Cross-site validation of lung cancer diagnosis by electronic nose with deep learning: a multicenter prospective study.

BACKGROUND: Although electronic nose (eNose) has been intensively investigated for diagnosing lung cancer, cross-site validation remains a major obstacle to be overcome and no studies have yet been performed. METHODS: Patients with lung cancer, as well as healthy control and diseased control groups, were prospectively recruited from two referral centers between 2019 and 2022. Deep learning models for detecting lung cancer with eNose breathprint were developed using training cohort from one site and then tested on cohort from the other site. Semi-Supervised Domain-Generalized (Semi-DG) Augmentation (SDA) and Noise-Shift Augmentation (NSA) methods with or without fine-tuning was applied to improve performance. RESULTS: In this study, 231 participants were enrolled, comprising a training/validation cohort of 168 individuals (90 with lung cancer, 16 healthy controls, and 62 diseased controls) and a test cohort of 63 individuals (28 with lung cancer, 10 healthy controls, and 25 diseased controls). The model has satisfactory results in the validation cohort from the same hospital while directly applying the trained model to the test cohort yielded suboptimal results (AUC, 0.61, 95% CI: 0.47─0.76). The performance improved after applying data augmentation methods in the training cohort (SDA, AUC: 0.89 [0.81─0.97]; NSA, AUC:0.90 [0.89─1.00]). Additionally, after applying fine-tuning methods, the performance further improved (SDA plus fine-tuning, AUC:0.95 [0.89─1.00]; NSA plus fine-tuning, AUC:0.95 [0.90─1.00]). CONCLUSION: Our study revealed that deep learning models developed for eNose breathprint can achieve cross-site validation with data augmentation and fine-tuning. Accordingly, eNose breathprints emerge as a convenient, non-invasive, and potentially generalizable solution for lung cancer detection. CLINICAL TRIAL REGISTRATION: This study is not a clinical trial and was therefore not registered.

Development of a Dual MOS Electronic Nose/Camera System for Improving Fruit Ripeness Classification.

Electronic nose (E-nose) systems have become popular in food and fruit quality evaluation because of their rapid and repeatable availability and robustness. In this paper, we propose an E-nose system that has potential as a non-destructive system for monitoring variation in the volatile organic compounds produced by fruit during the maturing process. In addition to the E-nose system, we also propose a camera system to monitor the peel color of fruit as another feature for identification. By incorporating E-nose and camera systems together, we propose a non-destructive solution for fruit maturity monitoring. The dual E-nose/camera system presents the best Fisher class separability measure and shows a perfect classification of the four maturity stages of a banana: Unripe, half-ripe, fully ripe, and overripe.

A Low Noise CMOS Readout Based on a Polymer-Coated SAW Array for Miniature Electronic Nose.

An electronic nose (E-Nose) is one of the applications for surface acoustic wave (SAW) sensors. In this paper, we present a low-noise complementary metal-oxide-semiconductor (CMOS) readout application-specific integrated circuit (ASIC) based on an SAW sensor array for achieving a miniature E-Nose. The center frequency of the SAW sensors was measured to be approximately 114 MHz. Because of interference between the sensors, we designed a low-noise CMOS frequency readout circuit to enable the SAW sensor to obtain frequency variation. The proposed circuit was fabricated in Taiwan Semiconductor Manufacturing Company (TSMC) 0.18 µm 1P6M CMOS process technology. The total chip size was nearly 1203 × 1203 µm². The chip was operated at a supply voltage of 1 V for a digital circuit and 1.8 V for an analog circuit. The least measurable difference between frequencies was 4 Hz. The detection limit of the system, when estimated using methanol and ethanol, was 0.1 ppm. Their linearity was in the range of 0.1 to 26,000 ppm. The power consumption levels of the analog and digital circuits were 1.742 mW and 761 µW, respectively.

A bio-inspired two-layer sensing structure of polypeptide and multiple-walled carbon nanotube to sense small molecular gases.

