Shresth Gupta

•The objective is to cover the maximum part of target without affecting the OARs.•An iterative reweighting algorithm based approach is used to achieve optimality.•09 equally separated beams with dose-volume limits on the OAR have used to treat PTV.•Considered multiple samples from CORT dataset to make a judgment.•Optimal leaf trajectory matching by keeping radiation delivery fixed for each samples. For a Fluence map optimization problem, we present a unique iterative reweighting algorithm-based method to achieve optimality. The purpose of planning is to determine a dose distribution in such a manner that it should cover the maximum part of the target without affecting the functionality of organ at risk. We suggest a unique methodology to solve dose-volume bounds while maintaining their non-convexity, as compared to earlier approaches that dedicated to convex estimation. The suggested approach is cooperative to competent procedures centered on partial minimization and certainly adjusts to tackle maximum-dose bounds. Fluence mapping is used for the inverse planning which determines the intensity profile of each beam. For the analysis, 09 equally separated beams with dose-volume limits on the organ at risk have used to treat malignancies in the prostate. Cumulative dose-volume histogram is used for the treatment plans quality measurement. We considered 4-different samples from CORT dataset named prostate, Liver, TG119 phantom, and Head & Neck to make a judgment about the generalizability of the performance of the proposed algorithm, and after getting this we finally compared our result with other method using prostate sample to validate our approach. After comparing with other methods centered on dose volume limits we found that in our approach maximum dose can deliver within minimum time across the planned target. For the optimization of trajectory we kept, the radiation delivery rate fixed to its extreme level and employed the independence property of each single leaf pair equally. By making this provision for all total T delivery times, trade-off between distribution time and fluency mapping quality is produced.

Conference Title: 2022 44th Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC) Conference Start Date: 2022, July 11 Conference End Date: 2022, July 15 Conference Location: Glasgow, Scotland, United KingdomWith the rapid development of communications, information technology, and Internet of Things (IoT), photoplethysmography (PPG) has achieved prevalence in telemedicine and remote health monitoring using wearable devices. As these devices are resource-constrained, efficient compression techniques are necessary for optimal storage and power consumption management with sustained clinical morphology of PPG signal. This work presents a new approach for PPG compression based on the energy of discrete cosine transform (DCT). With better data compression, a significant reduction of noise is also obtained in the reconstructed signal. The performance evaluation is done on multiple publicly available databases, an average compression ratio of 37.46 is achieved on MIMIC-II database with percentage root-mean-square difference of 0.0688, which is significantly higher than the existing PPG-compression approaches in the literature. Clinical relevance- The work is intended to promote the integration of this combined compression and denoising algorithm for PPG-based devices. The approach is helpful to impart optimization of memory and preserve the clinically relevant information of PPG signals in terms of its useful fiducial points. Further it can amplify ubiquitous computing health monitoring with faster diagnosis and feedback.

Conference Title: TENCON 2022 - 2022 IEEE Region 10 Conference (TENCON) Conference Start Date: 2022, Nov. 1 Conference End Date: 2022, Nov. 4 Conference Location: Hong Kong, Hong KongPhotoplethysmogram (PPG) signals are often assessed using the transmission and reflection techniques, which detect light passed through or reflected by tissue. The transmission technique is appropriate for body sites with thin tissue, such as the index finger. In contrast, the reflection approach applies to most body sites. As each body site impacts inducing noise while acquiring PPG, this study investigates the quality of PPG signal obtained from five body sites, i.e., fingertip, forehead, toe, wrist-upper, and wrist-lower. After denoising with the Savitzky Golay filter, we calculated popular state-of-the-art parameters to quantify the obtained denoised signal to indicate the quality of PPG. Simulation results after denoising and analysis reveal that the PPG acquired from fingertip exhibits the highest quality and yields a signal-to-noise ratio $SNR=39.8672dB$ with a minimum percentage root mean square difference $(PRD)$ of 1.0154 as compared to the PPG acquired from other body sites.

Proper blood pressure measurement is essential for diagnosis and appropriate medical treatment for cardiovascular disease and hypertension. Numerous attempts were made to derive meaningful features from photoplethysmogram (PPG) to correlate with mean arterial pressure (MAP) and pulse pressure (PP), however, satisfactory performance could not be achieved yet. This letter proposed an enhanced index termed dynamic large artery stiffness index to analyze the contour of PPG between systolic and diastolic points and employed its contribution toward noninvasive cuffless blood pressure estimation and as a significant biomarker for cardiovascular status. The proposed index demonstrated a good association with MAP and PP with average Pearson correlation coefficients of 0.87 and 0.65, respectively. The proposed index is combined with 12 other prominent PPG features to estimate systolic, diastolic, and mean arterial pressure. The estimation results yield mean absolute errors (with standard deviation) of (0.55 \pm 1.62), (0.32 \pm 0.89), and (0.35 \pm 0.95) for SBP, DBP, and MAP, respectively, on MIMIC-I database while (2.41 \pm 5.26), (0.41 \pm 1.50), and (0.65 \pm 1.34) on UCI repository database. It achieves grade-A by comparison with the global standard benchmark of British Hypertension Society for all the three estimations, i.e., SBP, DBP, and MAP.

