Effect of Image Noise on the Classification of Skin Lesions Using Deep Convolutional Neural Networks

Skin lesions are in a category of disease that is both common in humans and a major cause of death.The classification accuracy of skin lesions is a crucial determinant of the success rate of curing lethal diseases.Deep Convolutional Neural Networks(CNNs)are now the most prevalent computer algorithms for the purpose of disease classification.As with all algorithms,CNNs are sensitive to noise from imaging devices,which often contaminates the quality of the images that are fed into them.In this paper,a deep CNN(Inception-v3)is used to study the effect of image noise on the classification of skin lesions.Gaussian noise,impulse noise,and noise made up of a compound of the two are added to an image dataset,namely the Dermofit Image Library from the University of Edinburgh.Evaluations,based on t-distributed Stochastic Neighbor Embedding(t-SNE)visualization,Receiver Operating Characteristic(ROC)analysis,and saliency maps,demonstrate the reliability of the Inception-v3 deep CNN in classifying noisy skin lesion images.

Compensation of image noise induced by difference in optical characteristics of array-LED chips of LED print head

There are various applied electro-optical devices, which utilize light emitting didoe(LED) chip array for applications to displays and opto-electronic sensors. In those devices, it is the one of the critical technical issues to minimize uncertain fluctuations including optical power and optical density. Due to variation in operating environment of a device, those are not corrected precisely by controlling parameters based on simple relation between parameters and resultant abovementioned outputs.Therefore, there is essential need to correct outputs in real-time based on correction function generated from the consideration on various operation condition. In this article, we introduce an output correction method through reporting real-time image noise reduction in the application to electro-photography with LED print head. In the technology of LED print head, as differences in optical characteristics between each LED cause vertical image noise, it should be corrected in order to obtain images that are comparable or better in quality compared to those produced by the conventional laser scanning method. Even though it seems that the method used to obtain uniform light power from each LED can solve this problem, it does not work well for high-resolution printing. Therefore, a scan method involving correction by a printed and scanned pattern is introduced through this work. The scan method is composed of correction patterns to minimize printing noise by its shape, the correction algorithm to calculate the optimized value and the printing algorithm to control gray levels in real-time precisely. We believe that the developed correction method upgrades the printing quality of the LPH printer better than commercial printers. The developed correction method can also be applied to various application areas that use an array-type light source such as display systems and lighting systems.

Modeling random telegraph signal noise in CMOS image sensor under low light based on binomial distribution

The random telegraph signal noise in the pixel source follower MOSFET is the principle component of the noise in the CMOS image sensor under low light. In this paper, the physical and statistical model of the random telegraph signal noise in the pixel source follower based on the binomial distribution is set up. The number of electrons captured or released by the oxide traps in the unit time is described as the random variables which obey the binomial distribution. As a result,the output states and the corresponding probabilities of the first and the second samples of the correlated double sampling circuit are acquired. The standard deviation of the output states after the correlated double sampling circuit can be obtained accordingly. In the simulation section, one hundred thousand samples of the source follower MOSFET have been simulated,and the simulation results show that the proposed model has the similar statistical characteristics with the existing models under the effect of the channel length and the density of the oxide trap. Moreover, the noise histogram of the proposed model has been evaluated at different environmental temperatures.

Terahertz Three-Dimensional Imaging Based on Computed Tomography with Photonics-Based Noise Source

Computed tomography has been proven to be useful for non-destructive inspection of structures and materials. We build a three-dimensional imaging system with the photonically generated incoherent noise source and the Schottky barrier diode detector in the terahertz frequency band （90–140GHz）. Based on the computed tomography technique, the three-dimensional image of a ceramic sample is reconstructed successfully by stacking the slices at different heights. The imaging results not only indicate the ability of terahertz wave in the non-invasive sensing and non-destructive inspection applications, but also prove the effectiveness and superiority of the uni-traveling-carrier photodiode as a terahertz source in the imaging applications.

