March 13, 2025

Breathing is something we do constantly and unconsciously, without ever taking a break. In recent years, research has been published on the relationship between breathing and physical discomfort such as low back pain. NEC has developed technology that outputs highly accurate waveform graphs showing the movements of the chest and abdomen during breathing. We spoke with the researcher and business manager about the purpose and expected value of this technology.

Focusing on the breathing cycle, as well as subtle breathing movement and quality

―What kind of technology is respiratory visualization technology?

Yasukawa: This technology analyzes the movements of the chest and abdomen with millimeter precision and outputs a respiratory waveform graph. By using a type of distance sensor called LiDAR (Light Detection and Ranging), we capture the 3D shape of the torso and estimate the positions of the sternum and navel in real time from the images. This allows us to visualize the physical movement that occurs during breathing. In the waveform graph that is output, the vertical axis represents the magnitude of the movement and the horizontal axis represents time. From this, we can observe the changes and pace of the chest and abdomen movement during breathing. We can understand a variety of information in line with the interests and perspectives of respiratory trainers and therapists, such as the degree to which the chest and abdomen movements are synchronized, and how long the exhalation phase lasts.

Okita: Although we are using LiDAR, there is no need for bulky equipment. The technology we developed incorporates LiDAR into a tablet device, making it easy to utilize in a non-contact manner. We will soon release a dedicated application that can be installed on a tablet, as a service for professionals who provide training and other services that focus on breathing movement.

The intended users are people working in places such as osteopathic or chiropractic clinics and personal gyms. We believe this technology can be incorporated into new service offerings, such as showing clients a screen of their respiratory waveforms while giving explanations, or showing the differences before and after providing treatment. In addition, voice trainers and yoga instructors have shown interest during our demonstrations, so we are considering these potential applications as well.

―So, why visualize breathing movement in the first place?

Yasukawa: My interest was originally sparked by a suggestion from a professor at Institute of Science Tokyo (formerly Tokyo Medical and Dental University), who we were working with on a joint research project. He mentioned that the relationship between breathing and orthopedic conditions had been gaining attention recently. Indeed, since around 2010, research papers focusing on the relationship between the respiratory system and motor dysfunction have increased, and research findings demonstrating that low back pain can be improved by breath training have been reported. In addition, there has been a recent increase in the number of applications and products focused on breathing, including wearable devices. However, these are mostly focused on measuring the breathing cycle, and there are few examples of technologies aimed at addressing the quality of breathing itself. Therefore, we started researching ways to visualize breathing.

Estimating the real-time positions of the sternum and navel with high accuracy

―What kinds of technologies are used?

Yasukawa: LiDAR sensors only measure the distance to an object, so at this point, we can only detect 3D shapes. The system itself still does not understand where the chest or abdomen is. First, we used human keypoints detection technology to roughly estimate the area. Of course, this alone does not provide sufficient performance, so we had to make adjustments to further improve the accuracy. To do so, we collaborated with Institute of Science Tokyo to measure breathing movement in healthy individuals by simultaneously using motion capture technology and distance sensors. By comparing the distance sensor data to the motion data captured with reflective markers on the body, we investigated which points in the distance images would provide the highest accuracy. As a result, we were able to estimate with high accuracy the position of the center of the sternum and the point directly above the navel, which enabled us to produce a suitable respiratory waveform. We then developed a prototype using large 3D cameras and GPU computers, which eventually led to the current system.

Okita: Originally, this system required large cameras and GPUs, but with the service we are about to release, we have managed to condense it into a tablet device. As a result, we have achieved high precision in analyzing movements down to the millimeter level, while also providing convenience. Furthermore, for commercialization, we received support from prominent professionals and their communities in the field of breathing, as well as many other specialty doctors, osteopathic clinic managers, and physical therapists. We have incorporated the feedback from people working in the field into the application's features and interface.


Yasukawa: Actually, before becoming a researcher, I worked as a physical therapist for about 10 years, so I am also drawing from my clinical experience.

In addition to commercializing and applying the technology, we have also presented it at academic conferences. In 2022, we presented at the prestigious Society of Instrument and Control Engineers (SICE) conference, and in 2023, we presented at the Annual Research Meeting of the Japanese Orthopaedic Association, which has a long history.

Advancing toward "correct breathing" through data accumulation

―Please tell us about the future prospects and goals for this technology.

Okita: First, I want to make sure we have a successful service launch, which is just around the corner. As for future developments, I want to closely monitor the user response. As of now, the value we envision is that practitioners at osteopathic or chiropractic clinics and personal gyms can provide clear explanations and advice to clients while viewing data together on the same screen. I want to keep tracking the value of the technology based on its actual usage, to understand how it can affect the performance of practitioners and how it can contribute to the physical health of end users.

Honestly, I believe there are still some unknowns with this service. This is an unusual endeavor, even in global terms, so it is uncertain how society will accept it. However, it also means there is plenty of room for growth. I believe there are many potential uses and needs in industries we have not anticipated yet, so I want to keep my ears open to the possibilities and continue expanding the business.


Yasukawa: Right. Quantifying and visualizing breathing is an area that has yet to be explored much. We are presently receiving inquiries from university professors who are interested in using the technology for research purposes, so I hope we can contribute to the academic field as well.

Also, as our tool becomes more widely adopted, the accumulated data could lead to new insights. Personally, I am interested in the theme of clearly defining what is "correct breathing." If we can identify breathing patterns that contribute to mental and physical health, it may also help address issues like low back and shoulder discomfort for end users. Such an achievement would be a great reward as a researcher, and it would bring me a step closer to being the kind of researcher I have always aspired to be.

Furthermore, it would be great if, in addition to the respiratory waveform output, we could also provide suggestions based on what these waveforms represent in terms of health status. I hope we can explore such developments in the future by incorporating NEC's expertise in big data analysis and assets such as large language models (LLM).

Go back to Featured Technologies