Category: Nevin Manimala

Med Phys. 2023 Mar 10. doi: 10.1002/mp.16343. Online ahead of print.

ABSTRACT

BACKGROUD: Multiparametric magnetic resonance imaging (mp-MRI) is introduced and established as a noninvasive alternative for Prostate cancer (Pca) detection and characterization.

PURPOSE: To develop and evaluate a mutually communicated deep learning segmentation and classification network (MC-DSCN) based on mp-MRI for prostate segmentation and prostate cancer (PCa) diagnosis.

METHODS: The proposed MC-DSCN can transfer mutual information between segmentation and classification components and facilitate each other in a bootstrapping way. For classification task, the MC-DSCN can transfer the masks produced by the coarse segmentation component to the classification component to exclude irrelevant regions and facilitate classification. For segmentation task, this model can transfer the high-quality localization information learned by the classification component to the fine segmentation component to mitigate the impact of inaccurate localization on segmentation results. Consecutive MRI exams of patients were retrospectively collected from two medical centers (referred to as center A and B) . Two experienced radiologists segmented the prostate regions, and the ground truth of the classification refers to the prostate biopsy results. MC-DSCN was designed, trained, and validated using different combinations of distinct MRI sequences as input (e.g., T2-weighted and apparent diffusion coefficient) and the effect of different architectures on the network’s performance was tested and discussed. Data from center A was used for training, validation, and internal testing, while another center’s data was used for external testing. The statistical analysis is performed to evaluate the performance of the MC-DSCN. The DeLong test and paired t-test were used to assess the performance of classification and segmentation, respectively.

RESULTS: In total, 134 patients were included. The proposed MC-DSCN outperforms the networks that were designed solely for segmentation or classification. Regarding the segmentation task, the classification localization information helped to improve the IOU in center A: from 84.5% to 87.8% (p < 0.01) and in center B: from 83.8% to 87.1% (p < 0.01), while the AUC of PCa classification was improved in center A: from 0.946 to 0.991 (p < 0.02) and in center B: from 0.926 to 0.955 (p < 0.01) as a result of the additional information provided by the prostate segmentation.

CONCLUSION: The proposed architecture could effectively transfer mutual information between segmentation and classification components and facilitate each other in a bootstrapping way, thus outperforming the networks designed to perform only one task. This article is protected by copyright. All rights reserved.

PMID:36905102 | DOI:10.1002/mp.16343

Sensors (Basel). 2023 Mar 5;23(5):2839. doi: 10.3390/s23052839.

ABSTRACT

Emotional intelligence (EI) is a critical social intelligence skill that refers to an individual’s ability to assess their own emotions and those of others. While EI has been shown to predict an individual’s productivity, personal success, and ability to maintain positive relationships, its assessment has primarily relied on subjective reports, which are vulnerable to response distortion and limit the validity of the assessment. To address this limitation, we propose a novel method for assessing EI based on physiological responses-specifically heart rate variability (HRV) and dynamics. We conducted four experiments to develop this method. First, we designed, analyzed, and selected photos to evaluate the ability to recognize emotions. Second, we produced and selected facial expression stimuli (i.e., avatars) that were standardized based on a two-dimensional model. Third, we obtained physiological response data (HRV and dynamics) from participants as they viewed the photos and avatars. Finally, we analyzed HRV measures to produce an evaluation criterion for assessing EI. Results showed that participants’ low and high EI could be discriminated based on the number of HRV indices that were statistically different between the two groups. Specifically, 14 HRV indices, including HF (high-frequency power), lnHF (the natural logarithm of HF), and RSA (respiratory sinus arrhythmia), were significant markers for discerning between low and high EI groups. Our method has implications for improving the validity of EI assessment by providing objective and quantifiable measures that are less vulnerable to response distortion.

PMID:36905043 | DOI:10.3390/s23052839

Sensors (Basel). 2023 Mar 3;23(5):2789. doi: 10.3390/s23052789.

