Innovations in Medical Sensors for Biomedical Electronics Applications in Healthcare

Students can refer this post to start working on their projects. Students can take ideas and start the project

Biomedical electronics research is being driven by the aging “baby-boomer” population and their medical needs. This phenomenon is spurring fast development of new biotechnologies and need for access to innovative means of medical diagnosis and treatment in preventive medicine. Subsequently, the technologies of implants and advanced wireless electronic media will help alleviate rising medical costs in today’s society and extend the average longevity with a quality life in our later years.

This post will discuss the brain, heart and lung in relation to advances in technology that will improve the bridge between engineering, biology and medicine that will enhance the function of these organs.

The article will show how miniaturization, portability, connectivity, humanization, security and reliability in new devices have advanced this effort to improve on the fragile nature and amazing balance needed in aging or diseased/damaged organs in the human body.

The brain
Closed-loop deep brain stimulation (CDBS), for people with epilepsy, Parkinson’s Disease (PD) or even obsessive compulsive disorder (OCD), is a prime example of implementation of a biomedical electronic solution that offers an enhancement to the quality of life in someone afflicted with such conditions.

The DBS system detects a patient’s electroencephalogram (EEG) and automatically generates DBS electrical pulses to prevent the onset of an epileptic seizure or even helps lessen the tremors of PD. DBS sends specific stimuli to different regions of the brain. DBS is used in patients who are resistant to drug treatment and those who suffer from motor fluctuations and tremors.

To date only Medtronic has an FDA-approved DBS product. Their bilateral DBS devices, which were approved by FDA in 2002 consists of two neurostimulators, one for each side of brain. Similar to a cardiac pacemaker, the DBS uses a neurostimulator to generate and deliver high-frequency electric pulses into subthalamic nucleus (STN) or globus pallidus internus (GPi) portions of the brain through extension wires and electrodes. The Medtronics Soletra neurostimulator is one of the most advanced battery-operated devices.

The neurostimulator is typically programmed post-surgery by trained technicians to find the most effective signal parameters for alleviating Parkinson’s symptoms. See Figure 1 for a simple block diagram of Medtronic’s standard DBS product.

Figure 1: Block diagram of the Medtronic deep brain stimulation system, which uses a neurostimulator to generate and deliver high-frequency electric pulses into portions of the brain. (Courtesy of Medtronics.)

A proposed basic design of a CDBS is as follows1:

The CDBS device directly interfaces with recording and stimulation electrodes. Eight recording electrodes are implanted in the motor cortex and 64 stimulating electrodes are implanted in the STN portion of the brain. The 64-channel point-controllable stimulation enables the formation of various stimulus patterns for the most effective treatment of Parkinson’s symptoms.

The collected neural signals from the embedded microelectrodes should be conditioned using eight front-end low-noise neural amplifiers (LNAs). Due to the low-level amplitude of the neural spikes, integrated pre-amplifiers are sometimes used to amplify the small signals before the data conversion. The front-end design needs to be low noise to guarantee the signal integrity.

The front-end band-pass LNAs typically have a gain on the order of 100 and the LNA input design needs to minimize 1/f noise. A switched capacitor technique can be used for resistor emulation and 1/f noise reduction. The switched-capacitor circuit modulates the signal so that 1/f noise may be reduced to below thermal noise. The switched-capacitor amplifying filter performs well in recording neural spikes and field potential, simultaneously.

The LNA’s are in turn multiplexed to a single high-dynamic-range logarithmic amplifier front end into an analog-to-digital converter (ADC) making analog automatic gain control unnecessary.

To cover the entire range of both small-signal neural spikes and large-signal local field potential (LFP) responses from the brain stimulation, the high-dynamic-range ADC is needed to digitize all the desired neural information. The logarithmic amplifier, used in front of the ADC, is able to achieve this needed dynamic range. Logarithmic encoding is well-suited to neural signals and is efficient, since a large dynamic range can be represented with a short word length. To save area and power consumption, the relatively large-dynamic-range ADC is used, making analog automatic gain control unnecessary.

The ADC needs a digital filter which separates the low frequency neural field potential signal from the neural spike energy. Separation of the low-frequency field potential from the higher frequency spike energy can be done with a 22-tap finite-impulse-response (FIR) Butterworth-type digital filter.

