The use of Artificial Intelligence (AI) has the potential to transform all industrial sectors. Intelligent CIO talks to Pierre Brunswick, CEO, NeuroMem, about how technological advancements are being utilised in the healthcare sector offering the potential to save thousands, if not millions, of lives.
Efficiency in the healthcare sector has been increased through the use of pattern recognition and improved applications of neural processing.
Perhaps most significantly it has been utilised in the early detection of cervical cancer. This is one area where the use of Artificial Intelligence (AI) can prove especially beneficial as it avoids the laborious and time-consuming process of manually classifying pap smears for cervical cancer detection.
But the use of such technology even has the potential to detect and avoid an epidemic of disease across multiple continents or even worldwide.
The story of the application of AI in the healthcare sector is underway but we are surely only at the tip of the iceberg.
Can you explain how your solutions have been utilised in the early detection of cervical cancer?
Artificial intelligence (AI) is helping the healthcare sector to increase efficiency through the use of pattern recognition and improved applications of neural processing locally vis-à-vis moving data offsite to a cloud server.
One of the earliest applications was development by Dr Manan Suri, Assistant Professor with the Department of Electrical Engineering, Indian Institute of Technology-Delhi (IIT-Delhi), who has been using NeuroMem technology to develop proof of concept neuromorphic/AI applications in fields such as healthcare (to augment diagnosis of cervical cancer from pap smears tests available publicly).
Dr Suri and his team worked on an ultra-fast, efficient, low-power, neuromorphic hardware-based solution for cervical cancer diagnosis. Manual classification of pap smears for cervical cancer detection is a laborious and time-consuming process.
They trained a bio-inspired, dedicated ASIC on parameters of cell nuclei of ground truth images for normal/abnormal cell classification. They have been able to validate their approach on both single- and multi-cell images.
Given the number of cervical cancer suspect cases and the difference that a timely diagnosis can make, there is a strong need for development of intelligent, low-cost, portable, low-power preliminary diagnostic hardware.
Progress in the field of high-performance imaging hardware for smartphones has created a favorable pre-cursor for development of true portable mobile on-the-fly diagnostic systems. Bio-inspired neuromorphic hardware or artificial neural networks (ANNs) in hardware have proven to be promising in the field of image processing, speech recognition, pattern recognition and medical diagnostics.
Can you explain how the system is used to detect anomalies?
At NeuroMem, we truly believe that nonstop learning, and pattern recognition offered by our technology can become practical and ubiquitous only if it can rely on components inspired by the human brain (which we call neuromorphic memories), merging storage and neurons per cell, with massively parallel interconnected cells operating at low power.
Pattern recognition offers massively parallel computing, which means that no matter how big or small the dataset, it offers the sale latency by probing all the neurons simultaneously. If there is no answer, it means that in few nano seconds, there is an anomaly detection. Then the user can decide whether the system needs to learn or flag it as an anomaly that needs further action.
What are the main benefits of this process?
At the moment, most of the solutions (even the hardware ones) are relying on cloud-based software. Our Neuromorphic technology bypasses this and builds the pattern learning capabilities into the chip without the necessity of sending information to the cloud.
This means that there is no need to send private data to the cloud or to a server when a chip can protect you just as well.
Pattern recognition/anomaly detection also means that we do not need to have a pre-set database to start using our technology. The system can learn on the go, and because it is focused on pattern recognition and anomaly detection, it has a high rate of accuracy, does not intrude on privacy unless necessary and offers on-site functionality.
Are there any other areas of healthcare where the technology could be used to potentially save lives?
One of the key areas of focus for us is to work with research institutes and universities to create solutions that would help them detect and avoid pandemic cases. This is possible since the neurons in our chip can recognise millions of bacteria in a millisecond time frame, allowing the doctor to focus immediately on the one unknown (anomaly) virus that could prove to be the pandemic.
This would be useful for organisations and WHO to help prevent global pandemics or flag potential so that medical devices can be updated to find the right solutions and have the resources ready to combat these.