ABSTRACT

Based on the observations made in rheumatology clinics, autoimmune disease (AD) patients on immunosuppressive (IS) medications have variable vaccine site inflammation responses, whose study may help predict the long-term efficacy of the vaccine in this at-risk population. However, the quantitative assessment of the inflammation of the vaccine site is technically challenging. In this study analyzing AD patients on IS medications and normal control subjects, we imaged the inflammation of the vaccine site 24 h after mRNA COVID-19 vaccinations were administered using both the emerging photoacoustic imaging (PAI) method and the established Doppler ultrasound (US) method. A total of 15 subjects were involved, including 6 AD patients on IS and 9 normal control subjects, and the results from the two groups were compared. Compared to the results obtained from the control subjects, the AD patients on IS medications showed statistically significant reductions in vaccine site inflammation, indicating that immunosuppressed AD patients also experience local inflammation after mRNA vaccination but not in as clinically apparent of a manner when compared to non-immunosuppressed non-AD individuals. Both PAI and Doppler US were able to detect mRNA COVID-19 vaccine-induced local inflammation. PAI, based on the optical absorption contrast, shows better sensitivity in assessing and quantifying the spatially distributed inflammation in soft tissues at the vaccine site.

PMID:36904999 | DOI:10.3390/s23052789

Sensors (Basel). 2023 Mar 3;23(5):2793. doi: 10.3390/s23052793.

ABSTRACT

(1) Background: One effect of microgravity on the human body is fluid redistribution due to the removal of the hydrostatic gravitational gradient. These fluid shifts are expected to be the source of severe medical risks and it is critical to advance methods to monitor them in real-time. One technique to monitor fluid shifts captures the electrical impedance of segmental tissues, but limited research is available to evaluate if fluid shifts in response to microgravity are symmetrical due to the bilateral symmetry of the body. This study aims to evaluate this fluid shift symmetry. (2) Methods: Segmental tissue resistance at 10 kHz and 100 kHz was collected at 30 min intervals from the left/right arm, leg, and trunk of 12 healthy adults over 4 h of 6° head-down-tilt body positioning. (3) Results: Statistically significant increases were observed in the segmental leg resistances, first observed at 120 min and 90 min for 10 kHz and 100 kHz measurements, respectively. Median increases were approximately 11% to 12% for the 10 kHz resistance and 9% for the 100 kHz resistance. No statistically significant changes in the segmental arm or trunk resistance. Comparing the left and right segmental leg resistance, there were no statistically significant differences in the resistance changes based on the side of the body. (4) Conclusions: The fluid shifts induced by the 6° body position resulted in similar changes in both left and right body segments (that had statistically significant changes in this work). These findings support that future wearable systems to monitor microgravity-induced fluid shifts may only require monitoring of one side of body segments (reducing the hardware needed for the system).

PMID:36904995 | DOI:10.3390/s23052793

Sensors (Basel). 2023 Mar 1;23(5):2708. doi: 10.3390/s23052708.

ABSTRACT

Topic modeling is a machine learning algorithm based on statistics that follows unsupervised machine learning techniques for mapping a high-dimensional corpus to a low-dimensional topical subspace, but it could be better. A topic model’s topic is expected to be interpretable as a concept, i.e., correspond to human understanding of a topic occurring in texts. While discovering corpus themes, inference constantly uses vocabulary that impacts topic quality due to its size. Inflectional forms are in the corpus. Since words frequently appear in the same sentence and are likely to have a latent topic, practically all topic models rely on co-occurrence signals between various terms in the corpus. The topics get weaker because of the abundance of distinct tokens in languages with extensive inflectional morphology. Lemmatization is often used to preempt this problem. Gujarati is one of the morphologically rich languages, as a word may have several inflectional forms. This paper proposes a deterministic finite automaton (DFA) based lemmatization technique for the Gujarati language to transform lemmas into their root words. The set of topics is then inferred from this lemmatized corpus of Gujarati text. We employ statistical divergence measurements to identify semantically less coherent (overly general) topics. The result shows that the lemmatized Gujarati corpus learns more interpretable and meaningful subjects than unlemmatized text. Finally, results show that lemmatization curtails the size of vocabulary decreases by 16% and the semantic coherence for all three measurements-Log Conditional Probability, Pointwise Mutual Information, and Normalized Pointwise Mutual Information-from -9.39 to -7.49, -6.79 to -5.18, and -0.23 to -0.17, respectively.