Using digital filters instead of analog or mixed-signal filters provides many advantages. First of all, a digital filter is programmable so that its operation may be adjusted without modifying hardware while generally an analog filter may be changed only by modifying the design. A digital filter is used for diplexers to separate two frequency bands of spikes and LFPs. While analog filter circuits are subject to drift and are dependent on temperature, a digital filter does not suffer from these issues, and is extremely robust with respect to both time and temperature.

The electrical stimulator generates 64 channels of biphasic charge-balanced current stimulation. A dedicated controller generates these stimulation patterns via an I/O channel to control the 64 current-steering DACs. The 64 DACs can be formed as a cascade of a single shared 2-bit coarse current DAC and 64 individual bi-directional 4-bit fine DACs or other similar configuration.

The DAC has 48 possible current values. One can use a fine ADC and a polarity switch selects the positive or negative DAC output to achieve charge-balanced bi-phasic stimulation, helping to reduce the risk of long-term tissue damage.

Figure 2 below is a single chip for the CDBS system2 that when interfaced with a microprocessor gives a complete CDBS system. “A microprocessor gives information to the chip about where and how, and the chip takes care of the rest,” says Michael Flynn, a leader of the project.

Figure 2: Typical closed-loop deep brain stimulation (CDBS) chip system block diagram (Courtesy of PHOTO: Michael Flynn & Daryl Kipke in article by Morgen E. Peck, “A Chip to Better Control Brain Stimulators for Parkinson’s” 2008).

In the medical electronics arena, Freescale has partnered with Cactus Semiconductor, which does custom analog design. Cactus Semiconductor’s medical focus in integrated circuit design encompasses both implantable and portable applications such as neuro-stimulation, pacing, defibrillation, ultrasound, and medical monitoring (e.g., glucose meters). (See sidebar.)

Freescale has suitable medical solutions with low-power microcontrollers, integrated analog front ends (AFEs) and algorithms that consume less power. Their wireless communication solutions assure low-power operating modes as well as good sleep modes with quick wake-up.

For the next generation of DBS and tools to help researchers understand the mystery of the brain, Medtronic is developing the Bi-directional Brain-machine interface3 (BMI). This technique promises to be a major development tool in the frontier of brain research once all lab tests are completed and approved for use with human brain studies in the near future. Right now it is in preclinical research. There are no approved products yet.

As illustrated in the functional block diagram of Figure 3, the neural interface (NI) technology core is the existing stimulator and telemetry system found in released neurostimulators (ActivaPC by Medtronic).

Figure 3: In this functional block diagram for a bi-directional neural interface system, the neural interface (NI) technology core is the existing stimulator and telemetry system found in released neurostimulators (Courtesy of IOP publishing).

Referencing Figure 4, the sensor hardware, algorithm processor and firmware partition are inserted into the existing infrastructure with well-defined signal pathways in the physical and algorithmic domains.

Figure 4: The sensor hardware, algorithm processor and firmware partition in the bi-directional brain-machine interface prototype are inserted into the existing infrastructure with well-defined signal pathways in the physical and algorithmic domains. (Courtesy of IOP publishing).

The heart
“Small size,” “wireless,” and “non-contact” are words that could never have been associated with ECG devices in the past. New developments in electronics now enable more compact/portable designs, some that include wireless capability and even those sensors that require no physical or resistive contact with the body.

Small size in ECG designs are accomplished with such integrated circuit developments as Texas Instruments’ highly integrated ADS1298R AFE that also includes a fully-integrated, respiration impedance measurement function. See Figure 5 below that shows an integrated AFE device like the ADS1298 plus the other important parts of an ECG architecture.

Figure 5: An electrocardiogram solution showing an integrated analog front end (AFE) device (Courtesy of Texas Instruments).

ECG System Functionality and Evolution4 Basic functions of an ECG machine include ECG waveform display, either through LCD screen or printed paper media, and heart rhythm indication as well as simple user interface through buttons. More features, such as patient record storage through convenient media, wireless/wired transfer and 2D/3D display on large LCD screen with touch screen capabilities, are required in more and more ECG products.

Multiple levels of diagnostic capabilities are also assisting doctors and people without specific ECG training to understand ECG patterns and their indication of a certain heart condition (We will discuss Monebo algorithms below for this added benefit). After the ECG signal is captured and digitized, it will be sent for display and analysis, which involves further signal processing.