Monitoring Mean Arterial Pressure (MAP) helps calculate the arteries' flow, resistance, and pressure. It allows doctors to check how well the blood flows through our body and reaches all major organs. Photoplethysmogram technology is gaining momentum and popularity in smart wearable devices to monitor cuff-less blood pressure (BP). However, the performance reliability of the existing PPG-based BP estimation devices is still poor. Inaccuracy in estimating systolic and diastolic blood pressure leads to an overall imprecision in resultant MAP values. Hence, there is a need for robust and reliable MAP estimation algorithms. This work exploits the moving slope features of PPG contour in its first and second derivatives that directly correlate with MAP and does not require estimating systolic and diastolic blood pressure values. The proposed approach is evaluated using two different data sets (i.e., MIMIC-I and MIMIC-II) to demonstrate the robustness and reliability of the work for personalized non-invasive BP monitoring devices to estimate MAP directly. A mean absolute error of 1.28mmHg and a standard deviation of 2.50mmHg is obtained with MIMIC-II data-set using GridSearchCV random forest regressor that outperformed most of the existing related works.

Conference Title: TENCON 2022 - 2022 IEEE Region 10 Conference (TENCON) Conference Start Date: 2022, Nov. 1 Conference End Date: 2022, Nov. 4 Conference Location: Hong Kong, Hong KongColorectal cancer (CRC) is a significant public health concern worldwide. Its early detection is critical since it dictates treatment options and significantly impacts survival time. A pathologist can make a histological diagnosis based on histological images obtained from a colonoscopy biopsy. The traditional visual assessment is time-consuming and highly unreliable because of the subjectivity of the evaluation. On the other hand, current approaches primarily rely on the use of diverse combinations of textual features and classifiers, as well as transfer learning, to classify distinct organizational types. However, the classification remains difficult since histological pictures comprise various tissue types and properties. In this work, we propose a deep learning technique based on pre-trained Convolutional Neural Networks (CNNs) for distinguishing eight classes of adenocarcinomas from healthy tissues. We explored multiple CNN architectures such as RESNET50, INCEPTIONV3, VGG 16, VGG19, RESNET152V2 and heuristically searched the best architecture for CRC detection by varying different optimizers. It was found that INCEPTIONV3 is the best performing model with 89.8% average accuracy for eight CRC classes.

—Wearable health monitoring devices based on pho-toplethysmogram (PPG) can estimate valuable physiological parameters and are often regarded as a touchstone for indicating the cardiovascular status of a person. Severe motion artifacts are generally observed in the acquired PPG signals from the biosensors in the wearables when the PPG acquisition is performed during different physical activities. Classical filters can remove thermal noise and electromagnetic interference from the raw PPG signals, but they fail to handle the motion artifacts that cover a wider range of signal frequencies. This work introduces a new motion artifact removal scheme that exploits the sparsity of PPG signals in the tunable-Q factor wavelet transform (TQWT) domain using the basis pursuit denoising (BPDN) scheme. The proposed scheme deftly removes noise while preserving the original morphology of the PPG signal without using any reference accelerometer sensor data. In particular, a regularized SALSA-based BPDN algorithm is employed for sparse recovery, which utilizes a soft thresholding approach to denoise the noisy PPG signals. TQWT can be tuned according to the oscillatory behavior of the raw PPG signal. At the same time, BPDN exploits the sparsity of the wavelet coefficients to yield an artifact-free reconstruction of PPG. Further, heart rate estimation is performed using the denoised PPG signal with support vector regression to demonstrate the clinical information preservation during the proposed denoising process. The proposed denoising scheme is evaluated on two publicly available databases, and an overall minimum mean absolute error (MAE) of 1.03 and 0.76 beats per minute (bpm) is achieved. The suggested denoising approach has demonstrated better performance with the lowest MAE compared to the previously reported methods.

