Use Cases

Early detection of Blood Diseases with AI

How Artificial Intelligence improves early detection of blood diseases

Every day, millions of single blood cells are evaluated for disease diagnostics in medical laboratories and clinics. Mostly this repetitive task is done manually, by trained cytologists who inspect cells from microscopic blood images, one after another.

How can AI help?

Blood laboratory tests are used to confirm, exclude, classify, or monitor diseases and ultimately guide treatments, by analyzing and counting the different types of blood cells in the patient’s organism.
Traditionally, blood cells are counted manually using a hemocytometer along with other laboratory equipment and chemical compounds, which is a time-consuming and tedious process.

Machine Learning (ML) - a subset of Artificial Intelligence - can help to automatically identify and count the different types of blood cells using computer vision and classification algorithms. Moreover, the machine learning algorithm can “learn” to recognize which cells are healthy and which are not, opening the door for early detection of a particular blood disease.

For example, if a patient is suffering from leukemia, it may be hard to identify the disease at an early stage of development only from microscopic blood images. As a result, most patients get diagnosed when the disease has advanced enough, leading to expensive and extremely painful treatment (especially for children).

AI can help improve early detection of blood irregularities, and hence allow medical practitioners to help their patients on time, providing the appropriate treatment while the disease is still at an early stage.

Kantify’s work in early detection of Leukemia

Recently, Kantify has developed an AI model that can detect leukemia at a very early stage. By analyzing microscopic blood images of patients, our AI model can count red and white blood cells, detect if there are any burst blood cells, and if yes, determine the type of the abnormality. The way it works is by taking an image and feeding it through an artificial neural network.

The challenges in early detection of leukemia

Leukemia counts as the most common cancer in children and young adults, and is largely undetected in its early stage. Mostly because medical professionals are usually unable to detect it until more noticeable symptoms start emerging, and the disease is already in a pretty advanced stage. This is partly because tests for leukemia are pretty expensive, painful (especially for young children), and hence, rarely done as a preventive measure.

In addition, analyzing the results of a patients’ complete blood count (CBC) can be expensive for medical organizations, as well as time-consuming, and may suffer from classification variability.

What are the benefits of using AI for early detection of leukemia?

Patient-friendly tests

Leukemia tests are frankly painful and disturbing, especially for children. The testing usually includes two steps: a bone marrow aspiration, and a bone marrow biopsy. Hence, why medical practitioners usually direct their patients for these tests when they are almost certain of the presence of cancerous cells in the patient organism - which is often when the disease is more easily noticeable from the blood results and it has reached a more advanced stage.

Introducing prevention tests

Because the testing for leukemia is so painful and disturbing, hardly ever are they made preventively. If you have a history of this disease in your family, you might be interested in getting tested pervasively or test your child. AI models can help avoid the already painful process of testing, and analyze an unlimited amount of patient’s blood images more fast, precise, and cheaper.

Timely detection of blood cell irregularities

The diagnosis of cancerous blood cells can be a complex, multi-step process. Consequently, some patients experience delayed diagnosis, often associated with missed opportunities in detection of early “signals” in the blood, pointing out to the disease. Sometimes, delayed detection can lead to no major harm in the patient’s health, but sometimes it is a missed opportunity to prevent the progression of the disease. An AI-based blood cell analysis can help in early detection of (sometimes even minor) blood cell irregularities, ultimately leading to timely detection and diagnosis of potential pathologies before they advance. It can ultimately help improve healthcare services for both practitioners and patients.

Let’s get in touch!

AI is transforming the future of life sciences, and early detection of diseases is one of the most promising areas of development. By powering a new generation of systems that equip clinicians with smart tools when delivering care, AI will lead the way in a new era of exciting breakthroughs in patient care. Let’s get in touch to discover how you can use AI in blood cell analysis.