IMAGE IMPROVEMENT IN THE WAVELET DOMAIN FOR OPTICAL COHERENCE TOMOGRAMS

In this paper,an image processing method for improving the quality of optical coherence tomography(OCT)images is proposed.Wavelet denoising based on context modeling and contrast enhancement by means of the contrast measure in the wavelet domain is carried out on the OCT images in succession.Three parameters are selected to assess the effectiveness of the method.It is shown from the results that the proposed method can not only enhance the contrast of images,but also improve signal-to-noise ratio.Compared with two other typical algorithms,it has the best visual effect.

SAR Change Detection Algorithm Combined with FFDNet Spatial Denoising

Objectives:When detecting changes in synthetic aperture radar(SAR)images,the quality of the difference map has an important impact on the detection results,and the speckle noise in the image interferes with the extraction of change information.In order to improve the detection accuracy of SAR image change detection and improve the quality of the difference map,this paper proposes a method that combines the popular deep neural network with the clustering algorithm.Methods:Firstly,the SAR image with speckle noise was constructed,and the FFDNet architecture was used to retrain the SAR image,and the network parameters with better effect on speckle noise suppression were obtained.Then the log ratio operator is generated by using the reconstructed image output from the network.Finally,K-means and FCM clustering algorithms are used to analyze the difference images,and the binary map of change detection results is generated.Results:The experimental results have high detection accuracy on Bern and Sulzberger’s real data,which proves the effectiveness of the method.

Computed tomography colonography and radiation risk:How low can we go?

Computed tomography colonography(CTC)has become a key examination in detecting colonic polyps and colorectal carcinoma(CRC).It is particularly useful after incomplete optical colonoscopy(OC)for patients with sedation risks and patients anxious about the risks or potential discomfort associated with OC.CTC's main advantages compared with OC are its non-invasive nature,better patient compliance,and the ability to assess the extracolonic disease.Despite these advantages,ionizing radiation remains the most significant burden of CTC.This opinion review comprehensively addresses the radiation risk of CTC,incorporating imaging technology refinements such as automatic tube current modulation,filtered back projections,lowering the tube voltage,and iterative reconstructions as tools for optimizing low and ultra-low dose protocols of CTC.Future perspectives arise from integrating artificial intelligence in computed tomography machines for the screening of CRC.

Quasi noise-free digital holography

One of the main drawbacks of Digital Holography(DH)is the coherent nature of the light source,which severely corrupts the quality of holographic reconstructions.Although numerous techniques to reduce noise in DH have provided good results,holographic noise suppression remains a challenging task.We propose a novel framework that combines the concepts of encoding multiple uncorrelated digital holograms,block grouping and collaborative filtering to achieve quasi noise-free DH reconstructions.The optimized joint action of these different image-denoising methods permits the removal of up to 98%of the noise while preserving the image contrast.The resulting quality of the hologram reconstructions is comparable to the quality achievable with non-coherent techniques and far beyond the current state of art in DH.Experimental validation is provided for both singlewavelength and multi-wavelength DH,and a comparison with the most used holographic denoising methods is performed.

Biophotons Contribute to Retinal Dark Noise

The discovery of dark noise in retinal photoreceptors resulted in a long-lasting controversy over its origin and the underlying mechanisms.Here,we used a novel ultra-weak biophoton imaging system（UBIS） to detect biophotonic activity（emission） under dark conditions in rat and bullfrog（Rana catesbeiana） retinas in vitro.We found a significant temperature-dependent increase in biophotonic activity that was completely blocked either by removing intracellular and extracellular Ca^（2＋）together or inhibiting phosphodiesterase 6.These findings suggest that the photon-like component of discrete dark noise may not be caused by a direct contribution of the thermal activation of rhodopsin,but rather by an indirect thermal induction of biophotonic activity,which then activates the retinal chromophore of rhodopsin.Therefore,this study suggests a possible solution regarding the thermal activation energy barrier for discrete dark noise,which has been debated for almost half a century.