As technology evolves, more components are integrated into printed circuit boards (PCBs) and the PCB layout increases. Because small defects on signal trace can cause significant damage to the system, PCB surface inspection is one of the most important quality control processes. Owing to the limitations of manual inspection, significant efforts have been made to automate the inspection by utilizing high resolution CCD or CMOS sensors. Despite the advanced sensor technology, setting the pass/fail criteria based on small failure samples has always been challenging in traditional machine vision approaches. To overcome these problems, we propose an advanced PCB inspection system based on a skip-connected convolutional autoencoder. The deep autoencoder model was trained to decode the original non-defect images from the defect images. The decoded images were then compared with the input image to identify the defect location. To overcome the small and imbalanced dataset in the early manufacturing stage, we applied appropriate image augmentation to improve the model training performance. The experimental results reveal that a simple unsupervised autoencoder model delivers promising performance, with a detection rate of up to 98% and a false pass rate below 1.7% for the test data, containing 3900 defect and non-defect images.

Diabetes has recently emerged as a worldwide problem, and diabetic retinopathy is an abnormal state associated with the human retina. Due to the increase in daily screen-related activities of modern human beings, diabetic retinopathy is more prevalent among adults, leading to minor and major blindness. Doctors and clinicians are unable to perform early diagnoses due to the large number of patients. To solve this problem, this study introduces a classification model for retinal images that distinguishes between the various stages of diabetic retinopathy. This work involves deploying deep and densely connected networks for retinal image analysis with training from scratch. Dense connections between the convolutional layers of the network are an essential factor to enhance accuracy owing to the deeper supervision between layers. Another factor is the growth rate that further assists our model in learning more sophisticated feature maps regarding retinal images from every stage of the network. We compute the area under the curve, sensitivity, and specificity, particularly for messidor-2 and EyePACS. Compared to existing approaches, our method achieved better results, with an approximate rise rate of 0.01, 0.03, and 0.01, respectively. Therefore, computer-aided programs can help in diagnostic centers as automated detection systems.

This paper presents a novel amplifier that satisfies both low distortion and high efficiency for high-frequency wireless ultrasound systems with limited battery life and size. While increasing the amplifier efficiency helps to address the problems for wireless ultrasound systems, it can cause signal distortion owing to harmonic components. Therefore, a new type of class F amplifier is designed to achieve high efficiency and low distortion. In the amplifier, the resonant circuit at each stage controls the harmonic components to reduce distortion and improve efficiency. Transformers with a large shunt resistor are also helpful to reduce the remaining noise in the input signal. The proposed class F amplifier is tested using simulations, and the voltage and current waveforms are analyzed to achieve correct operation with adequate efficiency and distortion. The measured performance of the class F amplifier has a gain of 23.2 dB and a power added efficiency (PAE) of 88.9% at 25 MHz. The measured DC current is 121 mA with a variance of less than 1% when the PA is operating. We measured the received echo signal through the pulse-echo response using a 25-MHz transducer owing to the compatibility of the designed class F amplifier with high- frequency transducers. The measured total harmonic distortion (THD) of the echo signal was obtained as 4.5% with a slightly low ring-down. The results show that the low THD and high PAE of the new high-efficiency and high-voltage amplifier may increase battery life and reduce the cooling fan size, thus providing a suitable environment for high-frequency wireless ultrasound systems.

This paper describes the design of a front-end circuit consisting of an integrated preamplifier with a Sallen-Key Butterworth filter for very-high-frequency ultrasonic transducers and a low-power handheld receiver. This preamplifier was fabricated using a 0.18-μm 7WL SiGe bi-polar complementary metal oxide semiconductor (BiCMOS) process. The Sallen-Key filter is used to increase the voltage gain of the front-end circuit for high-frequency transducers which are generally low in sensitivity. The measured peak voltage gain of the frontend circuits for the BiCMOS preamplifier with the Sallen-Key filter was 41.28 dB at 100 MHz with a-6-dB bandwidth of 91%, and the dc power consumption of the BiCMOS preamplifier was 49.53 mW. The peak voltage gain of the front-end circuits for the CMOS preamplifier with the Sallen-Key filter was 39.52 dB at 100 MHz with a-6-dB bandwidth of 108%, and the dc power consumption of the CMOS preamplifier was 43.57 mW. Pulse-echo responses and wire phantom images with a single-element ultrasonic transducer have been acquired to demonstrate the performance of the front-end circuit.