BLW Filtering for ECG Systems: Getting IEC/FDA compliance right

,

Baseline wander (BLW) is one of the most persistent — and misunderstood — problems in ECG systems. Caused by respiration, movement, and electrode drift, BLW introduces low-frequency drift that can distort key waveform features and disrupt clinical analytics.

While BLW suppression is essential and mandated in all ECG systems, developers often face two key challenges:

  • Interpreting the filtering requirements in standards such as IEC 60601-2-25 and 60601-2-47, which define passband specifications but fall short on practical implementation guidance
  • Designing real-time, morphology-preserving filters that are efficient enough to run on embedded platforms such as modern low power microcontrollers used in modern wearable and portable devices.

In both the United States and Europe, regulatory bodies reference the IEC 60601-2-25 and 60601-2-47 standards when evaluating ECG systems. In the United States, the FDA uses them as a key part of its review process, while in Europe they form the basis for CE marking under the MDR (Medical Device Regulation, EU 2017/745). As such, adhering to IEC 60601-2-25 and 60601-2-47 is essential for global regulatory compliance.

Despite the growing use of AI and ML in ECG signal processing, current IEC and FDA standards have not yet caught up. There is no formal guidance on how to validate AI-based filtering or classification pipelines, leaving developers without a clear compliance path when using these emerging techniques.

As a result, most developers still rely on traditional digital filtering methods — especially when targeting regulatory approval. This is particularly true in wearable and portable devices, where Arm Cortex-M microcontrollers, such as the STM32 family, are the dominant hardware platform. Yet even here, many continue to follow legacy filtering advice inherited from the analog era, without re-evaluating what’s truly required for today’s real-time digital systems.

In this article, we break down what the standards actually require, explaining where the widely-cited 0.05 Hz cutoff comes from, and show how to design real-time high-pass filters that balance signal quality, compliance, and power efficiency — all within the constraints of modern embedded platforms such as the STM32 microcontroller.

What the standards don’t say, and why this is confusing

While IEC 60601-2-25 and IEC 60601-2-47 do define passband requirements (e.g., 0.05–150 Hz or 0.67–40 Hz), but they don’t offer practical engineering guidance for how to meet those requirements on resource-constrained platforms like wearables.

They don’t cover things such as:

  • How to implement the filter in a real-time embedded system
  • Trade-offs between FIR vs IIR in morphology-sensitive signals
  • What to do about SoCs with built-in analog filtering
  • Energy/performance implications of filter design choices
  • How to practically validate the signal chain for compliance — e.g. when conducting the ‘sinewave test’, should the amplitude of the 0.67-40Hz test sinewaves  be measured continuously or averaged over time? Should I use RMS or peak amplitude when comparing the amplitudes with the 5Hz reference amplitude?

This leaves engineers with many unanswered questions:

  • Is a 0.05 Hz cutoff always necessary?
  • What level of phase distortion is acceptable?
  • Can I use an IIR filter and still be compliant?
  • How do I qualify an SoC with undocumented internal filters?
  • What measurement methodology should I use for compliance testing?

As a result, developers often either overdesign their filters based on worst-case diagnostic assumptions — or under-design, thinking any high-pass filter will suffice. Neither approach guarantees compliance or robust signal quality.

Baseline wander (BLW) is a well-known issue in ECG systems, arising from slow movements such as respiration, torso motion and electrode impedance shifts. As a consequence, BLW introduces a low-frequency drift, distorting the ECG baseline and compromising the accuracy of QRS detection, HRV analysis, and waveform interpretation.

While baseline wander (BLW) suppression is essential — and mandated in all ECG systems, developers often face two key challenges:

  • Interpreting the filtering requirements in standards such as IEC 60601-2-25 and 60601-2-47, which define passband specifications but fall short on practical implementation guidance
  • Designing real-time, morphology-preserving filters that are efficient enough to run on embedded platforms such as modern low power microcontrollers

These issues often lead to overengineered solutions, misinterpretation of compliance requirements, or filters that distort clinically relevant features — all of which can compromise both product performance and regulatory approval.

0.05 Hz: A Diagnostic-Grade requirement from IEC standards

The 0.05 Hz lower cutoff frequency originates from IEC 60601-2-25, which governs diagnostic electrocardiographs. It is also referenced in IEC 60601-2-47 for ambulatory ECG systems when high-fidelity diagnostic performance is required.

The mandated system bandwidth of 0.05-150 Hz is intended to preserve:

  • ST-segment deviations
  • T-wave alternans and morphology
  • Other slow-changing diagnostic features

This specification is well-suited to clinical and hospital environments but presents serious challenges for embedded systems.

Why 0.05 Hz is difficult to implement in realtime

To meet this requirement, developers usually consider two filtering options:

  • IIR filters: These are computationally efficient but exhibit non-linear phase response, causing phase distortion that alters the shape of the QRS complex. This makes IIR filters unsuitable for ECG applications where signal morphology is critical.
  • FIR filters with linear phase: These preserve morphology but require very long filter lengths—often extending to 5-10k filter coefficients for a 0.05 Hz cutoff—resulting in increased memory use, latency, and computational cost.

Choosing the best cut-off frequency

Wearable ECG systems operate under constraints in power and processing. These systems typically target a bandwidth of 0.5-40 Hz. This is sufficient for preserving the key features of the ECG waveform, while also easing implementation on embedded hardware. This range aligns with reduced fidelity requirements in IEC 60601-2-47 for wearable or monitoring devices.