In this paper, we propose a bio-inspired, two-layer, multiple-walled carbon nanotube (MWCNT)-polypeptide composite sensing device. The MWCNT serves as a responsive and conductive layer, and the nonselective polypeptide (40 mer) coating the top of the MWCNT acts as a filter into which small molecular gases pass. Instead of using selective peptides to sense specific odorants, we propose using nonselective, peptide-based sensors to monitor various types of volatile organic compounds. In this study, depending on gas interaction and molecular sizes, the randomly selected polypeptide enabled the recognition of certain polar volatile chemical vapors, such as amines, and the improved discernment of low-concentration gases. The results of our investigation demonstrated that the polypeptide-coated sensors can detect ammonia at a level of several hundred ppm and barely responded to triethylamine.

A miniature electronic nose system based on an MWNT-polymer microsensor array and a low-power signal-processing chip.

This article introduces a power-efficient, miniature electronic nose (e-nose) system. The e-nose system primarily comprises two self-developed chips, a multiple-walled carbon nanotube (MWNT)-polymer based microsensor array, and a low-power signal-processing chip. The microsensor array was fabricated on a silicon wafer by using standard photolithography technology. The microsensor array comprised eight interdigitated electrodes surrounded by SU-8 "walls," which restrained the material-solvent liquid in a defined area of 650 × 760 µm(2). To achieve a reliable sensor-manufacturing process, we used a two-layer deposition method, coating the MWNTs and polymer film as the first and second layers, respectively. The low-power signal-processing chip included array data acquisition circuits and a signal-processing core. The MWNT-polymer microsensor array can directly connect with array data acquisition circuits, which comprise sensor interface circuitry and an analog-to-digital converter; the signal-processing core consists of memory and a microprocessor. The core executes the program, classifying the odor data received from the array data acquisition circuits. The low-power signal-processing chip was designed and fabricated using the Taiwan Semiconductor Manufacturing Company 0.18-µm 1P6M standard complementary metal oxide semiconductor process. The chip consumes only 1.05 mW of power at supply voltages of 1 and 1.8 V for the array data acquisition circuits and the signal-processing core, respectively. The miniature e-nose system, which used a microsensor array, a low-power signal-processing chip, and an embedded k-nearest-neighbor-based pattern recognition algorithm, was developed as a prototype that successfully recognized the complex odors of tincture, sorghum wine, sake, whisky, and vodka.

Fusion of memristor and digital compute-in-memory processing for energy-efficient edge computing.

Artificial intelligence (AI) edge devices prefer employing high-capacity nonvolatile compute-in-memory (CIM) to achieve high energy efficiency and rapid wakeup-to-response with sufficient accuracy. Most previous works are based on either memristor-based CIMs, which suffer from accuracy loss and do not support training as a result of limited endurance, or digital static random-access memory (SRAM)-based CIMs, which suffer from large area requirements and volatile storage. We report an AI edge processor that uses a memristor-SRAM CIM-fusion scheme to simultaneously exploit the high accuracy of the digital SRAM CIM and the high energy-efficiency and storage density of the resistive random-access memory memristor CIM. This also enables adaptive local training to accommodate personalized characterization and user environment. The fusion processor achieved high CIM capacity, short wakeup-to-response latency (392 microseconds), high peak energy efficiency (77.64 teraoperations per second per watt), and robust accuracy (<0.5% accuracy loss). This work demonstrates that memristor technology has moved beyond in-lab development stages and now has manufacturability for AI edge processors.

Towards a chemiresistive sensor-integrated electronic nose: a review.

Electronic noses have potential applications in daily life, but are restricted by their bulky size and high price. This review focuses on the use of chemiresistive gas sensors, metal-oxide semiconductor gas sensors and conductive polymer gas sensors in an electronic nose for system integration to reduce size and cost. The review covers the system design considerations and the complementary metal-oxide-semiconductor integrated technology for a chemiresistive gas sensor electronic nose, including the integrated sensor array, its readout interface, and pattern recognition hardware. In addition, the state-of-the-art technology integrated in the electronic nose is also presented, such as the sensing front-end chip, electronic nose signal processing chip, and the electronic nose system-on-chip.

