The combination of quantum and classic computing supports the initial diagnosis of breast cancer

Quantum computing is still in the early stages of development, but researchers have explored its potential use on a large scale. A recent study conducted at Sao Paulo State University (UNESP) in Brazil suggested a hybrid quantum classical model to support the diagnosis of breast cancer from medical images.

Work was published as part of the work 2025 IEEE 38th International Symposium for Computer Based Medical System (CBMS)Organized by the Institute of Electrical and Electronics Engineers (IEEE). In the publication, the authors describe a hybrid neural network that combines quantum and classic layers, known as the Covoloshol Nural Network (QNN) using a point of view. They applied the model to mammography and ultrasound images so that the lesions could be classified as benign or deadly.

“The thing we wanted to bring to this work was a very basic architecture that used quantum computing, but it contained at least quantum and classical devices,” says Yasmin Rodrigs, the first author of the study. This is part of its scientific launch project, which is overseen by the UNESP’s Borrow Campus, a full professor of computing department, Geo Paulo Papa. Papa also authored the article.

He chose breast cancer as a computer testing target as it is the most common type of women worldwide. In 2022, 2.3 million cases and 670,000 deaths were recorded. In order to increase the chances of treatment and survival, initial detection is very important. However, traditional methods, such as mammography, relies heavily on human interpretation, which can lead to variations in diagnosis.

Papa says, “Although it is easy to follow theoretically, mammography is still a test that depends heavily on performing professional processes.”

What distinguishes the UNESP’s work from other artificial intelligence measures in health is that it is the use of a quantum convention layer along with the classic layer.

Papa says, “Like the classic convention, the goal of quantum conductivity is to remove local features from pictures such as structured data. But it does more effectively by taking advantage of the quantum system’s features, such as superposes and confusion, which is more effective and more efficient.” In the study, the quantum layer, consists of four causes (quantum bits), changed the traditional process of removal from the images.

“What we did, was that the four cobbut quantum circuit had to move the images through the doors and logic doors. It helped us to get the necessary measurements. Then, the images went into simple classic layers, which finalized.”

This study did not use a real quantum processor, but rather a classic platform such as the Penny Lane Framework, which reproduces the ideal conduct of quantum circuit without environmental noise.

There are some real quantum computers in the world. They are all in the experimental phase and there are a limited number of quetts, of which are more than a few dozen to more than a thousand. They must be cooled near the absolute zero (-273 ° C), in most cases, in most cases. Therefore, when consumers are available, their use is inappropriately expensive.

Rodrigus explains, “Simulators we work fully on classical platforms, do not use real quaties, but assess how the circuits will behave in the ideal quantum world. They are mistaken, free from environmental changes that affect existing quantum computers.”

According to the researcher, although very easy, the artificial quantum circuit has already shown promising results.

“The best performing classical network had 11 million parameters. We, with the quantum layer, had about 5,000 5,000. This changes everything.”

One of the basic concept of physics behind the model is super position. The superpoints differentiate the cobbat from the classic butt.

Rodrigus explains, “It helps to refer to the representation known as the Baloch sphere, to understand the superposition. We can imagine this circle as a football ball, where each point at the level represents a potential quantum state between the state of 1, between 1 and the state of 1, between 1 and 1.

“When this system is on the North or South Pole, we are 100 % sure that the cobbut is in 0 or 1 state respectively. But at any other place in the circle, we enter the area of ​​probability: it is in a super position with different possibilities of measuring the cobbat 0 or 1.”

There are many potential superpoints states. For example, a cobbut can be in a state where it is likely to be 70 % of 0 and 30 % likely to be 1, or any other combination. Due to these superposing states, Cobs can process more information than classical bits. This is one of the reasons that quantum computing has the potential to cross classic computing. It has the maximum capacity and speed of processing. However, many technical challenges have to be overcome before understanding this ability on a large scale. . =

Papa said, “It is expected that people will have quantum computers, as they have traditional computers today.”

In the study, the information that encoded in cobs was pixels of mammography and ultrasound images. Sometimes it was just a pixel, and sometimes it was more than one. The model was tested with two databases: breast (with ultrasound images) and BCDR (with segmed memoogram). Even despite just four quartes circuits, the hybrid network performed competitively. In the best case, it achieved 86.1 % in the test set of 87.2 % accuracy and authentication set.

Rodrigus commented, “The idea was to create an architecture that can be used and further developed in other studies.”

Other requests

Although this research focuses on breast cancer, the authors said that developed architecture can be applied in other areas as well. For example, it can be used to analyze brain lesions or to classify tissues in microscopy images.

The researcher concluded that “we are taking the first step towards a new computing model for medical diagnosis. This is a passionate field in the coming years.