Where does BLW actually occur?

In practice, the majority of BLW energy is concentrated below 0.5 Hz. This is supported by:

  • Real-world ECG datasets
  • Clinical literature on motion and respiration artefacts
  • Empirical testing using wearable sensors under normal conditions

Sources of BLW such as respiration (~0.2–0.3 Hz), slow body movement, and electrode drift typically lie in the <0.5 Hz region. This means: a high-pass filter with a cutoff of 0.5 Hz or 0.67 Hz will suppress most baseline drift without distorting the QRS complex or T-wave morphology.

This insight is essential for wearable and mobile ECG systems, where developers must balance:

  • Signal fidelity
  • Computational efficiency
  • Power consumption
  • Compliance with IEC 60601-2-47 (where applicable)

Over-filtering (e.g., using a 1 Hz high-pass cutoff) can begin to distort critical ECG features, particularly the ST segment and T-wave. On the other hand, under-filtering may leave residual baseline drift that impacts analytics such as heart rate variability and R-peak detection.

The following animation demonstrates how the QRS is warped when using a 1st order IIR filter.

The 0.5–0.67 Hz cutoff range is often the optimal balance in wearable ECG systems — effective in suppressing baseline wander while preserving essential waveform morphology.

In fact, IEC 60601-2-47 stipulates that a 0.67Hz cut-off may be used if no phase distortion is introduced. We can therefore conclude that a signal bandwidth of 0.67-40Hz is IEC compliant for ambulant systems.

When implemented with a linear-phase FIR filter, this approach meets both signal quality and regulatory requirements — and remains computationally feasible for embedded targets such as the STM32.

Implementation on STM32 microcontrollers

STM32 microcontrollers built on the Arm Cortex-M architecture (e.g. M4, M7 cores) are a popular choice for wearable ECG systems. They offer a solid combination of processing performance, on-chip memory, and energy efficiency — all critical factors for medical-grade real-time signal processing.

When implementing high-pass filters for baseline wander suppression, several hardware and algorithmic factors must be considered:

  • Hardware Floating-Point Support expedites RAD: devices like the STM32F4, F7, and H7 feature single-precision floating-point units (FPUs). This allows developers to expedite RAD (rapid application development) by prototyping and validating filters quickly without needing to deal with fixed-point scaling and rounding errors.
  • FIR Filters preserve morphology, but are computationally heavy: Linear-phase FIR filters are strongly recommended for ECG applications because they preserve temporal relationships of the waveform, which is essential for correct analysis.
    However,
    • A 0.5 Hz cutoff at 200 Hz sampling typically requires 100s of filter coefficients, which is manageable on most STM32 devices (Cortex-M4/M7), and just about manageable on ultralow power Cortex-M0+ devices.
    • In contrast, a 0.05 Hz cutoff typically requires 1000s of filter coefficients (10k+ in some cases), resulting in high memory usage and processing load — often beyond what’s practical for real-time wearable designs.
  • Affect on battery life: Long FIR filters keep the FPU active nearly continuously, which prevents the microcontroller from entering low-power states. This significantly impacts battery life — especially in wearables running at higher sample rates, e.g. 500 Hz. For this reason, the cutoff frequency must be carefully selected to balance signal quality with energy efficiency.
  • Avoid IIR Filters for morphology-critical Applications: While IIR filters are computationally attractive due to their low order, they introduce non-linear phase distortion. This warps the QRS complex, alters timing relationships between ECG segments, and undermines compliance with medical waveform standards. For any application where waveform shape matters, IIR filters should be avoided.

Ultimately, for wearable ECG systems targeting compliance and signal integrity, linear-phase FIR filters on STM32 microcontrollers provide the most practical and reliable foundation for real-time baseline wander removal.

As a closing remark, it’s certainly true that many developers — especially in research or offline processing environments — use zero-phase IIR filtering by applying the filter forward and backward through the signal. While this technique avoids phase distortion and preserves waveform morphology, it is inherently non-causal and unsuitable for real-time applications. This distinction is critical, as filters that rely on forward-backward (zero-phase) processing cannot be implemented in real-time embedded systems, making them unsuitable for wearable applications that operate on streaming data.

Biomedical SoCs: Powerful, but require full signal-chain qualification

Several IC vendors — including Analog Devices (ADI), Texas Instruments (TI), and STMicroelectronics — have developed highly integrated biomedical SoCs that feature built-in analog front ends (AFEs). These devices often include:

  • Programmable gain amplifiers
  • Instrumentation amplifiers
  • Integrated analog filtering
  • High-resolution ADCs with CIC (Cascaded Integrator-Comb) filters

While these SoCs offer impressive integration and are well-suited for compact, low-power designs, they also introduce a significant challenge, i.e. Developers must still qualify the entire signal chain — including analog and digital filter stages — to ensure IEC compliance.

Typical challenges include:

  • Measuring the overall frequency response of the signal path
  • Evaluating the effects of CIC decimation filters and analog high-pass/low-pass stages
  • Understanding the behaviour of any hidden gain or filtering elements inside the AFE or ADC

Because much of this behaviour is not fully documented, it often requires empirical validation using signal injection and sweep testing — a tedious and time-consuming process that many teams underestimate.

