Fig 1 - uploaded by Mohamad Sawan
Content may be subject to copyright.
Source publication
In this paper, we present a fully integrated front-end of a portable spectroreflectometry-based brain imaging system dedicated for acquisition of modulated optical signals at a frequency of 1 Hz to 25 kHz. The proposed front-end preamplifier is composed of a photodetector, a transimpedance preamplifier, a two-stage voltage amplifier and a mixer. St...
Context in source publication
Context 1
... or its power dissipation. A predetermined number of emitters and receivers, two distinct chips, are placed noninvasively on the surface of the scalp of the subject. The proposed front-end involves modulated optical signals up to 50 kHz in the NIR region (735-850 nm). A block diagram of the complete front-end receiver chip is illustrated in Fig. ...
Similar publications
The barrier layer in InAs/GaSb LWIR nBn detector is usually composed of AlGaSb alloy, which has a non-negligible valence band offset and is sensitive to chemical solutions. In this work, we investigated a type-II superlattice (T2SL) barrier that is homogeneous with the T2SL absorber layer in order to resolve these drawbacks of the AlGaSb barrier. T...
Citations
... Afterwards, the increased signal is provided to the subsequent stage and provide the input to the ADC.Optical receiver's bandwidth and power consumption entails a trade-off for FD-NIRS. The mean time from source to detector is approximately 1 ns for 720 nm NIR light for human skull [7] [8]. During the process, 1 ns signal delay is experienced, for which the next phase resolving circuit needs to be capable of extracting that small delay. ...
In this paper, design and analysis of High Gain CMOS Transimpedance Amplifier (HGCTIA) is presented which is greatly necessitated for frequency domain functional Near Infrared Spectroscopy (FD-f NIRS). Generally, feedback resistor and drain resistor play a major role in deciding TIA gain and bandwidth. Most of the research work have utilized only feedback resistor and drain resistor. Here, the gain improvement is greatly achieved by replacing the drain resistor by an active inductor which is designed using CMOS transistor. The active inductor loaded based TIA exhibits only 0.01mW power dissipation which is comparatively lesser than previous research works. Also the proposed design achieves maximum value of transimpedance gain 140dB, bandwidth of 10 MHz, where average noise over the bandwidth is 4.6pA/√Hz with 0.68nA rms . Input referred noise is 3.38pA/√Hz. These results are obtained by simulating the proposed approach using cadence 45nm technology. The simulated results shows that the proposed HGCTIA is superior in terms of power, gain and bandwidth.
... Unfortunately, the amplitude of some of these artifacts can be large as compared to the actual EEG signal originating from brain. In addition, processing of high-resolution long term EEG [15]. Whatever the artifacts are, they are superimposed on the background activities. ...
... Human brain layers and their electrical model with a wet-EEG electrode, (a) equivalent electrical model of an EEG electrode on the human head, (b) EEG electrode frequency response[15]. ...
... 15 shows the amplitude-frequency characteristic of EEG, electrocardiography (ECG) and electromyography (EMG) as recorded by wet surface electrodes. It shows that the EEG signals are corrupted by 1/f noise (generated by CMOS components) and 50/60Hz interference generated by power lines. ...
