CAD systems for COVID-19 diagnosis and disease stage classification by segmentation of infected regions from CT images.

BACKGROUND: Here propose a computer-aided diagnosis (CAD) system to differentiate COVID-19 (the coronavirus disease of 2019) patients from normal cases, as well as to perform infection region segmentation along with infection severity estimation using computed tomography (CT) images. The developed system facilitates timely administration of appropriate treatment by identifying the disease stage without reliance on medical professionals. So far, this developed model gives the most accurate, fully automatic COVID-19 real-time CAD framework. RESULTS: The CT image dataset of COVID-19 and non-COVID-19 individuals were subjected to conventional ML stages to perform binary classification. In the feature extraction stage, SIFT, SURF, ORB image descriptors and bag of features technique were implemented for the appropriate differentiation of chest CT regions affected with COVID-19 from normal cases. This is the first work introducing this concept for COVID-19 diagnosis application. The preferred diverse database and selected features that are invariant to scale, rotation, distortion, noise etc. make this framework real-time applicable. Also, this fully automatic approach which is faster compared to existing models helps to incorporate it into CAD systems. The severity score was measured based on the infected regions along the lung field. Infected regions were segmented through a three-class semantic segmentation of the lung CT image. Using severity score, the disease stages were classified as mild if the lesion area covers less than 25% of the lung area; moderate if 25-50% and severe if greater than 50%. Our proposed model resulted in classification accuracy of 99.7% with a PNN classifier, along with area under the curve (AUC) of 0.9988, 99.6% sensitivity, 99.9% specificity and a misclassification rate of 0.0027. The developed infected region segmentation model gave 99.47% global accuracy, 94.04% mean accuracy, 0.8968 mean IoU (intersection over union), 0.9899 weighted IoU, and a mean Boundary F1 (BF) contour matching score of 0.9453, using Deepabv3+ with its weights initialized using ResNet-50. CONCLUSIONS: The developed CAD system model is able to perform fully automatic and accurate diagnosis of COVID-19 along with infected region extraction and disease stage identification. The ORB image descriptor with bag of features technique and PNN classifier achieved the superior classification performance.

Formal reasoning about synthetic biology using higher-order-logic theorem proving.

Synthetic biology is an interdisciplinary field that uses well-established engineering principles for performing the analysis of the biological systems, such as biological circuits, pathways, controllers and enzymes. Conventionally, the analysis of these biological systems is performed using paper-and-pencil proofs and computer simulation methods. However, these methods cannot ensure accurate results due to their inherent limitations. Higher-order-logic (HOL) theorem proving is proposed and used as a complementary approach for analysing linear biological systems, which is based on developing a mathematical model of the genetic circuits and the bio-controllers used in synthetic biology based on HOL and analysing it using deductive reasoning in an interactive theorem prover. The involvement of the logic, mathematics and the deductive reasoning in this method ensures the accuracy of the analysis. It is proposed to model the continuous dynamics of the genetic circuits and their associated controllers using differential equations and perform their transfer function-based analysis using the Laplace transform in a theorem prover. For illustration, the genetic circuits of activated and repressed expressions and autoactivation of protein, and phase lag and lead controllers, which are widely used in cancer-cell identifiers and multi-input receptors for precise disease detection, are formally analyzed.

Error metrics determination in functionally approximated circuits using SAT solvers.

Approximate computing is an emerging design paradigm that offers trade-offs between output accuracy and computation efforts by exploiting some applications' intrinsic error resiliency. Computation of error metrics is of paramount importance in approximate circuits to measure the degree of approximation. Most of the existing techniques for evaluating error metrics apply simulations which may not be effective for evaluation of large complex designs because of an immense increase in simulation runtime and a decrease in accuracy. To address these deficiencies, we present a novel methodology that employs SAT (Boolean satisfiability) solvers for fast and accurate determination of error metrics specifically for the calculation of an average-case error and the maximum error rate in functionally approximated circuits. The proposed approach identifies the set of all errors producing assignments to gauge the quality of approximate circuits for real-life applications. Additionally, the proposed approach provides a test generation method to facilitate design choices, and acts as an important guide to debug the approximate circuits to discover and locate the errors. The effectiveness of the approach is demonstrated by evaluating the error metrics of several benchmark-approximated adders of different sizes. Experimental results on benchmark circuits show that the proposed SAT-based methodology accurately determines the maximum error rate and an average-case error within acceptable CPU execution time in one go, and further provides a log of error-generating input assignments.

FPGA Modeling and Optimization of a SIMON Lightweight Block Cipher.

Security of sensitive data exchanged between devices is essential. Low-resource devices (LRDs), designed for constrained environments, are increasingly becoming ubiquitous. Lightweight block ciphers provide confidentiality for LRDs by balancing the required security with minimal resource overhead. SIMON is a lightweight block cipher targeted for hardware implementations. The objective of this research is to implement, optimize, and model SIMON cipher design for LRDs, with an emphasis on energy and power, which are critical metrics for LRDs. Various implementations use field-programmable gate array (FPGA) technology. Two types of design implementations are examined: scalar and pipelined. Results show that scalar implementations require 39% less resources and 45% less power consumption. The pipelined implementations demonstrate 12 times the throughput and consume 31% less energy. Moreover, the most energy-efficient and optimum design is a two-round pipelined implementation, which consumes 31% of the best scalar's implementation energy. The scalar design that consumes the least energy is a four-round implementation. The scalar design that uses the least area and power is the one-round implementation. Balancing energy and area, the two-round pipelined implementation is optimal for a continuous stream of data. One-round and two-round scalar implementations are recommended for intermittent data applications.

Optic disc detection in retinal fundus images using gravitational law-based edge detection.

Diabetic retinopathy is one of the primary causes of vision loss worldwide. Early detection of the condition is critical for providing adequate treatment of this ailment to prevent vision loss. This detection is achieved by processing retinal fundus images. A key step in detecting diabetic retinopathy is identifying the optic disc in these images. The optic disc is similar in color and contrast to the exudates that indicate diabetic retinopathy. Hence, the optic disc has to be removed from the fundus image before exudates can be detected. Detecting the optic disc is also required in algorithms used for blood vessel segmentation in fundus images. Therefore, there is a need for approaches that accurately and quickly detect optic disc. This paper proposes a simple, deterministic, and time-efficient approach for optic disc detection by adapting an edge detection algorithm inspired by the gravitational law. Our method introduces novel pre- and post-detection steps that aim to increase the accuracy of the adapted detection method. In addition, a candidate selection technique is proposed to decrease the number of missed optic discs. The proposed methodology was found to have a detection rate of 100, 97.75, 92.90, and 95 % for DRIVE, DiaRet, DMED, and STARE datasets, respectively, which is comparatively better than existing optic disc detection schemes. Experimental results showed an average running time of 0.40 s per image, which is significantly lower than available methods published in the literature.