Without this validation, there’s a real risk of:

  • Non-compliant frequency response
  • Undetected distortion of QRS or ST segments
  • Unexpected interaction between analog filtering and digital signal processing

Bottom line: integrated SoCs are powerful, but compliance is never automatic. It’s up to the developer to fully characterize and correct the signal path to meet the requirements of IEC 60601-2-47 — especially when targeting diagnostic or screening-grade wearable ECG systems.

Recommendations based on legacy systems

The commonly referenced 0.05 Hz high-pass cutoff originates from the behaviour of first-order analog filters used in ECG systems developed in the 1970s and 1980s. These filters had slow roll-off and introduced significant phase distortion, but were accepted at the time due to the limitations of analog circuitry and minimal awareness of morphology distortion effects.

Over time, this cutoff value was incorporated into standards such as IEC 60601-2-25, and later IEC 60601-2-47, often without re-evaluation in the context of modern digital systems. As a result, much of today’s ECG design guidance continues to reflect analog-era limitations, despite dramatic advances in hardware and signal processing.

It’s also worth noting that much of the legacy design advice found in textbooks and reference designs stems from the ‘analog era’ — when discrete op-amps, analog filters, and limited processing power dictated system architecture. While those principles served their purpose, modern biomedical SoCs and real-time digital filtering pipelines operate under entirely different constraints and possibilities.

As a result, engineers must challenge inherited assumptions, revalidate their signal path in the digital domain, and adopt design practices suited to real-time embedded systems — not outdated analog models.

This shift in mindset is essential not only for meeting IEC compliance in modern systems, but for achieving robust, efficient, and clinically reliable signal quality — especially in wearable and low-power applications.

Designing for Clarity, Compliance and Real-time constraints

Tools such as the ASN Filter Designer simplify ECG filter development by providing:

  • Reference designs for designing linear phase IEC compliant ECG filter cascades (0.5-40Hz)
  • High accuracy frequency response charts
  • The ability to load and stream ECG dataset to visualize filter performance in real-time
  • Code export options for Arm processors (ANSI C) and Python and Matlab

This allows engineers to deploy real-time, medically robust filters quickly and with confidence.

Key Takeaways

A 0.05 Hz high-pass cutoff is referenced in both IEC 60601-2-25 (diagnostic ECG) and IEC 60601-2-47 (ambulatory ECG) for systems intended to support full diagnostic fidelity. However, this requirement is not appropriate for most wearable applications.

In wearable environments, baseline wander is dominated by motion and respiration artefacts, most of which lie below 0.5 Hz. As such, filtering at 0.05 Hz does not sufficiently suppress BLW, and imposes significant implementation burdens — including high memory usage, increased latency, high computation requirements and greater power consumption.

Crucially, IEC 60601-2-47 permits a high-pass cutoff of 0.67 Hz for ambulatory systems, provided that no phase distortion is introduced. When implemented using a linear-phase FIR filtering, a bandwidth of 0.67–40 Hz is both IEC-compliant and technically practical for real-time implementation on embedded platforms such as the STM32.

AI-based ECG processing

While the current standards, including IEC 60601-2-47, provide guidance for traditional digital filtering approaches, it’s important to note that they have not yet caught up with AI-based ECG processing techniques. As of today, there are no formal standards, validation protocols, or compliance pathways defined for ECG systems using AI/ML for baseline correction or morphology analysis.

This lack of regulatory clarity means that, for now, classical digital filtering — particularly linear-phase FIR filters — remains the most robust, transparent, and standards-aligned approach for ensuring ECG signal quality and achieving IEC/FDA compliance in wearable and ambulatory systems.

ASN Solutions and Expertise

At ASN, we’ve supported numerous international clients in designing medically robust ECG systems — including helping them achieve FDA and IEC 60601-2-47 compliance. From real-time FIR filter design to signal chain validation and compliance testing workflows, we offer practical, implementation-ready solutions tailored for modern embedded systems and wearable applications.

Author

  • Sanjeev is a RTEI (Real-Time Edge Intelligence) visionary and expert in signals and systems with a track record of successfully developing over 25 commercial products. He is a Distinguished Arm Ambassador and advises top international blue chip companies on their AIoT/RTEI solutions and strategies for I4.0, telemedicine, smart healthcare, smart grids and smart buildings.

    View all posts

/0 Comments/by

Choosing the right Arm Cortex-M processor for your biomedical smart product: a practical guide

, ,

Advances in telemedicine healthcare products over the past decades have been truly miraculous with ingenious little devices invented by start-ups as well as by larger corporations, .e.g Apple’s smart watch and the Fitbit. These advancements have been facilitated by the availability of low-cost microcontrollers offering algorithmic functionality, allowing developers to implement wearables with excellent battery life and edge based real-time data analysis.

Over 90% of the microcontrollers used in the smart product market are powered by so called Arm Cortex-M processors that offer a combination of high algorithmic performance, low-power and security. The Arm Cortex-M4 is a very popular choice with hundreds of silicon vendors (including ST, TI, NXP, ADI, Nordic, Microchip, Renesas), as it offers DSP (digital signal processing) functionality traditionally found in more expensive devices and is low-power. Arm and its rich eco system of partners provide developers with easy-to-use tooling and tried and tested software libraries, such as the CMSIS-DSP and CMSIS-NN frameworks for algorithm development and machine learning.