Epilepsy is one of the most common neurological diseases which has approximately one percent population of the global community. Although there are some drugs to reduce severe seizures and surgery methods to prevent them, the origin of this chronic disease is not known clearly. Researchers mainly focus on electrical response of the brain in electroencephalography (EEG) signal to investigate how electrochemistry change at the brain cell membrane is affecting the appearance of an epileptic condition. On the other side, some pioneering research on metabolic changes in the cerebral are trying to clarification what exactly happens on a cellular level. It is principally done by the functional magnetic resonance imaging (fMRI) system, but for the purpose of investigation into the cerebral cortex, the Near-infrared spectroscopy (NIRs) system is the best alternative technique. It has not only better temporal resolution compare to the fMRI machine, but the NIRs equipment is also portable which provides a wide range of motion. The aim of the present thesis was to determine new quantitative diagnostic criteria for patients with epileptic seizures, capable of differentiating among pre-ictal activities, inter-ictal abnormal activities and variations in ictal zone using NIRs signal. In addition, using selected features to achieve a seizure prediction system. Twelve patients with parietal lobe epilepsies in volunteers were recruited. The interest region was shaved and NIRs-EEG probe was placed individually for each subject. Subsequently, they were asked to sleep during night in the special dark room of the sleep clinic and signals were recorded continuously. Ten onset/offset seizures for each patient were labeled by a neurologist specialist according to EEG signal. Pre-processed NIRs signals were mapped from the time-domain to the time-frequency domain through discrete wavelet transform (DWT) method. Non-linear dynamical features based on higher-order statistics/spectra (HOS) as well as several widely-used conventional second-order measures and Shannon entropy were employed in DWT coefficients at every level. The results demonstrated the potential of low frequency oxy-hemoglobin and deoxy-hemoglobin changes in DWT approximation components-based features to act as specific marker for pre-ictal region outperforming the conventional approaches. Additionally, detail components-based indices, especially, entropy and skewness provide some intuition about the relationship between cortical hemodynamic changes and EEG spikes that corresponding to inter-ictal time. Finally, to find the most accurate seizure prediction model, selected features were applied to the three different supervised machine learning classifiers. These experiments showed that support vector machine (SVM) with radial basis function (RBF) kernel had the best performance. In conclusion, this study illuminated that there is significant information about different stages of seizures even in signals obtained from the simple setup NIRs system having the capability to be individually employed as a seizure prediction device.
... 28 Nevertheless, several studies reported working fNIRS instruments with SNR in the range of 40 dB. 24,29,30 Drift in the raw measurements could lead to undesired concentration divergence of O 2 Hb and HHb without physiological origin and should consequently be minimized. A sampling frequency above 6 Hz 3 is important for accurately resolving the time course of the physiological signals (typically in the range of 0.1 to 1.2 Hz) and decreasing the influence of electronic noise. ...
With the aim of transitioning functional near-infrared spectroscopy (fNIRS) technology from the laboratory environment to everyday applications, the field has seen a recent push toward the development of wearable/miniaturized, multiwavelength, multidistance, and modular instruments. However, it is challenging to unite all these requirements in a precision instrument with low noise, low drift, and fast sampling characteristics. We present the concept and development of a wearable fNIRS instrument that combines all these key features with the goal of reliably and accurately capturing brain hemodynamics. The proposed instrument consists of a modular network of miniaturized optode modules that include a four-wavelength light source and a highly sensitive silicon photomultiplier detector. Simultaneous measurements with short-separation (7.5 mm; containing predominantly extracerebral signals) and long-separation (20 mm or more; containing both extracerebral and cerebral information) channels are used with short-channel regression filtering methods to increase robustness of fNIRS measurements. Performance of the instrument was characterized with phantom measurements and further validated in human in vivo measurements, demonstrating the good raw signal quality (signal-to-noise ratio of 64 dB for short channels; robust measurements up to 50 mm; dynamic optical range larger than 160 dB), the valid estimation of concentration changes (oxy- and deoxyhemoglobin, and cytochrome-c-oxidase) in muscle and brain, and the detection of task-evoked brain activity. The results of our preliminary tests suggest that the presented fNIRS instrument outperforms existing instruments in many aspects and bears high potential for real-time single-trial fNIRS applications as required for wearable brain-computer interfaces.
... Since the PPG signal consists of an ac signal and a dc component related to physiological conditions of individuals. To avoid degradation of dynamic range, dc component should be rejected; however, the pulse width modulation (PWM) technique is usually used in light sources to reduce power consumption and extend lifetime of LEDs, thus the dc rejection problem is more complex to solve (Normandin et al. 2005). Another issue for PPG front-end readout is the realization of mismatch current source caused by CMOS process. ...
