Biometric Authentication Gets a New Champion: Innovative ECG-Based Identification Technology by Pratek

In the field of biometric technology, in addition to the well-known fingerprint recognition systems, some novel and uncommon identification methods have emerged in recent years, attracting significant public attention and sparking intense curiosity. For instance, facial recognition technology and iris recognition technology are both considered cutting-edge, high-precision frontier technologies.

According to statistical data from the Qianzhan Industry Research Institute, the market size for biometric identification was RMB 42.3 billion in 2010, RMB 51.3 billion in 2011, RMB 59.6 billion in 2012, and RMB 91.0 billion in 2015, with projections indicating it will reach RMB 175 billion by 2020. In fact, it is not only the general public taking notice; internet giants with exceptional market foresight, such as BAT (Baidu, Alibaba, and Tencent), have long recognized the potential. They are increasingly investing in the biometric identification industry, aiming to build a complete industrial chain and establish large-scale commercial operations.

The most typical example is when Jack Ma personally demonstrated Alipay’s facial recognition technology at the CeBIT computer fair in Hanover, Germany, completing a “face-scan” payment. This innovation was soon applied to the account verification process on Taobao, where applicants must first pass “facial comparison” and “liveness detection” to be eligible for approval.

However, as an emerging technology, facial recognition is far from perfect and faces core challenges that urgently need to be addressed. For instance, influenced by multiple factors, current facial recognition rates remain low, falling well short of practical usability. Additionally, facial appearance changes with age, rendering facial recognition systems unable to adapt over time. The newer iris recognition technology also has significant drawbacks, such as the inability to miniaturize image acquisition devices, high equipment costs, and other barriers that hinder large-scale adoption. Looking back at traditional fingerprint recognition, it suffers from prominent issues including poor security and the tendency to leave behind traceable residues.

Therefore, although biometric technology is widely regarded as having a bright future, the widespread adoption of new technologies still faces significant challenges due to numerous unresolved contradictions. Now, Pratek, an entrepreneurial team from Guangdong, has taken a novel approach by developing a unique electrocardiogram (ECG) signal recognition technology. This innovation promises to enable payments in a more cost-effective and theft-resistant manner, effectively addressing the challenges of secure payment.

Team Pratek believes that ECG-based recognition technology not only meets the stability, uniqueness, and convenience required for fingerprint recognition but also offers unique anti-spoofing capabilities.

Biological studies indicate that ECG signals are bioelectric currents generated by cardiac contractions, with their waveform morphology determined by the anatomical characteristics of the human body and the heart. Although ECG signals can be influenced by factors such as physical condition and mental state, these factors only cause scaling or deformation of the waveform without altering its underlying structure. Consequently, each individual’s ECG remains unique. Leveraging these characteristics, Pratek has proposed a secure and convenient ECG-based authentication payment solution that effectively addresses security challenges in mobile internet and Internet of Things (IoT) applications.

ECG recognition consists of two phases: registration and verification. During the registration phase, a segment of ECG signal is first acquired, then segmented and aligned by cardiac cycles, followed by quality screening for each cycle. Feature extraction is performed on the N highest-quality ECG cycles selected through screening to generate a feature template, which is stored in the database to complete the registration process. In the verification phase, features are first extracted from the ECG signal of the user to be verified, and then compared with the feature templates in the database to produce an identification result.

ECG recognition terminals can be deployed according to different usage scenarios. Mature ECG sensors are already available on the market; they are compact and low-cost, thus enabling easy integration into any terminal device.

Pratek’s solution utilizes the currently popular wearable wristband as the terminal device, integrating ECG sensors into the band to serve as the system’s hardware entry point. Upon first use, users must authenticate their identity via the wristband’s companion app, after which the wristband captures the user’s ECG data to complete the registration process within the app. An encrypted channel is established between the wristband and the smartphone using asymmetric key algorithms to ensure the secure transmission of ECG data. Furthermore, the wristband incorporates a digital signature and verification module, providing data signing, signature verification, and encryption services, thereby meeting financial institutions’ requirements for non-repudiation and confidentiality in data transmission.

The founder of Pratek stated that this innovative technology leverages advanced techniques such as signal processing, pattern recognition, and machine learning—the most prominent technology in the era of big data. The Gypal 2.0 version has already been developed, achieving an accuracy rate of approximately 93%. The Gypal 3.0 version is currently under development and is expected to achieve an accuracy rate exceeding 97%, while the algorithms continue to require ongoing updates.

The Pratek team was established in October 2014. It comprises five members with an average age of 30. The team’s most prominent advantage lies in its cutting-edge technical expertise in computer science, particularly in deep learning. Below are the backgrounds of the team members:

Chen Zhixiao (Founder), Ph.D. from Beijing University of Posts and Telecommunications, majoring in Communication and Information Systems;
Zhang Han (Co-founder), Master’s degree from Carnegie Mellon University, USA, majoring in Natural Language Processing at the School of Computer Science;
Li Zhihuan (Co-founder, CTO), Master’s degree from Sun Yat-sen University, majoring in Pattern Recognition and Intelligent Systems;
Zou Jian (Co-founder, CEO), Bachelor’s degree from the University of Leicester, UK, majoring in Mechanical Engineering;
Kong Lingpeng (Technical Co-Founder), a Ph.D. candidate at Carnegie Mellon University in the United States, specializing in Natural Language Processing within the School of Computer Science.

Notably, the project made its debut by securing first place in the team competition of the 4th China Innovation and Entrepreneurship Competition (Guangdong Dongguan Division) in October 2015. In the same month, it also achieved an outstanding result by placing seventh in the team competition of the 4th China Innovation and Entrepreneurship Competition Finals (held in Tongxiang, Zhejiang).



Pratek’s startup mentor is Dr. Wang Yong from Shenzhen Sinoway Capital. He considers the Pratek team to be one of the most technically distinguished teams in the Dongguan region, boasting strong technical capabilities. It is also the first technology project in China to focus on identity recognition based on electrocardiogram (ECG) signals. All founders are seasoned professionals who have been actively working on the front lines for many years. Furthermore, as the project represents technological innovation rather than business model innovation centered on marketing and operations, it faces limited competition in the domestic market, thereby ensuring a high success rate.

Dr. Wang Yong believes that, with time, this ECG signal-based identity recognition technology is expected to rapidly expand into a wide range of application scenarios, such as access control systems and attendance tracking systems, ultimately reaching the field of financial payments.

Regarding its commercial strategy, Pratek’s founder stated that the company will initially provide algorithms and software systems, partnering with well-established hardware manufacturers (such as wearable band makers, smartphone manufacturers, and mobile terminal providers). In later stages, it plans to launch its own branded products to comprehensively penetrate the financial payment sector. The company projects capturing a 10% share of the biometric recognition market by 2020, equivalent to RMB 17.5 billion. Currently, Dongguan Rural Commercial Bank has expressed strong interest in the product and is poised to become one of the first early-adopter clients upon its official launch.

Therefore, Pratek’s long-term development plan is to enter low-barrier, high-end access control and attendance sectors in the short to medium term with ECG signal-based identity recognition products, while considering expansion into the smart healthcare and financial sectors in the medium to long term.

Pratek has secured Pre-A financing and plans to raise angel funding in the second half of the year; it is currently in the incubation phase. The team is seeking 4–5 junior software engineers and looks forward to collaborating with more medical institutions. We hope that hospitals can provide extensive electronic sample data, including ECGs (both normal and pathological samples).

If you are interested in this project or have a startup project that you would like to be covered, please contact the VCBeat team by adding WeChat ID: zhn88369