PMID:36904915 | DOI:10.3390/s23052708

Sensors (Basel). 2023 Mar 1;23(5):2712. doi: 10.3390/s23052712.

ABSTRACT

A wireless channel digital twin is a useful tool to evaluate the performance of a communication system at the physical or link level by generating the physical channel controllably. In this paper, a stochastic general fading channel model is proposed, which considered most of the channel fading types for various communication scenarios. By using the sum-of-frequency-modulation (SoFM) method, the phase discontinuity of the generated channel fading was well addressed. On this basis, a general and flexible generation architecture for channel fading was developed on a field programmable gate array (FPGA) platform. In this architecture, improved CORDIC-based hardware circuits for the trigonometric function, exponential function, and natural logarithm were designed and implemented, which improved the real-time performance of the system and the utilization rate of the hardware resources compared with the traditional LUT and CORDIC method. For a 16-bit fixed-point data bit width single-channel emulation, the hardware resource consumption was significantly reduced from 36.56% to 15.62% for the overall system by utilizing the compact time-division (TD) structure. Moreover, the classical CORDIC method brought an extra latency of 16 system clock cycles, while the latency caused by the improved CORDIC method was decreased by 62.5%. Finally, a generation scheme of a correlated Gaussian sequence was developed to introduce a controllable arbitrary space-time correlation for the channel generator with multiple channels. The output results of the developed generator were consistent with the theoretical results, which verified the correctness of both the generation method and hardware implementation. The proposed channel fading generator can be applied for the emulation of large-scale multiple-input, multiple-output (MIMO) channels under various dynamic communication scenarios.

PMID:36904913 | DOI:10.3390/s23052712

Sensors (Basel). 2023 Feb 28;23(5):2656. doi: 10.3390/s23052656.

ABSTRACT

We have developed the New Jersey Institute of Technology-Home Virtual Rehabilitation System (NJIT-HoVRS) to facilitate intensive, hand-focused rehabilitation in the home. We developed testing simulations with the goal of providing richer information for clinicians performing remote assessments. This paper presents the results of reliability testing examining differences between in-person and remote testing as well as discriminatory and convergent validity testing of a battery of six kinematic measures collected with NJIT-HoVRS. Two different groups of persons with upper extremity impairments due to chronic stroke participated in two separate experiments. Data Collection: All data collection sessions included six kinematic tests collected with the Leap Motion Controller. Measurements collected include hand opening range, wrist extension range, pronation-supination range, hand opening accuracy, wrist extension accuracy, and pronation-supination accuracy. The system usability was evaluated by therapists performing the reliability study using the System Usability Scale. When comparing the in-laboratory collection and the first remote collection, the intra-class correlation coefficients (ICC) for three of the six measurements were above 0.900 and the other three were between 0.500 and 0.900. Two of the first remote collection/second remote collection ICCs were above 0.900, and the other four were between 0.600 and 0.900. The 95% confidence intervals for these ICC were broad, suggesting that these preliminary analyses need to be confirmed by studies with larger samples. The therapist’s SUS scores ranged from 70 to 90. The mean was 83.1 (SD = 6.4), which is consistent with industry adoption. There were statistically significant differences in the kinematic scores when comparing unimpaired and impaired UE for all six measures. Five of six impaired hand kinematic scores and five of six impaired/unimpaired hand difference scores demonstrated correlations between 0.400 and 0.700 with UEFMA scores. Reliability for all measures was acceptable for clinical practice. Discriminant and convergent validity testing suggest that scores on these tests may be meaningful and valid. Further testing in a remote setting is necessary to validate this process.

PMID:36904860 | DOI:10.3390/s23052656

Sensors (Basel). 2023 Feb 28;23(5):2654. doi: 10.3390/s23052654.