The choice is vast, and can be very confusing. Therefore, here are some practical hints and tips for both managers and developers to help you decide which Arm Cortex-M processor is best for your biomedical product.

Which Arm Cortex-M processor do I choose for my biomedical application?

The Arm Cortex-M0+ processor is an ultra-low power 32-bit processor designed for very low-cost IoT applications, such as simple wearable devices. The low price point is comparable with equivalent 8-bit devices, but with 32-bit performance. Microcontrollers built around the M0+ processor provide developers with excellent battery life (months to years), a rich peripheral set and a basic amount of connectivity and computational performance. The latter means that only simple algorithms can be implemented, such as algorithms for correcting baseline wander and minimizing the effects of motion artefacts using accelerometer data via an adaptive filter, such as the NLMS algorithm. Although for PPG pulse rate measurement applications, the sampling rate is typically 50Hz, leaving the processor plenty of time to perform various simpler algorithmic operations, such as digital filtering and zero-crossing detection.

For high performance PPG applications, sampling rates in the order of 500Hz are typically used. These types of applications usually look at more biomedical features, such as identifying the Systolic and Diastolic phases and finding the Dicrotic notch using feature extraction algorithms and ML models. These extra functionalities provide a significant strain on the processor’s abilities, and as such are beyond the abilities of the M0+.

The Cortex-M3 is a step up from the M0+, offering better computational performance but with less power efficiency. The extra processing power, rich hardware peripheral set for connecting other sensors and connectivity options makes the M3 a very good choice for developers looking to develop slightly more advanced wearable products, such as the Fitbit device that is based on ST’s low-power STM32L series of microcontrollers.

High performance wearables and beyond

The Arm Cortex-M4 processor and its more powerful bigger brother the Cortex-M7 are highly-efficient embedded processors designed for IoT applications that require decent real-time signal processing performance and memory. Depending on the flavour of the processor, the M4F/M7F processors implement DSP hardware accelerated instructions, as well hardware floating point support. This lends itself to the efficient implementation of much more computationally intensive biomedical DSP and ML algorithms needed for more advanced telemedicine products.

The hardware floating point support unit expedites RAD (rapid application development), as algorithms and functions developed in Matlab or Python can be ported to C for implementation without the need for a lengthy data arithmetic quantisation analysis. Microcontrollers based on the M4F or M7F, usually offer many of the hardware peripheral and connectivity advantages of the M3, providing developers with a very powerful, low power development platform for their telemedicine application.

The Arm Cortex-M33 is a step up from the M4 focusing on algorithms and hardware security via Arm’s TrustZone technology and memory-protection units. The Cortex-M33 processor attempts to achieve an optimal blend between real-time algorithmic performance, energy efficiency and system security.

State-of-the art AI microcontrollers

Released in 2020, the Arm Cortex-M55 processor and its bigger brother the Cortex-M85 are targeted for AI applications on microcontrollers. These processors feature Arm’s Helium vector processing technology, bringing energy-efficient digital signal processing (DSP) and machine learning (ML) capabilities to the Cortex-M family. In November 2023, Arm announced the release of Cortex-M52 processor for IoT applications. This processor looks to replace the older M33 processor, as it combines Helium technology with Arm TrustZone technology.

Although the IP for these processors is available for licencing, only a few IC vendors have developed a microcontroller, e.g. Samsung’s Exynos W920 SoC that has been specifically designed for the wearables market. The SoC packs two Arm Cortex-A55 processors, and the Arm Mali-G68 GPU using state-of-the art 5nm semiconductor technology. The chipset also features a dedicated low-power Cortex-M55 display processor for handling AoD (Always-on Display) tasks – although a little over the top for simple wearable devices, the Exynos processor family certainly seems like an excellent choice for building next generation AI capable low-power wearable products.

So, which one do I choose?

The compromise for biomedical product developers when choosing an M4, M7 or M33 based microcontroller over an M3 device usually comes down to a trade-off between algorithmic performance, security requirements and battery life. If good battery life and simple algorithms are key, then M3 devices are a good choice. However, if more computationally intensive analysis algorithms are required (such as ML models), then the M4 or M7 should be used.

As mentioned earlier, the Armv7E-M architecture used in M4/M7 processors supports a DSP extension that implements an SIMD (single instruction, multiple data) architecture extension that can significantly improve the performance of an algorithm. The hardware floating point unit is very good for expediting MAC (multiply and accumulate) operations used in digital filtering, requiring just three cycles to complete. Other DSP operations such as add, subtract, multiply and divide require just one cycle to complete.

The M7 out performs its M4 little brother by offering approximately twice the computational performance and some devices even offer hardware double precision floating point support which make M4/M7 processors attractive for high accuracy algorithms needed for medical analysis.

If data security is paramount, for example protecting and securing transferring patient data to a cloud service, then the M33 or the M52 (when avalaible) are good choices. These devices also offer a high level of protection against tampering and running of authorised code via TrustZone’s trusted execution environment.

Some IC vendors now offer hybrid micro-controllers that implement multi-processors on chip, such as ST’s ST32Wx family that combine the M0+ and M4 in order to get the advantages of each processor and maximise battery life. 

