intelligence can help in early stage breast cancer detection

Read Article

By Prashant Akhawat, COO, Telerad Tech

Artificial Intelligence (AI) is breaking the boundaries of traditional software development into complex neural networks, data analytics, and computer vision algorithms to achieve human cognition of data. AI is all set to bring an exemplary shift across various industry verticals and is powered by the ever-evolving availability of data apart from the swift advancement of data mining and inference analytics.

In medical imaging, there is immense opportunity to improve patient care using AI. Not only will AI boost the efficiency of radiologists, it will also identify exams with critical findings and help radiologists to prioritize the cases that need immediate attention. It will offer ways to bridge gaps in accessibility, availability, point in time care, improved detection, and diagnosis.

Today, the healthcare industry is facing a global shortage of radiologists wherein radiologist is to patient ratio in USA 1:10,000; India 1:100,000; Bangladesh 1:10,00,000 and even worse situation in the African region. Governments across countries are under huge cost pressure to provide efficient and affordable healthcare.

Early breast cancer diagnosis with AI
Gradually, AI is being recognized as one of the major elements in the healthcare space. Breast cancer is one of the most frequently diagnosed cancer and is the leading cause of death among women worldwide. Amongst the cancer category, breast cancer ranks one among Indian females with a rate as high as 25.8 per 100,000 women and mortality of 12.7 per 100,000 women, according to the health ministry.

In India, 1 in 22 urban and 1 in 32 rural women are diagnosed with breast cancer with the mortality rate of 50%. The most common cause of death is late detection of breast cancer. Taking the global data, every 19th second a woman is diagnosed with breast cancer and every 74th second someone dies from breast cancer. The mortality rate in developing nations is much higher than developed nations. Approximately 425,000 women died from breast cancer around the world in the year 2010, at this rate 10.6 million women will die from breast cancer during next 25 years. This calls for urgent improvement in diagnosis and management of the disease in all areas of the world.

In our recent breakthrough innovations in our AI Lab, we developed AI Algorithm – MammoAssist to analyze Mammography exams. This algorithm can help in early stage breast cancer detection and generate a full-fledged detailed report with all critical clinical findings such as Micro Calcification, Macro Calcification, Clustered Calcification, Lesions & Lymph Node Analysis, Architectural Distortion, Bilateral Asymmetry, Breast Parenchymal Composition, Size, Shape, Location and Density analysis with BIRADS (Breast Imaging Reporting & Data System) score. It enhances the ability of a radiologist to report cases with very high accuracy, efficiency and provides a standard interface to communicate with healthcare systems through industry standard protocols.

This can also can help in reducing the government’s burden by improving patient care specifically in mass screening programs. It will help the radiologists to identify if there is a high risk of malignancy like hidden tumor – distinguishing which lesion would become more cancerous and should be removed or which lesion is benign.

Therefore, AI’s decision-making algorithms with high specificity, sensitivity and accuracy will help in identifying true negatives and will highlight positive cases with critical findings and BIRADS classification in turn leading to improved patient care.

Medical Image Analytics Algorithms are advancing towards detection, diagnosis, and prognosis and sooner we will be observing software taking a leap in efficiently recommending a treatment plan.

At Telerad Tech, we are targeting to release 40+ algorithms by year 2019 and already achieved a significant breakthrough’s in Brain Strokes, Chest X-Ray, Fractures, Tuberculosis, Ureteric Stones and Cardiovascular diseases.

In the near future, AI and deep learning algorithms will be virtualized brains behind Medical Image Analytics and will help in achieving very high levels of accuracy in diagnosing various clinical conditions. AI has the potential to provide full support to government’s mass screening programs as well as to uplift underdeveloped healthcare systems.

If you have an interesting article / experience / case study to share, please get in touch with us at [email protected]

Advertisement