A 4.5 µW Miniaturized 3-Channel Wireless Intra-Cardiac Acquisition System.

This article presents a chip designed for wireless intra-cardiac monitoring systems. The design consists of a three-channel analog front-end, a pulse-width modulator featuring output-frequency offset and temperature calibration, and inductive data telemetry. By employing a resistance boosting technique in the instrumentation amplifier feedback, the pseudo-resistor exhibits lower non-linearity, leading to a total harmonic distortion of below 0.1%. Furthermore, the boosting technique enhances the feedback resistance, leading to a reduction in the size of the feedback capacitor and, consequently, the overall size. To make the modulator's output frequency resilient to temperature and process changes, coarse and fine-tuning algorithms are used. The front-end channel is capable of extracting the intra-cardiac signal with an effective number of bits of 8.9, while exhibiting an input-referred noise of less than 2.7 µVrms, and consuming 200 nW per channel. The front-end output is encoded by an ASK-PWM modulator, which drives an on-chip transmitter at 13.56 MHz. The proposed System-on-Chip (SoC) is fabricated in a 0.18 µm standard CMOS technology and consumes 4.5 µW while occupying 1.125 mm2.

A 5.3 pJ/Spike CMOS Neural Array Employing Time-Modulated Axon-Sharing and Background Mismatch Calibration Techniques.

Inspired by the human brain, spiking neuron networks are promising to realize energy-efficient and low-latency neuromorphic computing. However, even state-of-the-art silicon neurons are orders of magnitude worse than biological neurons in terms of area and power consumption due to the limitations. Moreover, limited routing in typical CMOS processes is another challenge for realizing the fully-parallel high-throughput synapse connections compared to biological synapses. This paper presents an SNN circuit that utilizes resource-sharing techniques to address the two challenges. Firstly, a comparator sharing neuron circuit with a background calibration technique is proposed to shrink the size of a single neuron without performance degradation. Secondly, a time-modulated axon-sharing synapse system is proposed to realize a fully-parallel connection with limited hardware overhead. To validate the proposed approaches, a CMOS neuron array is designed and fabricated under a 55-nm process. It consists of 48 LIF neurons with 3125 neurons/mm 2 area density, power consumption of 5.3 pJ/spike, and equivalent 2304 fully parallel synapses providing a unit throughput of 5500 events/s/neuron. It proves the proposed approaches are promising to realize a high-throughput high-efficiency SNN with CMOS technology.

Polymer/Ordered mesoporous carbon nanocomposite platelets as superior sensing materials for gas detection with surface acoustic wave devices.

We have prepared nanocomposites of polymers and platelet CMK-5-like carbon and have demonstrated their superior performance for gravimetric gas detection. The zirconium-containing platelet SBA-15 was used as hard template to prepare CMK-5-like carbon, which was then applied as a lightweight and high-surface-area scaffold for the growth of polymers by radical polymerization. Mesoporous nanocomposites composed of four different polymers were used as sensing materials for surface acoustic wave devices to detect ppm-level ammonia gas. The sensors showed much better sensitivity and reversibility than those coated with dense polymer films, and the sensor array could still generate a characteristic pattern for the analyte with a concentration of 16 ppm. The results show that the nanocomposite sensing materials are promising for highly sensitive gravimetric-type electronic nose applications.

An analog multilayer perceptron neural network for a portable electronic nose.

This study examines an analog circuit comprising a multilayer perceptron neural network (MLPNN). This study proposes a low-power and small-area analog MLP circuit to implement in an E-nose as a classifier, such that the E-nose would be relatively small, power-efficient, and portable. The analog MLP circuit had only four input neurons, four hidden neurons, and one output neuron. The circuit was designed and fabricated using a 0.18 µm standard CMOS process with a 1.8 V supply. The power consumption was 0.553 mW, and the area was approximately 1.36 × 1.36 mm2. The chip measurements showed that this MLPNN successfully identified the fruit odors of bananas, lemons, and lychees with 91.7% accuracy.