Finally, advances in semiconductor technology means that a modern M4F processor produced with 40nm process technology may match or even surpass the energy efficiency of an M3 produced with 90nm technology from several years ago. As such, higher performance processors that were until several years too costly and energy inefficient for low-cost wearables products are rapidly becoming a viable solution to this exciting marketplace.

Author

  • Sanjeev is a RTEI (Real-Time Edge Intelligence) visionary and expert in signals and systems with a track record of successfully developing over 25 commercial products. He is a Distinguished Arm Ambassador and advises top international blue chip companies on their AIoT/RTEI solutions and strategies for I4.0, telemedicine, smart healthcare, smart grids and smart buildings.

    View all posts

/by

Building real-time FDA compliant biomedical wearable products: challenges and commercial solutions

,

Recent research suggests that ECG wearables devices (such as smart watches) are now medically suitable for providing predictive insights into serious heart conditions such as atrial fibrillation (A-Fib). These advancements have been facilitated by the availability of low-cost microcontrollers offering algorithmic functionality, allowing developers to implement wearables with excellent battery life and edge-based real-time data analysis.

Although the international research community has produced many innovative high-performance ECG and PPG biomedical algorithms, these are unfortunately limited to offline clinical analysis in Matlab or Python. As such, very little emphasis has been placed on building commercial real-time wearables algorithms on microcontrollers, leading manufacturers to conduct the research themselves and to design suitable candidates. 

This is further complicated by the requirement of manufacturers on how they will implement a developed algorithm in real-time on a low-cost microcontroller and still achieve decent battery life.

Arm Cortex-M microcontrollers

Over 90% of the microcontrollers used in the smart product market are powered by so-called Arm Cortex-M processors that offer a combination of high algorithmic performance, low-power and security. The Arm Cortex-M4 is a very popular choice with hundreds of silicon vendors (including ST, TI, NXP, ADI, Nordic, Microchip, Renesas), as it offers DSP (digital signal processing) functionality traditionally found in more expensive devices and is low-power.

The Cortex-M4F device offers floating point support, helping with RAD (rapid application development) as designs can be easily ported from Matlab/Python to C without the need of performing a detailed quantisation arithmetic analysis. As such, a design cycle can be cut from months to weeks, offering organisations a significant cost saving.

Arm and its rich ecosystem of partners provide developers with easy-to-use tooling and tried and tested software libraries, such as the CMSIS-DSP and CMSIS-NN frameworks and ASN’s DSP filtering library for algorithm development and machine learning.

FDA compliance

The AHA (American Heart Association) provides developers with guidelines for developing FDA-compliant ECG monitoring products. These are broken down into the following three categories: 

  1. Diagnostic: 0.05Hz -150Hz
  2. Ambulatory (wearables): 0.67Hz – 40Hz
  3. ST segment: 0.05Hz

The ECG measurements must be FDA compliant with IEC 60601-2 2-47 standards for ambulatory ECG, but what are the criteria and challenges?

Challenges with ECG/PPG measurements

Modelling the QRS complex found in ECG data is extremely difficult, as to date there is no concrete model available.  This is further complicated by the variety of ECG data depending on the position of the lead on the patient’s body and illnesses. The following list summaries the typical challenges faced by algorithm developers:

  1. Accurate baseline wander (BLW) removal remains one of the most challenging topics in ECG analysis.
  2. The BLW must be removed for accurate clinical analysis.
  3. BLW manifests itself as low-frequency ‘wander’ (typically <0.5Hz) from EMG and torso movement.
  4. QRS width widening and amplitude distortion due to filtering invalidates clinical analysis.
  5. Reducing EMG and measurement noise without altering the temporal biomedical relationships of the ECG signal.
  6. 50/60Hz powerline interference can swamp the ECG signal – this is primarily attributed to pickup by the long high impedance measurement cables. This is typically problematic for extended bandwidth wearable applications that go beyond 40Hz.
  7. Glitches, sudden movement and poor sensor contact with the skin: This is related to BLW, but usually manifests itself as abrupt glitches in the ECG measurement data. The correction algorithm must discriminate between these undesirable events and normal behaviour.
  8. IEC 60601-2 2-47 frequency response specifications:
    • Bandwidth: 0.67Hz – 40Hz.
    • Passband ripple: < ±0.5dB
    • Maximum ±10% amplitude error: most biomedical SoCs make use of a Sigma-Delta ADC, leading to amplitude droop.

Shortcomings with ECG/PPG algorithms

A mentioned in the previous section, much research has been conducted over the years with mixed results. The main shortcomings of these methods are summarised below:

  1. Computationally heavy: most algorithms have been designed for research in Matlab and not for real-time, e.g. wavelets have excellent performance but have high computational cost, leading to poor battery life and the need for an expensive processor.
  2. Large latency and warping: digital filtering chain introduces large latency, computational cost and can warp the characteristics of the biomedical features.
  3. Overlapping frequencies: there are many examples of unwanted noise overlapping the delicate ECG data, hence the popularity of time-frequency analysis, such as wavelets.
  4. Mixed results regarding BLW removal: spline removal is excellent, but it has high computational cost and has the added difficulty of finding good correction points between the QRS complexes. Linear phase FIR filtering is a good compromise but has very high computational cost (typically >1000 filter coefficients) due to the high sampling rate to cut-off frequency ratio. Non-linear phase IIR filter has low computational cost, but warps the ECG features, and is therefore unsuitable for clinical analysis.
  5. AI based kernel filters: ‘black box filter’ based on massive training data. Moderate implementation cost with performance dependent on the variety of training data, leading to unpredictable results in some cases.
  6. PPG analysis: has the added difficulty of eliminating motion from the measurement data, such as when walking or running. Although a range of tentative algorithms has been proposed by various researchers using accelerometer measurement data to correct the PPG data, very few commercial solutions are currently based on this technology.

