In this article, you will learn about a real-world example of the use of artificial intelligence in medical imaging. Read on to learn the details of how various deep learning models are combined to analyze images taken with a microscope.
Artificial Intelligence in Medical Imaging
You may have read use cases where AI is used in medical diagnosis to differentiate between images showing pathological and non-pathological features (e.g. differentiating benign or malignant skin lesions), and also to detect areas of interest in images (e.g. the presence of tumours).
At Capillary.io we have focused on another medical test that is somewhat less well known but widely used in rheumatology and internal medicine: nailfold capillaroscopy.
What is a Capillaroscopy?
Capillaroscopy consists of observing the blood capillaries at the base of the patient’s nails (nail bed) using a microscope called a capillaroscope and helps to determine the state of the patient’s vascular system in a simple, fast and non-invasive way.
Capillaroscopy is frequently used for the diagnosis and follow-up of some autoimmune diseases such as scleroderma, dermatomyositis or mixed connective tissue disease.
Recognition of Blood Capillaries in Microscope Photos
So how can AI help improve the practice of capillaroscopy? And what types of deep learning models are used to do so?
In summary, during a capillaroscopy, the physician will observe the capillaries with the microscope and take pictures of them to later count them, decide if each one of them presents a normal shape or not and calculate some statistical metrics such as capillary density, percentage of dilations, presence or not of giant capillaries and hemorrhages, average capillary size, etc.
With AI we can improve the overall process for physicians, who no longer have to analyze each image and calculate the metrics manually as the system calculates them instantly when all the necessary data is available. In addition, the doctor will be able to correct mistaken model predictions, as no perfect AI system can work with 100% precision and recall, and the quality of the images may vary depending on the doctor’s own skill and the microscope used.
The task of locating capillaries and the task of classifying them into different types can be unified into a single task. In the field of machine learning for computer vision, most object detection models have been designed to do both at the same time in a way that is transparent to the developer. Such models, such as the well-known Retinanet and YOLO, output both the rectangle that frames each object in the image (bounding box) and the type of object (label class).
In conclusion, deep learning models can be of great use to physicians to improve the quantity and quality of the data they obtain from their patients with imaging for diagnostics. AI is therefore a support tool that speeds up the diagnosis and monitoring of diseases. By properly integrating AI into medical software we can help the doctor make better decisions.
- 📩 Want the latest on technological innovations sent to your inbox? Subscribe to our newsletter!
- 🎙 The Detechtor Podcast is available on all podcast players!
- 📲 Let’s connect! Follow us on Twitter | Instagram | Facebook | Youtube