A Review of Emerging Electromagnetic-Acoustic Sensing Techniques for Healthcare Monitoring.

Conventional electromagnetic (EM) sensing techniques such as radar and LiDAR are widely used for remote sensing, vehicle applications, weather monitoring, and clinical monitoring. Acoustic techniques such as sonar and ultrasound sensors are also used for consumer applications, such as ranging and in vivo medical/healthcare applications. It has been of long-term interest to doctors and clinical practitioners to realize continuous healthcare monitoring in hospitals and/or homes. Physiological and biopotential signals in real-time serve as important health indicators to predict and prevent serious illness. Emerging electromagnetic-acoustic (EMA) sensing techniques synergistically combine the merits of EM sensing with acoustic imaging to achieve comprehensive detection of physiological and biopotential signals. Further, EMA enables complementary fusion sensing for challenging healthcare settings, such as real-world long-term monitoring of treatment effects at home or in remote environments. This article reviews various examples of EMA sensing instruments, including implementation, performance, and application from the perspectives of circuits to systems. The novel and significant applications to healthcare are discussed. Three types of EMA sensors are presented: (1) Chip-based radar sensors for health status monitoring, (2) Thermo-acoustic sensing instruments for biomedical applications, and (3) Photoacoustic (PA) sensing and imaging systems, including dedicated reconstruction algorithms were reviewed from time-domain, frequency-domain, time-reversal, and model-based solutions. The future of EMA techniques for continuous healthcare with enhanced accuracy supported by artificial intelligence (AI) is also presented.

An electronic-nose sensor node based on a polymer-coated surface acoustic wave array for wireless sensor network applications.

This study developed an electronic-nose sensor node based on a polymer-coated surface acoustic wave (SAW) sensor array. The sensor node comprised an SAW sensor array, a frequency readout circuit, and an Octopus II wireless module. The sensor array was fabricated on a large K(2) 128° YX LiNbO3 sensing substrate. On the surface of this substrate, an interdigital transducer (IDT) was produced with a Cr/Au film as its metallic structure. A mixed-mode frequency readout application specific integrated circuit (ASIC) was fabricated using a TSMC 0.18 µm process. The ASIC output was connected to a wireless module to transmit sensor data to a base station for data storage and analysis. This sensor node is applicable for wireless sensor network (WSN) applications.

A single-walled carbon nanotube network gas sensing device.

The goal of this research was to develop a chemical gas sensing device based on single-walled carbon nanotube (SWCNT) networks. The SWCNT networks are synthesized on Al(2)O(3)-deposted SiO(2)/Si substrates with 10 nm-thick Fe as the catalyst precursor layer using microwave plasma chemical vapor deposition (MPCVD). The development of interconnected SWCNT networks can be exploited to recognize the identities of different chemical gases by the strength of their particular surface adsorptive and desorptive responses to various types of chemical vapors. The physical responses on the surface of the SWCNT networks cause superficial changes in the electric charge that can be converted into electronic signals for identification. In this study, we tested NO(2) and NH(3) vapors at ppm levels at room temperature with our self-made gas sensing device, which was able to obtain responses to sensitivity changes with a concentration of 10 ppm for NO(2) and 24 ppm for NH(3).

A local weighted nearest neighbor algorithm and a weighted and constrained least-squared method for mixed odor analysis by electronic nose systems.

A great deal of work has been done to develop techniques for odor analysis by electronic nose systems. These analyses mostly focus on identifying a particular odor by comparing with a known odor dataset. However, in many situations, it would be more practical if each individual odorant could be determined directly. This paper proposes two methods for such odor components analysis for electronic nose systems. First, a K-nearest neighbor (KNN)-based local weighted nearest neighbor (LWNN) algorithm is proposed to determine the components of an odor. According to the component analysis, the odor training data is firstly categorized into several groups, each of which is represented by its centroid. The examined odor is then classified as the class of the nearest centroid. The distance between the examined odor and the centroid is calculated based on a weighting scheme, which captures the local structure of each predefined group. To further determine the concentration of each component, odor models are built by regressions. Then, a weighted and constrained least-squares (WCLS) method is proposed to estimate the component concentrations. Experiments were carried out to assess the effectiveness of the proposed methods. The LWNN algorithm is able to classify mixed odors with different mixing ratios, while the WCLS method can provide good estimates on component concentrations.