It would seem that ECG and PPG analysis has some major obstacles to overcome, especially when considering how to deploy the algorithms on low-power microcontrollers.

The future: ASN’s real-time RCF algorithm and Advanced Analytics

Together with cardiologists from Medisch Spectrum Twente, ASN’s advanced analytics team developed the RCF (retrospective collaborative filtering) algorithm that uses time-frequency analysis to enhance the ECG data in real-time.

The essence of RCF algorithm centres around a highly optimised set of polynomial cleaning filters with different frequency characteristics that are applied to different segments of the QRS complex for enhancement. This has some synergy with wavelets, but it does not suffer from the computational burden associated with wavelet analysis.

The polynomial filters are peak preserving, meaning that they preserve the delicate biomedical peaks while smoothing out the unwanted noise/ripple. The polynomial fitting operation also overcomes the challenge of overlapping frequency content, as data within a specified region is smoothed out by the relevant filter.

RCF is further strengthened by the BLW killer IP block that implements a highly computationally efficient linear phase 0.67Hz highpass filter. The net effect is an FDA-compliant signal chain suitable for clinical analysis. The complete signal chain is extremely computationally efficient, and as such is suitable for Arm’s popular M3 and M4 Cortex-M processor families.

Real-time ECG feature extraction

The ECG waveform can be split up into segments, where each wave or segment represents a certain event in the cardiac cycle, as shown below:

As seen, the biomedical features are designated P, Q, R, S, T that define points in time within the cardiac cycle. The RCF algorithm is further strengthened with our state-of-the-art AAE (Advanced Analytics Engine) that automatically cleans and find these features for clinical analysis.

AAE supported analytics

  1. P-wave duration
  2. PR interval
  3. QRS duration
  4. QT duration (Bazett algorithm used for QTc)
  5. HR (RR interval)
  6. HRV (rMSSD algorithm used)

Armed with the real-time features, an ML model can be trained and provide valuable insights into patient health running on an edge processor inside a wearable device.

A-Fib

Atrial fibrillation (A-Fib) is the most frequent cardiac arrhythmia, affecting millions of people worldwide. An arrhythmia is when the heart beats too slowly, too fast, or in an irregular way. Signs of A-Fib are an irregular beating pattern and no p-waves. Our AAE provides developers with all of the relevant features needed to build an ML model for robust A-Fib detection.

Let us help you build your product

By combining advanced low-power processor technology, advanced mathematical algorithmic concepts and medical knowledge, our solution provides developers with an easy way of building wearable products for medical use. The high accuracy of our Advanced Analytics Engine (AAE) has been verified by cardiologists, and can be used with an additional ML model or standalone to provide people with valuable insights into potentially fatal health conditions, such as A-Fib without the need for an expensive medical examination at a hospital.

ASN’s ECG algorithmic solutions are ideal for building next generation ECG and PPG wearable products on Arm Cortex-M microcontrollers (e.g. STM32F4, MAX32660) and bio-sensor SoCs (MAX86150). These algorithms can be easily used with industry standard biomedical AFEs, such as: MAX30003, AFE4500 and AFE4950.

Please contact us for more information and to arrange an evaluation.

Author

  • Sanjeev is a RTEI (Real-Time Edge Intelligence) visionary and expert in signals and systems with a track record of successfully developing over 25 commercial products. He is a Distinguished Arm Ambassador and advises top international blue chip companies on their AIoT/RTEI solutions and strategies for I4.0, telemedicine, smart healthcare, smart grids and smart buildings.

    View all posts
/by

FIR notch filter from a lowpass filter

,

Designs an FIR notch filter from a lowpass filter by computing the difference between the prototype lowpass filter and its amplitude complementary

Read more
/0 Comments/by

Sensors in Biomedical AIOT

,

Biomedical devices are at the forefront of AI and IOT (more often called AIOT). What is your most important reason to use sensors for biomedical devices?

Biomedical sensors for ai, iot and aiot to optimize

To control

  • Does the patient follow the medical instructions? Examples: is he doing his therapy on time and in the right way. Does he take his medications?  Especially groups of risk can be monitored so that timely action can be taken if necessary
  • Is treatment going well?  For both doctor and client alike. And even better:  You can optimize the healing process
  • Do medically devices still give the right measurement?
sensors biomedical devices optimize ai iot aiot

To optimize

  • Optimize your treatment: Compare the treatment results from your client with your other clients. And thus, find out point of improvement
  • Give attention for those who need it. Nobody wants to spend time unnecessary in a waiting room
  • Better use of existing resources
  • Connect systems with each other
  • Take the right decisions at the right time
  • Preventive maintenance Security

To innovate

  • Better serve your clients
  • Be at the forefront of medical developments
  • Track & trace
  • Create optimal circumstances with modern technology
sensors for biomedical devices iot ai aiot