Diabetes is a group of metabolic disorders characterized by hyperglycemia caused by a deficiency in insulin production, insulin action, or both. Type-II diabetes mellitus (DM-2) complications include retinopathy, cardiovascular disorder and diabetic neuropathy. The existing works for the diagnosis of DM-2 using photoplethysmogram (PPG) utilize several time-domain features and demographic parameters of the individuals. However, the current features do not indicate a clinical correlation between DM-2 and its functional influence on cardiovascular regulation. This work proposes a novel index, called dSVRI, as a feature based on systemic vascular resistance pathology to discriminate between a healthy and diabetic subject. The discrimination ability and the diagnostic performance of the proposed dSVRI were compared with the existing time-domain PPG features and its higher derivatives. In experiments with a publicly available data-set (having 219 subjects including healthy and DM-2 individuals), an accuracy of 98.52% is obtained using grid search random forest, which is significantly higher than the existing methods.

Continuous blood pressure monitoring is essential for persons at risk of hypertension and cardiovascular disease. This work presents the analysis of different temporal features of photoplethysmogram (PPG) useful for estimating cuffless blood pressure by utilizing several statistical test approaches to identify the features’ contribution to this estimation and their correlation with a target mean arterial pressure values. The regression is performed using a random forest regressor to estimate mean arterial pressure (MAP) with temporal features, and statistical analysis with a ranking of features is done after estimation using p-value, correlation, and z-test. The significant ranking temporal features are selected and used to estimate MAP, DBP, and SBP.

Wearable health-monitoring devices based on photoplethysmogram (PPG) can estimate valuable physiological parameters and are often regarded as a touchstone for indicating the cardiovascular status of a person. Severe motion artifacts are generally observed in the acquired PPG signals from the biosensors in the wearables when the PPG acquisition is performed during different physical activities. Classical filters can remove thermal noise and electromagnetic interference from the raw PPG signals, but they fail to handle the motion artifacts that cover a wider range of signal frequencies. This work introduces a new motion artifact removal scheme that exploits the sparsity of PPG signals in the tunable-[Formula Omitted] factor wavelet transform (TQWT) domain using the basis pursuit denoising (BPDN) scheme. The proposed scheme deftly removes noise while preserving the original morphology of the PPG signal without using any reference accelerometer sensor data. In particular, a regularized SALSA-based BPDN algorithm is employed for sparse recovery, which utilizes a soft thresholding approach to denoise the noisy PPG signals. TQWT can be tuned according to the oscillatory behavior of the raw PPG signal. At the same time, BPDN exploits the sparsity of the wavelet coefficients to yield an artifact-free reconstruction of PPG. Further, heart rate estimation is performed using the denoised PPG signal with support vector regression (SVR) to demonstrate the clinical information preservation during the proposed denoising process. The proposed denoising scheme is evaluated on two publicly available databases, and an overall minimum mean absolute error (MAE) of 1.03 and 0.76 beats per minute (bpm) is achieved. The suggested denoising approach has demonstrated better performance with the lowest MAE compared to the previously reported methods.

Conference Title: 2022 IEEE 9th Uttar Pradesh Section International Conference on Electrical, Electronics and Computer Engineering (UPCON) Conference Start Date: 2022, Dec. 2 Conference End Date: 2022, Dec. 4 Conference Location: Prayagraj, IndiaThe ratio of a variable measurement's maximum and minimum values is known as its dynamic range. This range, as it relates to photography and cinematography, is the proportion between an image's whitest (brightest) and darkest (darkest) values. The main goal here is to enhance the detailed visibility of scenes. In this research, we present a three-adaptive step approach for image dynamic range modification that is both efficient and effective in terms of visibility and ease in use. First, two Gamma functions are adaptively selected on the basis of the histogram of the brightness map independently. Second, to balance the amplification of details in different places, an adaptive fusion technique is presented to integrate the two modified luminance maps. Finally, we propose a method to restore the color to the fused image.

Mean Arterial Pressure (MAP) is defined as central pressure in the arteries of a person during a single cardiac cycle. It is regarded as an important bio-marker of blood perfusion in vital organs as compared to systolic blood pressure (SBP). The actual MAP can be determined by manual monitoring and complex calculations limited to occasional monitoring status. Growing personalized health care monitoring devices have already evinced a variety of health parameters to track on a daily basis with the additional advantage of continuous, noninvasive, and unobstructed measurement. This work proposes a direct strategy for the estimation of mean arterial pressure without using the systolic and diastolic BP values. By exploring 13 significant morphological features from a single PPG signal which are most related to the target MAP are derived such as Pulse Interval, Inflection Ratio etc. The estimation is performed using LSTM network with an architecture having 2- LSTM layers followed by a dropout and dense layer. With 942 subjects of UCI repository dataset our model achieves a remarkable mean absolute error of 1.48, standard deviation of 2.36 and pearson correlation coefficient of 0.96 which is better as compared to the existing works and even chalked up the British Hypertension Society (BHS) benchmark with grade A.