Freescale’s Dr. Jose Fernandez Villaseñor, Physician, Surgeon and Electrical Engineer, says, “There is always room for improvement in both surgical techniques and in the electronics and software used to control the (DBS) pacemaker. Electronics are particularly critical in that they must correctly detect when the patient’s brain cells are malfunctioning and in turn, when and when not to compensate. I believe research needs to be conducted on new control software and accuracy needs to be improved in sensors and processing units in order to reduce the likelihood of complications.” He goes on to say, “As technology providers, we can only hope to expedite this process by creating the most effective and safe solutions possible.”

Here is a quote from Tim Denison regarding Medtronic’s approach to sharing what it knows so we can engage the scientific community in the development of new technology as neural interfacing develops.

“Neural interfacing is a relatively new area, and there’s much we don’t know. Medtronic views the discovery, development and deployment of brain-machine interface technology a participative process. We’ve been open about sharing our models, so we can engage the world’s best scientific minds in the development of new tools that in the short term enable the next generation of therapies to treat chronic, debilitating conditions, like Parkinson’s disease, and over time, could potentially lead to new treatments based on the interpretation of brain signals.”

— Tim Denison, Ph.D., Engineering Director and Technical Fellow in the Neuromodulation division of Medtronic, Inc.

Signal Acquisition challenges:

    • Measurement of the ECG signal can be very challenging due to the presence of the large DC offset and various interference signals. This potential can be up to 300mV for a typical electrode. The interference signals include the 50/60-Hz interference from the power supplies, motion artifacts due to patient movement, radio frequency interference from electro-surgery equipment, defibrillation pulses, pace maker pulses, other monitoring equipment, etc.


    • Depending on the end equipment, different accuracies and bandwidth requirements will be needed in an ECG:
      • Standard monitoring needs frequencies between 0.05-30 Hz
      • Diagnostic monitoring needs frequencies from 0.05-1000 Hz


  • Some of the 50Hz/60Hz common mode interference can be cancelled with a high-input-impedance instrumentation amplifier (INA), which removes the AC line noise common to both inputs. To further reject line power noise, the signal is inverted and driven back into the patient through the right leg by an amplifier. Only a few micro amps or less are required to achieve significant CMR improvement and stay within the UL544 limit. In addition, 50/60Hz digital notch filters are used to reduce this interference further.

Analog front end options:

    • Optimizing the power consumption and the PCB area of the analog front end is critical for portable ECG’s. Due to technological advancements, there are now several front end options:
      • Using a low resolution ADC (needs all filters)
      • Using a high resolution ADC (needs fewer filters)
      • Using a sigma-delta ADC (needs no filters, no amplifier aside from INA, no DC offset)
      • Using a sequential Vs simultaneous sampling approach.


    • When a low resolution (16 bit) ADC is used, the signal needs to be gained up significantly (typically 100x – 200x) to achieve the necessary resolution. When a high resolution (24bit) sigma delta ADC is used, the signal needs a modest gain of 4 – 5x. Hence the second gain stage and the circuitry needed to eliminate the DC offset can be removed. This leads to an overall reduction in area and cost. Also the delta sigma approach preserves the entire frequency content of the signal and gives plenty of flexibility for digital post processing.


  • When using a sequential approach the individual channels converting the leads of an ECG are multiplexed to one ADC. This way there is a definite skew between adjacent channels. With the simultaneous sampling approach, a dedicated ADC is used for each channel and hence there is no skew introduced between channels.

Freescale has a great low-cost development board, called the MED-EKG Module, that is an extremely versatile system which allows designers to quickly prototype an electrocardiogram system. When used as part of the Freescale’s Tower system, designers get a fully functioning system that can then be easily modified, changed or upgraded to a custom design by replacing any of the individual modules that are a part of this kit with a custom designed board.

In addition, using the Monebo Kinetic ECG algorithm will allow the designer to provide the user signal processing and interpretation of the ECG waveform, thereby aiding health care professionals in assessing cardiac parameters. It provides highly accurate QRS (A combination of three of the graphical deflections seen on a typical electrocardiogram–It is usually the central and most visually obvious part of the tracing) detection and feature extraction, beat classification, interval measurement, and rhythm interpretation for up to 16 leads of captured ECG data.

Contactless ECG is no longer science fiction. Plessey Semiconductors and the University of Sussex have developed the electric potential integrated circuit (EPIC) sensor, an electric potential sensing (EPS) technology where an array of these sensors can just be held over the patient’s chest to obtain readings to give the equivalent of a 12 lead ECG without the mess of wiring, contact gel and electrodes that can easily become detached.