To save

  • Give the client the best care
  • Spend your budget where most needed
  • To prevent is better than to cure
  • Prevent greater suffering, avoid extra high costs
  • Nobody is waiting for unnecessary treatment
  • Preventive maintenance on medical devices prevents higher repair costs and downtime
/0 Comments/by

Biomedical IoT: to save and improve life

,

How biomedics can benefit from IoT:

  • Improve the quality and effectiveness of your medical treatment with IoT and thus the quality of life for your patient
  • Do you want to spend your time on patients who really need it? Treat more patients in a better way in the same time within the same budget
  • Improve and fasten your patient’s recovery… with knowledge of results of the treatment of thousands of patients
  • Be on the forefront of IoT and help your patients in ways that were not possible before
  • Better measurement with clean data
  • Security: get control of your patients’ privacy

Sensor devices improve the quality of treatment and thus of life itself. Through prevention, medical treatment itself or therapy. Sensors solutions help to improve life and can even help to save lives. No wonder medical devices are one of the fastest growing sectors in IoT.

Households want to benefit from the newest possibilities IoT devices deliver. For better health and independence as much and as long as possible. Sensor devices can especially help people with chronic diseases, elderly people and people who need constant supervision.

For medical service providers, sensor devices can help to provide the best service as possible in times of increasing medical demand but also on a growing focus on costs. So, budgets must be spent in the best way as possible. Sensors and other IoT devices can help medical facilities to treat more patients in a better way and more efficiently.

How Advanced Solutions helps

Our services can help you in the following ways:

  • IoT solutions
  • Sensor measurement
  • Security

IoT solutions

Our systems prevent accidents from happening and reduce the response time of healthcare professionals.

Sensor measurement

In biomedical applications, sensors are already used in a lot of ways. EMG and ECG are just some examples. For making the right decisions on treatment, the measurement has to be accurate. However, most sensor data suffer from the following problems:

  • Powerline interference, measurement noise and glitches
  • Baseline wander and offsets from analog instrumentation electronics
  • Biomedical sensors are sending weak signals. So, in some cases, those signals have to be strengthened before analysis is possible.

Sometimes the filtering is done in the hardware of the medical equipment. However, these are very costly solutions. And most of the times, a simpler solution and thus less costly solution is what is being sort after.

ASN Consultancy is the modern way of working of algorithm design to separate the wanted sensor signals from the undesirable unwanted signals. So, you can analyse and take action on clean and accurate sensor data.

/0 Comments/by

Practical noise reduction tips for biomedical ECG filters

, ,

In ECG signal processing, the Removal of 50/60Hz powerline interference from delicate information rich ECG biomedical waveforms is a challenging task! The challenge is further complicated by adjusting for the effects of EMG, such as a patient limb/torso movement or even breathing. A traditional approach adopted by many is to use a 2nd order IIR notch filter:

\(\displaystyle H(z)=\frac{1-2cosw_oz^{-1}+z^{-2}}{1-2rcosw_oz^{-1}+r^2z^{-2}}\)

where, \(w_o=\frac{2\pi f_o}{fs}\) controls the centre frequency, \(f_o\) of the notch, and \(r=1-\frac{\pi BW}{fs}\) controls the bandwidth (-3dB point) of the notch.

What’s the challenge?

As seen above, \(H(z) \) is simple to implement, but the difficulty lies in finding an optimal value of \(r\), as a desirable sharp notch means that the poles are close to unit circle (see right).

In the presence of stationary interference, e.g. the patient is absolutely still and effects of breathing on the sensor data are minimal this may not be a problem.

However, when considering the effects of EMG on the captured waveform (a much more realistic situation), the IIR filter’s feedback (poles) causes ringing on the filtered waveform, as illustrated below:

Contaminated ECG with non-stationary 50Hz powerline interference (FIR filtering), ECG sigal processing, ECG DSP, ECG measurement

Contaminated ECG with non-stationary 50Hz powerline interference (IIR filtering)

As seen above, although a majority of the 50Hz powerline interference has been removed, there is still significant ringing around the main peaks (filtered output shown in red). This ringing is undesirable for many biomedical applications, as vital cardiac information such as the ST segment cannot be clearly analysed.

The frequency reponse of the IIR used to filter the above ECG data is shown below.

IIR notch filter frequency response, ECG signal processing, ECG DSP, ECG measurement

IIR notch filter frequency response

Analysing the plot it can be seen that the filter’s group delay (or average delay) is non-linear but almost zero in the passbands, which means no distortion. The group delay at 50Hz rises to 15 samples, which is the source of the ringing – where the closer to poles are to unit circle the greater the group delay.

ASN FilterScript offers designers the notch() function, which is a direct implemention of H(z), as shown below:

ClearH1;  // clear primary filter from cascade
ShowH2DM;   // show DM on chart

interface BW={0.1,10,.1,1};

Main()

F=50;
Hd=notch(F,BW,"symbolic");
Num = getnum(Hd); // define numerator coefficients
Den = getden(Hd); // define denominator coefficients
Gain = getgain(Hd); // define gain

Savitzky-Golay FIR filters

A solution to the aforementioned mentioned ringing as well as noise reduction can be achieved by virtue of a Savitzky-Golay lowpass smoothing filter. These filters are FIR filters, and thus have no feedback coefficients and no ringing!

