According to a report from Apollo Hospitals, there are only about 10,000 trained radiologists for the current population of India, and AI-based radiology reporting could be a saviour to address the shortage of radiologists.
Synapsica, an AI-based radiology reporting start-up, has been using AI-powered technology that automates several aspects of radiology workflow, improves the quality of reports, and increases transparency between patients and doctors.
Meenakshi Singh, Co-Founder and CEO of Synapsica, saw the effect of the shortage of radiologists in her hometown in UP and decided to use her AI chops to find a faster and more efficient method to reduce the workload of doctors and radiologists, to help patients get their diagnoses sooner.
How is AI used in radiology workflow?
Explaining how AI is used in radiology workflow, Meenakshi said, “Once a patient scan is taken, and the information reaches our secure cloud server, the AI engine automatically creates biomarkers of pathologies that are shared with the reporting radiologist in a viewer that presents the patient’s scan. A layer of NLP on top of AI-generated bio-markers also pre-fills radiology reports with relevant clinical findings that can be edited by the radiologist, saving typing time. The AI engine also generates reporting instructions that automate tasks normally performed by back-office support personnel.”
About the technology stack used by Synapsica at their backend:
- Pytorch, cuDNN and Anaconda frameworks are used for building neural networks that are trained to detect pathologies on radiology scans.
- Apache, NodeJS, HAProxy, Docker and Nginx are the frameworks for hosting and managing servers.
- MongoDB and Redis are used for data storage.
- AWS Suite (EC2, S3, ELB, etc.) is used for integrated cloud deployment of all components, host AI and application servers, data storage and delivery.
Image source: Synapsica
Synapsica uses Pytorch for its AI models since it allows for easy and modular development in a familiar Python-like environment. Code debugging and finetuning is an iterative process in any software development work. Dynamic computational graphs in Pytorch make it relatively quick and easy to re-work pieces of code, reducing the time between iterations. Development is also made quicker by easy data parallelism that enables the usage of multiple GPUs for training neural networks.
An interesting feature in the AI engine
The start-up is also building new features that extend the capabilities of existing AI engines in automatically detecting pathologies in the spine. This will bring us closer to the vision of automating 60% of radiology reporting workloads over the next few years.
About Self-supervised learning
Synapsica’s existing AI algorithms use self-supervised learning methods that bring accuracy improvements in the range of 3-5% for our classifications models. However, their observation so far is that self-supervised learning works better for general classifications tasks rather than specialised medical imaging use cases where pathologies to be detected can be much more nuanced.
Other players developing AI-model
Other players building AI solutions for radiology workflows include Qure.ai, Predible Health and Deeptek.ai. Each of these players covers different aspects of radiology reporting, and Synapsica’s uniqueness lies in building spine focused workflows.
Low back and neck pain have ranked as the No. 1 cause of Years Lived with Disability (YLD) in the Global Burden of Disease survey conducted by WHO since 1990. We are helping make medical care better for patients suffering from spine-related pain.
Synapsica has been approaching diagnostic centres, hospitals and radiology reporting groups; they are planning several initiatives for partnering with radiology practices across the spectrum of clients, including having:
● Inside sales for small to mid-sized practices
● Direct sales to large practices and hospital groups across India, Africa, the Middle East, the UK and the US
● Pilot projects with very large practices to evaluate the operational value-added by AI-assisted reporting.
As the number of cases with pain is growing, so are the radiologists at Synapsica’s platform; the start-up is planning to expand in the US, the UK and Australia. They also plan to launch more AI-driven features that further reduce radiology workloads while creating more detailed and objective reports.
Since its inception in 2019, Synapsica has raised $4.2 million in funding. Its investors include Silicon Valley-based incubator Y Combinator, capital funds IvyCap Ventures, and Endiya Partners.