The lungs
A medical ventilator, also called respirator help or mechanical ventilator (MV), pushes air into the patient’s lungs. Ventilators are used, for example, in intensive care medicine for artificial ventilation or in home care to treat sleep apnea. Modern devices use intelligent circuits to mix the gas or set a fixed or modulated speed of the fan based on the input from sensors. The ST Microelectronics solution contains all semiconductors needed as well as certified software libraries to allow silent and reliable operation.

Mechanical ventilators have saved many lives since their development and usage in hospitals and care institutions. Patients in intensive care units (ICU) surviving on an MV for greater duration than one week have increased risk of medical complications such as ventilator-acquired pneumonia (VAP) and nosocomial infections and are seven times likely to die in the ICU. See Figure 6.

Figure 6: Typical respirator aid block diagram (Courtesy of ST Microelectronics).

Patients on an MV rapidly develop atrophied diaphragm muscles and will have greater difficulty of weaning from the ventilator as time goes on.

Avery Biomedical has developed a breathing pacemaker system that uses radio frequency (RF) coupled receivers to transcutaneously send both power and signal. This is important for two reasons:

1. There is no implanted battery and therefore nothing to wear-out internally. Barring mechanical damage, the implant is expected to last the lifetime of any patient it is put in regardless of their age.

2. There is no percutaneous connection between the implanted and external components. Since the patient’s skin is intact, there are no long-term issues with wound care management and no chronic infection risk.

The other key point is that the system operates on the principle of negative pressure ventilation. That is, by causing the diaphragm to contract the pressure inside the lungs becomes lower than atmospheric pressure and air is drawn in. This is physiologically correct and is we are breathing right now. Positive pressure ventilation (whether by mask or a mechanical ventilator) pushes air in which is not natural and is associated with a high risk of VAP, or ventilator associated pneumonia. VAP is the most common reason for hospital readmissions of ventilator-dependent patients. Reducing hospital readmissions (by paying less for them through Medicare/Medicaid) is a focus of the recent healthcare reforms. See Figures 7 and 8.

Figure 7: Breathing pacemaker showing implanted electrodes and RF receiver for lung phrenic nerve stimulation and external antenna which transmits RF signal to the implant for stimulation pacemaker function (Courtesy of Avery Biomedical).

Figure 8: Basic block diagram of the breathing pacemaker (Courtesy of Avery Biomedical).

For the next generation of devices5 in the works, new developments and even less invasive using endovascular electrodes suitable for percutaneous (any medical procedure where access to inner organs or other tissue is done via needle-puncture of the skin, rather than by using an “open” approach where inner organs or tissue are exposed) insertion only using local anesthesia in critically ill patients, the phrenic nerves may be electrically paced to maintain diaphragm strength and resistance to fatigue, improved ventilation, and the earlier possibility of being weaned from the MV. This technique also can shorten the ICU stay, reduce mortality and decrease hospital costs once it is approved by the FDA and associated agencies.

Rhythmic diaphragm contractions can be produced by proper nerve stimulation with this minimally invasive technique. The threshold potential required for phrenic nerve stimulation is 1.26 V. The required current to activate the nerve is predicted to be three times less in a shielded electrode than in a lead-type electrode. Balanced biphasic pulses of 180 us phase duration are typically used.

New commercial sensor and microcircuit innovations for hand-held devices like the iPhone, Blackberry and iPad require low cost, small size and low power. These efforts are spilling over into the biomedical electronics area and will bring more amazing solutions to improve implantables and ultimately eliminate the need for most medical implantables via non-contact stimuli and sensing devices like inductive power and data transmission and lower power RF devices.

1. “A Closed-Loop Deep Brain Stimulation Device with a Logarithmic Pipeline ADC” by Jongwoo Lee; A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy (Electrical Engineering) in The University of Michigan 2008.

2. Morgen E. Peck, “A Chip to Better Control Brain Stimulators for Parkinson’s” 2008.

3. “A chronic generalized bi-directional brain-machine interface”, A G Rouse, S R Stanslaski, P Cong, R M Jensen, P Afshar, D Ullestad, R Gupta, G F Molnar, D W Moran and T J Denison, Journal of Neural Engineering, J. Neural Eng. 8 (2011) 036018 (19pp)May 2011, IOP Publishing


5. “Diaphragm Pacing with Endovascular Electrodes”, Hoffer J.A., Tran BD, Tang JK, Saunders JTW, Francis CA, Sandoval RA, Meyyappan R, Seru S, Wang HDY, Nolette MA, Tanner AC.

Enhanced by Zemanta
Related Posts Plugin for WordPress, Blogger...