... The specific design challenges for TIA are how to successfully amplify the PD to large dynamic ranges, while keeping noise low and maintaining required response speed, i.e., a large-enough bandwidth, for processing sensed data. Many TIA circuits are reported in recent years (Razavi 2003;Chien et al. 2010;Wong et al. 2008;Normandin et al. 2005;Razavi 2000), such as common-gate TIA, resistive feedback TIA, differential TIA and capacitive feedback TIA. Common-gate TIAes (CG-TIA) employ open-loop topology with a low input impedance and high transimpedance gain. ...
This study presents a novel readout circuit that serves to evaluate quantity of skin melanin in a noninvasive way via utilizing a photoplethysmography (PPG) optical sensor. PPG sensors are widely used nowadays for noninvasive diagnosis due to their salient features such as low cost and easy-to-useness. PPG signal is known highly correlated with human skin condition and cardiovascular system health. A new readout circuit featuring high-bandwidth, low-noise and low-power is designed in this study to evaluate quantitatively the melanin content of human skin based on PPG signals with high accuracy. This new readout circuit is intended to convert and amplify the electric current of the photo diode in the PPG sensor to voltage signals, and then being further properly amplified to the full dynamic range of the chip circuit fabricated by the TSMC 0.18 μm process for maximum S/N ratios. The designed circuit has a front-end transimpedance amplifier (TIA), 60 dB in gain, with a degenerate resistor to perform a low noise less than 3 µV/√Hz while operating by a 3.3 V supply. In related circuit designs, the mismatch in current sources caused by CMOS process variation is mitigated by a current steering digital-analog-converter (DAC) with least significant bit (LSB) corresponding to 1.25 µA. Both TIA and DAC are designed with effort to render high accuracy in the digital outputs for the levels of sensed melanin. The power consumption of chip design is as low as 1.5 mW, while the error rate between measurement and simulation result is less than 3 %.
... Standard deviation and average values of the DC current consumption, bandwidth and gain are shown in Fig.8. [10][11][12][13]. It's clear that the proposed TIA achieves the highest FoM and the lowest power consumption. ...
... The TIA introduced in [11] offers high GBW at the expense of higher input referred noise. The design introduced in [12] operates at large input capacitance with a better sensitivity but the bandwidth is limited to 40kHz.The TIA introduced in [13] achieves a better gain and lower noise, however, its power consumption is much higher than our proposed design. ...
... The main design parameters of SiAPD are area, depletion region thickness, SNR and fabrication technology and architecture. High-level noise and low-efficiency of the currently available detectors [114], [115], encouraged us to design a new CMOS SiAPD with low-noise and high efficiency characteristics at NIR region. ...
... The amount of reflected NIR light photons detected by the photodetector depends on the power of the NIR light source, the attenuation due to biological tissue, and the light source-to-photodetector distance. The level of attenuation with a light source-tophotodetector distance of 4 cm, for a five-layer (scalp, skull, cerebrospinal fluid (CSF) layer, and gray and white matters) head model, can be approximated by 4.12 9 10 -4 cm -2 [20]. If we consider an NIR light source of 10 mW, the optical power seen by the photodetector is only 4.12 lW/cm 2 . ...
... Earlier silicon p-i-n photodiodes were successfully used for NIRS systems with an active area of *7.5 mm 2 [7,20]. Silicon avalanche photodiodes are also commercially available with comparable active area and are being used in several NIRS systems, such as, Hamamatsu C5460-01 device has a large active area of 7 mm 2 [5,6]. ...
... As such, the performance of SiAPD is degraded by a factor known as excess noise factor, F. Both excess noise and photocurrent of SiAPD increase as the avalanche gain increase, so the best SNR occurs at a certain gain. A SNR of *40 dB is needed for NIRS application [20]. Silicon avalanche photodiodes with dark current in nA range, and the generated photocurrent in hundreds of lA range confirms SNR of much higher than 40 dB. ...
