In this study, we propose a novel deep learning-based method to predict an optimized structure for a given boundary condition and optimization setting without using any iterative scheme. For this purpose,first, using open-source topology optimization code, datasets of the optimized structures paired with the corresponding information on boundary conditions and optimization settings are generated at low (32  32) and high (128  128) resolutions. To construct the artificial neural network for the proposed method, a convolutional neural network (CNN)-based encoder and decoder network is trained using the training dataset generated at low resolution. Then, as a two-stage refinement, the conditional generative adversarial network (cGAN) is trained with the optimized structures paired at both low and high resolutions, and is connected to the trained CNN-based encoder and decoder network. The performance evaluation results of the integrated network demonstrate that the proposed method can determine a nearoptimal structure in terms of pixel values and compliance with negligible computational time.

Double Patterning lithography is very fascinating way of lithography which is capable of pushing down the k1 limit below 0.25. By using double patterning lithography, we can delineate the pattern beyond resolution capability. Target pattern is decomposed into patterns within resolution capability and decomposed patterns are combined together through twice lithography and twice etch processes. Two ways, negative and positive, of doing double patterning process are contrived and studied experimentally. In this paper, various issues in double patterning lithography such as pattern decomposition, resist process on patterned topography, process window of 1/4 pitch patterning, and overlay dependent CD variation are studied on positive and negative tone double patterning respectively. Among various issues about double patterning, only the overlay controllability and productivity seemed to be dominated as visible obstacles so far.

The dry etching mechanism of lead zirconate titanate (PZT) films was studied in high density CF4 and Cl2/CF4 inductively coupled plasmas. The concentrations of atomic Cl and F as well as the flux and energy of bombarding ions were monitored as a function of etching parameters such as etching gas ratio, substrate bias voltage (Vs), induction coil power and process pressure. The compositions and chemical bonding states of etched PZT films were examined by X-ray photoelectron spectroscopy (XPS). The etching of PZT films in CF4-based plasma is chemically assisted sputter etching, and the dominant step of the overall etching process is either the formation or the removal of the etch by-products, depending on the etching conditions. The etching of PZT films in Cl2/CF4 mixed plasma is mainly dominated by the formation of metal chlorides which depends on the concentration of the atomic Cl and the bombarding ion energy. The PZT film shows a maximum etch rate in 90% Cl2/(Cl2+CF4) plasma where the concentration of atomic Cl is maximum. The etch selectivity of PZT to Pt is less than 1.3 in CF4-based plasma, where as more than 2 in Cl2/CF4 mixed plasma. The amount of sidewall residue is greatly reduced in Cl2/CF4 mixed plasma compared with in CF4 plasma. A more vertical etch profile of PZT films can be obtained by lowering the process pressure and increasing the substrate bias voltage.

Idiopathic infantile nystagmus (IIN) is the involuntary oscillation of the eyes with onset in the first few months of life. The most common form of inheritance is X-linked, and mutations in FRMD7 gene are a major cause. To identify the FRMD7 gene mutations associated with X-linked IIN, we performed PCR-based DNA direct sequencing in 4 affected subjects from 2 Korean families. We also assessed structural abnormalities of retina and optic nerve head using optical coherence tomography (OCT). Genetic analysis revealed a A>G transversion at nucleotide c.1, the first base of the start codon. This mutation leads to the loss of the primary start codon ATG for methionine, which is replaced by a triplet GTG for valine. The alternative in-frame start codon is not present around a mutation. OCT revealed the morphological changes within the optic nerve head, including shallow cup depth and small cup-to-disc ratio. In summary, we identified a novel start codon mutation within the FRMD7 gene of 2 Korean families. Our data expands the mutation spectrum of FRMD7 causing IIN. We also demonstrated abnormal developments of afferent system in patients with FRMD7 mutations using OCT, which may help to understand the etiological factor in development of nystagmus.

A comparative study assessing the effect of interface type on procedural skill transfer during virtual training is presented. The aim of this research is to evaluate the transferability of two aspects of procedural skills, that is, procedural knowledge and technical skills. We established one group with a lecture and three virtual training groups with a combination of output and input devices: a monitor and keyboard/mouse, a head-mounted display (HMD) and joypad, and an HMD and wearable sensors. The task for assessment was a lifeboat launching operation that requires a participant to memorize a 10-step procedure utilizing 14 different pieces of equipment that should be manipulated in each step. Before and after training, we evaluated the participants' procedural knowledge and technical skill on a real lifeboat. The monitor and keyboard/mouse group showed the best performance in a procedural knowledge assessment that addressed visually induced recollections from the real lifeboat. Alternatively, in the assessment of technical skills that determined manipulation ability that requires word-based mnemonics, the HMD and wearable sensors group outperformed the other groups. Moreover, the results showed that the virtual training was a more efficient training format for short-term training than a lecture due to the freedom of observation viewpoint, despite simulator sickness.