Diagnosing liver steatosis is an essential precaution for detecting hepatocirrhosis and liver cancer in the early stages. However, automatic diagnosis of liver steatosis from ultrasound (US) images remains challenging due to poor visual quality from various origins, such as speckle noise and blurring. In this paper, we propose a fully automated liver steatosis prediction model using three deep learning neural networks. As a result, liver steatosis can be automatically detected with high accuracy and precision. First, transfer learning is used for semantically segmenting the liver and kidney (L-K) on parasagittal US images, and then cropping the L-K area from the original US images. The second neural network also involves semantic segmentation by checking the presence of a ring that is typically located around the kidney and cropping of the L-K area from the original US images. These cropped L-K areas are inputted to the final neural network, SteatosisNet, in order to grade the severity of fatty liver disease. The experimental results demonstrate that the proposed model can predict fatty liver disease with the sensitivity of 99.78%, specificity of 100%, PPV of 100%, NPV of 99.83%, and diagnostic accuracy of 99.91%, which is comparable to the common results annotated by medical experts.
Despite the advancements in the diagnosis of early-stage cirrhosis, the accuracy in the diagnosis using ultrasound is still challenging owing to the presence of various image artifacts, which results in poor visual quality of the textural and lower-frequency components. In this study, we propose an end-to-end multistep network called CirrhosisNet that includes two transfer-learned convolutional neural networks for semantic segmentation and classification tasks. It uses a uniquely designed image, called an aggregated micropatch (AMP), as an input image to the classification network, thereby assessing whether the liver is in a cirrhotic stage. With a prototype AMP image, we synthesized a bunch of AMP images while retaining the textural features. This synthesis significantly increases the number of insufficient cirrhosis-labeled images, thereby circumventing overfitting issues and optimizing network performance. Furthermore, the synthesized AMP images contained unique textural patterns, mostly generated on the boundaries between adjacent micropatches (μ-patches) during their aggregation. These newly created boundary patterns provide rich information regarding the texture features of the ultrasound image, thereby making cirrhosis diagnosis more accurate and sensitive. The experimental results demonstrated that our proposed AMP image synthesis is extremely effective in expanding the dataset of cirrhosis images, thus diagnosing liver cirrhosis with considerably high accuracy. We achieved an accuracy of 99.95 %, a sensitivity of 100 %, and a specificity of 99.9 % on the Samsung Medical Center dataset using 8 × 8 pixels-sized μ-patches. The proposed approach provides an effective solution to deep-learning models with limited-training data, such as medical imaging tasks.
Chemical cell lysis is an interesting topic in the research to Lab-on-a-Disc (LOD) platforms on account of its perfect compatibility with the centrifugal spin column format. However, standard procedures followed in chemical cell lysis require sophisticated non-contact temperature control as well as the use of pressure resistant valves. These requirements pose a significant challenge thereby making the automation of chemical cell lysis on an LOD extremely difficult to achieve. In this study, an LOD capable of performing fully automated chemical cell lysis is proposed, where a combination of chemical and thermal methods has been used. It comprises a sample inlet, phase change material sheet (PCMS)-based temperature sensor, heating chamber, and pressure resistant valves. The PCMS melts and solidifies at a certain temperature and thus is capable of indicating whether the heating chamber has reached a specific temperature. Compared to conventional cell lysis systems, the proposed system offers advantages of reduced manual labor and a compact structure that can be readily integrated onto an LOD. Experiments using Salmonella typhimurium strains were conducted to confirm the performance of the proposed cell lysis system. The experimental results demonstrate that the proposed system has great potential in realizing chemical cell lysis on an LOD whilst achieving higher throughput in terms of purity and yield of DNA thereby providing a good alternative to conventional cell lysis systems.
Lab-on-a-disc (LOD) has emerged as a promising candidate for a point-of-care testing (POCT) device because it can effectively integrate complex fluid manipulation steps using multiple layers of polymeric substrates. However, it is still highly challenging to design and fabricate temperature measurement and heating system in non-contact with the surface of LOD, which is a prerequisite to successful realization of DNA amplification especially with a rotatable disc. This study presents a Lab-on-a-disc (LOD)-based automatic loop-mediated isothermal amplification (LAMP) system, where a thermochromic coating (<~420 µm) was used to distantly measure the chamber’s temperature and a micro graphite film was integrated into the chamber to remotely absorb laser beam with super high efficiency. We used a deep learning network to more consistently analyze the product of LAMP than we could with the naked eye. Consequently, both temperature heating and measurement were carried out without a physical contact with the surface of LOD. The experimental results show that the proposed approach, which no previous work has attempted, was highly effective in realizing LAMP in LOD.