This paper surveys recent research on CMOS silicon avalanche photodiodes (SiAPD) and presents the design of a SiAPD based
photoreceiver dedicated to near-infrared spectroscopy (NIRS) application. Near-infrared spectroscopy provides an inexpensive,
non-invasive, and portable means to image brain function, and is one of the most efficient diagnostic techniques of different
neurological diseases. In NIRS system, brain tissue is penetrated by near-infrared (NIR) radiation and the reflected signal
is captured by a photodiode. Since the reflected NIR signal has very low amplitude, SiAPD is a better choice than regular
photodiode for NIR signal detection due to SiAPD`s ability to amplify the photo generated signal by avalanche multiplication.
Design requirements of using CMOS SiAPDs for NIR light detection are discussed, and the challenges of fabricating SiAPDs using
standard CMOS process are addressed. Performances of state-of-the-art CMOS SiAPDs with different device structures are summarized
and compared. The efficacy of the proposed SiAPD based photoreceiver is confirmed by post layout simulation. Finally, the
SiAPD and its associated circuits has been implemented in one chip using 0.35 μm standard CMOS technology for an integrated
NIRS system.
... Micro-LED arrays, silicon probes with polymer light guides, and optorodes have been reportedly used for optical stimulation [21][22][23][24]. Despite growing demands, there have been few reports of the development of contact CMOS imagers that are designed for neuroscience applications [25,26]. To meet the need for compact neuroimaging tools using CMOS technologies, we have developed several types of CMOS image sensors. ...
We developed a complementary metal oxide semiconductor (CMOS) integrated device for optogenetic applications. This device can interface via neuronal tissue with three functional modalities: imaging, optical stimulation and electrical recording. The CMOS image sensor was fabricated on 0.35 μm standard CMOS process with built-in control circuits for an on-chip blue light-emitting diode (LED) array. The effective imaging area was 2.0 × 1.8 mm². The pixel array was composed of 7.5 × 7.5 μm² 3-transistor active pixel sensors (APSs). The LED array had 10 × 8 micro-LEDs measuring 192 × 225 μm². We integrated the device with a commercial multichannel recording system to make electrical recordings.
... In the past, Baker and Sarpeshkar and Ferrari et al. had developed sense amplifiers for biomedical applications, but the DC input range was either limited to one point [11] or too small [12], and the circuit also consumed huge power [12]. An AC coupling circuit using diode-connected transistors had also been implemented [13] to realize large impedance; however, it is sensitive to process variations. ...
This brief describes a fully integrated dual-loop transimpedance amplifier with bandpass response for wearable near-infrared (NIR) sensing operating at low frequency: one loop to lower the lower band cutoff frequency and another loop for self-regulating DC photocurrent to prevent saturation at later stages. The circuit was implemented in a 0.35- $muhbox{m}$ CMOS process and achieved a DC photocurrent rejection ranging from 2.7 to 15 $muhbox{A}$ and a $-$ 3-dB high-pass cutoff frequency from 0.5 to 110 Hz by using on-chip capacitors. Both total harmonic distortion and spurious-free dynamic range are better than $-$ 42 dB. It achieves an improvement of 390 times in capacitor reduction compared with the traditional DC rejection technique.
... As mentioned in previous section, there is a need for the development of a current-to-voltage converter that can regulate voltage output coming from a pulsed current source [3], [4]. A TIA contained in a feedback loop with error amplifier is necessary , such as the one shown in [13]. ...
A micro-power CMOS front-end, consisting of a transimpedance amplifier (TIA) and an ultralow cutoff frequency lowpass filter for the acquisition of photoplethysmographic signal (PPG) is presented. Robust DC photocurrent rejection for the pulsed signal source is achieved through a sample-and-hold stage in the feed-forward signal path and an error amplifier in the feedback path. Ultra-low cutoff frequency of the filter is achieved with a proposed technique that incorporates a pair of current-steering transistors that increases the effective filter capacitance. The design was realized in a 0.35-mum CMOS technology. It consumes 600 muW at 2.5 V, rejects DC photocurrent ranged from 100 nA to 53.6 muA, and achieves lower-band and upper-band - 3-dB cutoff frequencies of 0.46 and 2.8 Hz, respectively.
