AI in Medical Imaging

Introduction

AI is everywhere, be it e-commerce or healthcare. In the healthcare industry, medical imaging is the area where AI has seen the highest usage.

Medical imaging results are rich sources of patient data and require high precision. Based on medical imaging results, further care and treatment for a patient are decided. 

The use of AI in imaging is a way to reduce human errors and assist radiologists in reading medical imaging reports with accuracy. 

The advent of AI in Medical imaging is not a recent event. It has evolved over the years. The exponential growth of medical imaging data burdening physicians led to its faster adoption in the medical imaging field.

Medical imaging detects abnormalities by processing images. AI in medical imaging fastens the process and avoids the possibilities of human errors.

Machine learning and deep learning are subsets of Artificial intelligence enabling computer systems to learn from existing imaging reports. The more the imaging results are fed to the systems, the better the system learns.

Applications of AI in medical imaging

AI-powered medical imaging has numerous applications. Application areas are categorized into different sub-branches based on disease diagnostics and medical operation planning. 

The major use cases of AI in medical imaging are:

• Identifying cardiovascular abnormalities
• Detecting fractures and other musculoskeletal injuries
• Aiding in the diagnosis of neurological diseases 
• Screening for cancers like breast cancer, 
• Predicting Alzheimer’s disease
• Revaluating treatment
• Planning surgeries

Companies and startups focused on AI in medical imaging

• Google’s work on AI in medical imaging is noteworthy. Deepmind, a subsidiary of Google, is known as the mother of all AI. With the help of AI medical imaging, Google can now detect
  • Skin diseases 
  • Eye diseases
  • Screen Lung cancer 
  • Screen Breast Cancer 

• IBM Watson Health aims at faster processing of images generated from medical imaging and interpreting the information from the database of numerous processed images. It is a Fortune 100 company and a pioneer of AI-powered healthcare applications in India.

• Arterys is a pioneer in four-dimensional cloud-based imaging. It built the first tech product to visualise and quantify blood flow using MRI. It uses AI to reduce missed detections and provide preventive care.

• Butterfly network aims at detecting diseases in real-time during scanning. Providing both hardware and software solutions, Butterfly network carves a different perspective on medical imaging. It uses ultrasound-on-chip technology.

•  Zebra medical vision is the world’s first AI chest x-ray triage product to receive FDA clearance. Zebra med is an Israel based company that provides AI medical imaging solutions to radiologists at a fixed annual fee.

• Freenome uses AI to detect cancer by imaging blood cells. It has raised over $800 million in 6 years.

• Enlitic uses deep learning to help radiologists read images faster and with more accuracy. The company aims at equipping radiologists with algorithms that ease their work.

Many other companies contributing to developing AI in medical imaging are

• Gauss Surgical Inc. uses AI to detect surgical blood loss.
• Sigtuple is building intelligent AI based screening solutions to aid diagnostics.
• Caption Health provides AI software to interpret ultrasound exams.
• Behold.ai helps radiologists diagnose scans with artificial intelligence technologies.
• Viz.ai uses AI imaging to detect early signs of brain strokes.
• DiA Imaging uses AI-powered solutions for ultrasound image analysis.
• RetinAi is a discovery platform to detect eye diseases with AI imaging.
• Subtle Medical provides AI-powered medical imaging for faster and safer workflow. 
• BrainMiner is a UK based company that provides AI-based automated solutions to analyse brain images.
• Lunit, abbreviated from ‘learning unit’, develops software to process medical imaging data using deep learning technology.
• Aidoc is an Israel based company that provides AI tools for radiologists. It helps in detecting critical abnormalities through deep learning and AI algorithms.
• Kheiron Medical technologies help in the early screening of breast cancer.
• Braid Health provides an AI-based operating system for health data. It is a diagnostic collaboration platform for radiologists and healthcare providers.

You can read more about other works and research on AI in medical imaging here.

Implementation challenges

Though the implementation of AI in medical imaging changes the way diseases are diagnosed, reduces human errors, and decreases the burden of radiologists, it is difficult to implement. 

The major challenges faced while implementing AI-based solutions in medical imaging are

• AI screenings are not 100% accurate. Human interference is necessary.
• Organisation and pre-processing of data are difficult
• Infrastructure and health data sharing problems 
• High cases of false positives

Future of AI in medical imaging

AI has a bright future in medical imaging. Currently, we are at the beginning of the AI era and a lot of research and pilot projects are in progress. Also, the FDA approval process takes years before the technology is actually brought to use.

Google, IBM Watson, and other major players are working day in and day out to develop better AI-powered medical imaging solutions. AI solutions will impact medical imaging more quickly than other fields.

For patients, AI in medical imaging will mean better and quicker diagnostics.

For healthcare providers, it will help in making better decisions.

For radiologists, AI in medical imaging will help in easing the workflow and better diagnosis.