Development of a portable electronic nose system for the detection and classification of fruity odors.

In this study, we have developed a prototype of a portable electronic nose (E-Nose) comprising a sensor array of eight commercially available sensors, a data acquisition interface PCB, and a microprocessor. Verification software was developed to verify system functions. Experimental results indicate that the proposed system prototype is able to identify the fragrance of three fruits, namely lemon, banana, and litchi.

A Gas Mixture Prediction Model Based on the Dynamic Response of a Metal-Oxide Sensor.

Metal-oxide (MOX) gas sensors are widely used for gas concentration estimation and gas identification due to their low cost, high sensitivity, and stability. However, MOX sensors have low selectivity to different gases, which leads to the problem of classification for mixtures and pure gases. In this study, a square wave was applied as the heater waveform to generate a dynamic response on the sensor. The information of the dynamic response, which includes different characteristics for different gases due to temperature changes, enhanced the selectivity of the MOX sensor. Moreover, a polynomial interaction term mixture model with a dynamic response is proposed to predict the concentration of the binary mixtures and pure gases. The proposed method improved the classification accuracy to 100%. Moreover, the relative error of quantification decreased to 1.4% for pure gases and 13.0% for mixtures.

Detection of Cigarette Smoke Using a Surface-Acoustic-Wave Gas Sensor with Non-Polymer-Based Oxidized Hollow Mesoporous Carbon Nanospheres.

The objective of this research was to develop a surface-acoustic-wave (SAW) sensor of cigarette smoke to prevent tobacco hazards and to detect cigarette smoke in real time through the adsorption of an ambient tobacco marker. The SAW sensor was coated with oxidized hollow mesoporous carbon nanospheres (O-HMC) as a sensing material of a new type, which replaced a polymer. O-HMC were fabricated using nitric acid to form carboxyl groups on carbon frameworks. The modified conditions of O-HMC were analyzed with Scanning Electron Microscopy (SEM), Fourier transform infrared spectrometry (FTIR), and X-ray diffraction (XRD). The appropriately modified O-HMC are more sensitive than polyacrylic acid and hollow mesoporous carbon nanospheres (PAA-HMC), which is proven by normalization. This increases the sensitivity of a standard tobacco marker (3-ethenylpyridine, 3-EP) from 37.8 to 51.2 Hz/ppm and prevents the drawbacks of a polymer-based sensing material. On filtering particles above 1 µm and using tar to prevent tar adhesion, the SAW sensor detects cigarette smoke with sufficient sensitivity and satisfactory repeatability. Tests, showing satisfactory selectivity to the cigarette smoke marker (3-EP) with interfering gases CH4, CO, and CO2, show that CO and CO2 have a negligible role during the detection of cigarette smoke.

An Inductive Power and Data Telemetry Subsystem With Fast Transient Low Dropout Regulator for Biomedical Implants.

This paper presents a capacitorless low-dropout (LDO) regulator with fast transient response and data reverse telemetry circuit for fully implantable wireless transmission applications. We propose a novel hybrid feedback structure using high-frequency compensation technology to achieve a rapid transient response for the LDO regulator. To reduce the size of the implant and transmit neural recordings through the same coil without interfering with power transmission, the load-shift-key (LSK) modulation technique is adopted for back data telemetry. The proposed implantable chip, fabricated using commercial 0.18 µm complementary metal oxide semiconductor technology, yielded an output power of 15 mW. Under 1.15 V operation voltage, the maximum overshoot and undershoot voltages were less than 45 mV and 55 mV, respectively, for a 15 mA full-load current whose rising and falling time were less than 100 ns. The proposed LSK transceiver uses a digitized demodulator to improve bandwidth efficiency for low carrier frequency operation.