ABSTRACT

Straight bevel gears are widely used in mining equipment, ships, heavy industrial equipment, and other fields due to their high capacity and robust transmission. Accurate measurements are essential in order to determine the quality of bevel gears. We propose a method for measuring the accuracy of the top surface profile of the straight bevel gear teeth based on binocular visual technology, computer graphics, error theory, and statistical calculations. In our method, multiple measurement circles are established at equal intervals from the small end of the top surface of the gear tooth to the large end, and the coordinates of the intersection points of these circles with the tooth top edge lines of the gear teeth are extracted. The coordinates of these intersections are fitted to the top surface of the tooth based on NURBS surface theory. The surface profile error between the fitted top surface of the tooth and the designed surface is measured and determined based on the product use requirements, and if this is less than a given threshold, the product is acceptable. With a module of 5 and an eight-level precision, such as the straight bevel gear, the minimum surface profile error measured was -0.0026 mm. These results demonstrate that our method can be used to measure surface profile errors in the straight bevel gears, which will broaden the field of in-depth measurements for the straight bevel gears.

PMID:36904858 | DOI:10.3390/s23052654

Sensors (Basel). 2023 Feb 27;23(5):2607. doi: 10.3390/s23052607.

ABSTRACT

Cannabis is commercially cultivated for both therapeutic and recreational purposes in a growing number of jurisdictions. The main cannabinoids of interest are cannabidiol (CBD) and delta-9 tetrahydrocannabidiol (THC), which have applications in different therapeutic treatments. The rapid, nondestructive determination of cannabinoid levels has been achieved using near-infrared (NIR) spectroscopy coupled to high-quality compound reference data provided by liquid chromatography. However, most of the literature describes prediction models for the decarboxylated cannabinoids, e.g., THC and CBD, rather than naturally occurring analogues, tetrahydrocannabidiolic acid (THCA) and cannabidiolic acid (CBDA). The accurate prediction of these acidic cannabinoids has important implications for quality control for cultivators, manufacturers and regulatory bodies. Using high-quality liquid chromatography-mass spectroscopy (LCMS) data and NIR spectra data, we developed statistical models including principal component analysis (PCA) for data quality control, partial least squares regression (PLS-R) models to predict cannabinoid concentrations for 14 different cannabinoids and partial least squares discriminant analysis (PLS-DA) models to characterise cannabis samples into high-CBDA, high-THCA and even-ratio classes. This analysis employed two spectrometers, a scientific grade benchtop instrument (Bruker MPA II-Multi-Purpose FT-NIR Analyzer) and a handheld instrument (VIAVI MicroNIR Onsite-W). While the models from the benchtop instrument were generally more robust (99.4-100% accuracy prediction), the handheld device also performed well (83.1-100% accuracy prediction) with the added benefits of portability and speed. In addition, two cannabis inflorescence preparation methods were evaluated: finely ground and coarsely ground. The models generated from coarsely ground cannabis provided comparable predictions to that of the finely ground but represent significant timesaving in terms of sample preparation. This study demonstrates that a portable NIR handheld device paired with LCMS quantitative data can provide accurate cannabinoid predictions and potentially be of use for the rapid, high-throughput, nondestructive screening of cannabis material.

PMID:36904818 | DOI:10.3390/s23052607

Sensors (Basel). 2023 Feb 26;23(5):2591. doi: 10.3390/s23052591.

ABSTRACT

Cardiac and respiratory diseases are the primary causes of health problems. If we can automate anomalous heart and lung sound diagnosis, we can improve the early detection of disease and enable the screening of a wider population than possible with manual screening. We propose a lightweight yet powerful model for simultaneous lung and heart sound diagnosis, which is deployable in an embedded low-cost device and is valuable in remote areas or developing countries where Internet access may not be available. We trained and tested the proposed model with the ICBHI and the Yaseen datasets. The experimental results showed that our 11-class prediction model could achieve 99.94% accuracy, 99.84% precision, 99.89% specificity, 99.66% sensitivity, and 99.72% F1 score. We designed a digital stethoscope (around USD 5) and connected it to a low-cost, single-board-computer Raspberry Pi Zero 2W (around USD 20), on which our pretrained model can be smoothly run. This AI-empowered digital stethoscope is beneficial for anyone in the medical field, as it can automatically provide diagnostic results and produce digital audio records for further analysis.

PMID:36904794 | DOI:10.3390/s23052591