Savitzky-Golay (polynomial) smoothing filters or least-squares smoothing filters are generalizations of the FIR average filter that can better preserve the high-frequency content of the desired signal, at the expense of not removing as much noise as an FIR average. The particular formulation of Savitzky-Golay filters preserves various moment orders better than other smoothing methods, which tend to preserve peak widths and heights better than Savitzky-Golay. As such, Savitzky-Golay filters are very suitable for biomedical data, such as ECG datasets.

Eliminating the 50Hz powerline component

Designing an 18th order Savitzky-Golay filter with a 4th order polynomial fit (see the example code below), we obtain an FIR filter with a zero distribution as shown on the right. However, as we wish to eliminate the 50Hz component completely, the tool’s P-Z editor can be used to nudge a zero pair (shown in green) to exactly 50Hz.

The resulting frequency response is shown below, where it can be seen that there is notch at exactly 50Hz, and the group delay of 9 samples (shown in purple) is constant across the frequency band.

FIR Savitzky-Golay filter frequency response, ECG signal processing, ECG DSP, ECG measurement

FIR  Savitzky-Golay filter frequency response

Passing the tainted ECG dataset through our tweaked Savitzky-Golay filter, and adjusting for the group delay we obtain:

Contaminated ECG with non-stationary 50Hz powerline interference (FIR filtering), ECG signal processing, ECG digital filter, ECG filter designa

Contaminated ECG with non-stationary 50Hz powerline interference (FIR filtering)

As seen, there are no signs of ringing and the ST segments are now clearly visible for analysis. Notice also how the filter (shown in red) has reduced the measurement noise, emphasising the practicality of Savitzky-Golay filter’s for biomedical signal processing.

A Savitzky-Golay may be designed and optimised in ASN FilterScript via the savgolay() function, as follows:

ClearH1;  // clear primary filter from cascade

interface L = {2, 50,2,24};
interface P = {2, 10,1,4};

Main()

Hd=savgolay(L,P,"numeric");  // Design Savitzky-Golay lowpass
Num=getnum(Hd);
Den={1};
Gain=getgain(Hd);

Deployment

This filter may now be deployed to variety of domains via the tool’s automatic code generator, enabling rapid deployment in Matlab, Python and embedded Arm Cortex-M devices.

ASN Filter Designer box
Download Demo
Pricing and Licencing

Author

  • Sanjeev is a RTEI (Real-Time Edge Intelligence) visionary and expert in signals and systems with a track record of successfully developing over 25 commercial products. He is a Distinguished Arm Ambassador and advises top international blue chip companies on their AIoT/RTEI solutions and strategies for I4.0, telemedicine, smart healthcare, smart grids and smart buildings.

    View all posts

/by

Contactless vital life signs measurement… Fact or science-fiction?

Generations have been entertained by the gadgets and future technology portrayed in Sci-fi series, such as Star Trek, but is it all science fiction?


The Tricorder
One device that intrigued me for years was the so called ‘Tricorder’ and the ability for the doctor to read a person’s vital life signs or VLS (e.g. heart beat and respiration) with a handheld device from about a metre away.


Back to the 21st century!
With advances in radar technology over the last few years, a few chip manufacturers are now producing affordable radar devices suitable for biomedical VLS measurement. Radar technology that used to cost thousands of Euros, and was primarily aimed a military technology, is now available for a few hundred Euros, making it viable for home medical products.


Sounds great, but what can UWB (ultrawide band) pulse doppler radar do?

  • millimetre accuracy: allowing for detection of the smallest changes, such as respiration and heart rate from several metres away.
  • Penetrates duvets, blankets and clothes: The virtue of the small wavelength ensures accurate detection of humans in a bed or sitting in chair watching TV or reading a book.
  • Penetrates Walls and doors: Tracking of VLS and movement when radar is mounted within a ceiling or behind a wall – no ugly module on the wall!
  • High sensitivity: able to see the VLS from tiny premature babies.
  • Ultra-safe technology: RF emission is 0.01% the energy typically found in a household WiFi router – meaning that prolonged exposure will have no detrimental effects on human health.

After receiving a request from a client about monitoring the health of an eldery person living alone or in a nursing home, we decided to conduct a few tests of our own to see what was possible with this new technology.


VLS of a subject lying in bed
After building a demonstrator, and placing the radar sensor about 1 metre from a subject sleeping (similar to the doctor in Star Trek), we obtained the following waveform:

Captured VLS data captured from a UWB radar – containing both heart beat and respiration information

Wow! Was our initial reaction to the test data – this is millimetre movement through a duvet! Notice how slow the biomedical signal is, as an average adult’s respiration rate at rest is about 12bpm, requiring relatively long data acquisition times (tens of seconds) for meaningful data analysis.


What’s the respiration and heart beat?
Passing the signal through our algorithm, we could easily estimate the respiration (RR) and heartbeat (HB) from the plot (see the two red squares on the two peaks). However, in order to be objective, we attached a clip-on pulse oximeter to the subject finger, and as seen they matched very well.

What does this all mean for me?
Contactless VLS measurement for home use is closer than you think, and is certainly not science-fiction anymore. This technology opens up many possibilities for monitoring when normal sensors are infeasible, such as premature babies, patients with dementia and even sleep trend analysis. We’ll improve our algorithm in order to make it more robust and faster, but as seen our results are very promising indeed, and open up the possibility of contactless vital life signs (VLS) measurements for many practical applications!

/0 Comments/by