搜索结果：Handheld

Tomato leaf diseases substantially reduce tomato yields and quality and remain a persistent challenge for efficient crop management. Although deep learning-based detectors have achieved strong accuracy in controlled benchmarks, many existing solutions are still difficult to transfer to resource-constrained agricultural systems because they rely on high-end GPUs, consume considerable power, and often lose performance after deployment on embedded devices. To address this practical gap, this study proposes HGS-YOLO, a system-oriented deployable lightweight adaptation of YOLOv11 for leaf-level tomato disease detection, together with an end-to-end edge sensing pipeline for low-power agricultural deployment. The main contribution lies in the coordinated system-level co-design of model structure, optimization, and deployment rather than in a novel detector architecture. Specifically, YOLOv11 is adapted through three coordinated modifications: an HGNetV2 backbone for efficient feature extraction, an HS-FPN neck with channel attention for lightweight multi-scale fusion, and an MPDIoU loss function for more stable localization optimization. Beyond the model architecture, the study establishes a complete engineering pipeline that includes training, optimization, post-training quantization, and hardware deployment with BPU acceleration on a D-Robotics RDK X5 handheld platform. Comprehensive benchmark experiments indicate that HGS-YOLO achieves 93.6% mAP50 and 72.1% mAP@[0.5:0.95] with 86.5% recall, only 1.3 M parameters, and a 3.1 MB model size, substantially reducing the model complexity and storage cost relative to the YOLOv11 baseline. A three-seed retraining comparison shows that HGS-YOLO trades roughly 0.5 mAP50 points for this compactness (a statistically significant but small concession) and recovers the cost on the deployment side: on the RDK X5 chip, HGS-YOLO is the fastest, most memory-efficient, and lowest-power model among all compared detectors. Indoor deployment tests using separately collected tomato leaf samples further achieve 90.3% mAP50, 82.3% recall, 89.0% precision, 25.0 ± 0.4 ms end-to-end latency, 40.0 ± 0.6 FPS, and 9.8 ± 0.4 W average system power. After PTQ, the mAP50 drops from 93.6% to 93.0% on the same benchmark; because this figure was measured under controlled imaging conditions, it is presented as an in-distribution reference point rather than as evidence of robustness in the open field. We also took the handheld system into a working tomato greenhouse for a small outdoor field round, where it ran end-to-end and produced on-device disease detections under natural sunlight, specular highlights, partial occlusion, background clutter, and handheld motion blur. These results show that HGS-YOLO reaches a good balance of accuracy, efficiency, and deployability and that it works in the field on an independent small-scale test; validating it more widely across sites, seasons, and weather is left to future work.

To compare automated breast ultrasound (ABUS) and handheld breast ultrasound (HHUS) as adjuncts to digital mammography (DM) for preoperative assessment of local tumor extent in women with newly diagnosed early-stage breast cancer. In this prospective study conducted at three institutions, women with newly diagnosed early-stage breast cancer and no palpable axillary nodes scheduled for surgery underwent ABUS and HHUS adjunct to DM between Oct 2019 and Apr 2021. The primary outcome was the sensitivity of ABUS with DM (ABUS/DM) and HHUS with DM (HHUS/DM) for detecting additional ipsilateral or contralateral cancers beyond the index cancer. ABUS/DM was evaluated for noninferiority with a prespecified margin of 5%. Specificity was assessed as a secondary outcome. A total of 659 patients (mean age, 50.5 ± 9 years) were included. Seventy-nine patients (12.0%) had additional cancers: 64 ipsilateral (51 multifocal, 13 multicentric) and 15 contralateral. For index cancers, both ABUS/DM and HHUS/DM achieved 100% detection. ABUS/DM showed noninferior sensitivity to HHUS/DM for additional ipsilateral cancers (71.9% vs. 75.0%; p = 0.617) and higher sensitivity than HHUS/DM for contralateral cancers (86.7% vs. 60.0%; p = 0.046). Specificity did not differ significantly between ABUS/DM and HHUS/DM for additional ipsilateral (97.7% vs. 97.0%; p = 0.317) or contralateral (97.8% vs. 98.3%; p = 0.467) cancers. ABUS/DM showed noninferior sensitivity to HHUS/DM, with similar specificity for detecting additional ipsilateral and contralateral breast cancers in early-stage breast cancer. ABUS may serve as a feasible alternative to HHUS for preoperative staging in early-stage breast cancer. Question Is automated breast ultrasound (ABUS) noninferior to handheld ultrasound (HHUS), each with digital mammography (DM), in preoperative staging of tumor extent in early-stage breast cancer? Findings ABUS/DM demonstrated noninferior sensitivity to HHUS/DM, with similar specificity for detecting additional ipsilateral and contralateral cancers in early-stage breast cancer. Clinical relevance ABUS may serve as a feasible alternative to HHUS for preoperative staging in early-stage breast cancer, particularly in clinical settings where the HHUS use is constrained by limited staffing resources and time restrictions.

Diabetic foot ulcers (DFUs) are chronic, non-healing wounds that affect up to 34% of diabetic patients. DFUs are complicated by infection in nearly 60% of cases and frequently progress to amputation. DFU pathology is characterized by a persistent inflammatory state, impaired angiogenesis, and infection. This creates a complex microenvironment refractory to standard care, with fewer than 20% of DFUs healing within 8 weeks. In this review article, normal and pathophysiological processes of wound healing, current clinical management strategies, and adjunct therapeutics in the clinical pipeline are discussed, followed by recent advances in multifunctional bioengineered platforms. These platforms are categorized into three main systems: hydrogels, electrospun dressings, and 3D-bioprinted constructs, in addition to hybrid fabrication approaches and the integration of low-temperature plasma therapy as emerging multi-targeted strategies. For hydrogels, stimuli-responsive designs that respond to mechanical force, pH, glucose, and excess reactive oxygen species to actively modulate drug release and scaffold behavior are discussed. For electrospun scaffolds, strategies for controlled, multi-therapeutic delivery, including fiber blending, surface conjugation, and core-shell architectures are reviewed. Next, 3D bioprinting as a platform for patient-specific, cell-laden constructs is presented and covers major fabrication techniques and the emerging potential of handheld in situ bioprinters for accelerating clinical translation. Multi-targeted hybrid approaches that combine these platforms, along with the synergistic integration of low-temperature plasma therapy for broad-spectrum antimicrobial action, biofilm disruption, and immune modulation are emphasized. Unlike prior material-centric reviews, this review adopts a function-driven framework that organizes scaffold systems based on their ability to address key DFU pathologies, including infection, inflammation, impaired angiogenesis, and delayed healing, providing a more clinically relevant perspective. Finally, emerging directions such as artificial intelligence (AI)-guided design, in situ bioprinting, and recent clinical trends are discussed to bridge scaffold design with translational application. STATEMENT OF SIGNIFICANCE: Diabetic foot ulcers (DFUs) present a critical global health challenge characterized by a highly inflammatory microenvironment that remains refractory to standard care. This review elucidates the paradigm shift from passive wound dressings to "intelligent," multifunctional bioengineered scaffolds designed to actively modulate DFUs. We critically examine recent advances in stimuli-responsive hydrogels (pH-, glucose-, and reactive oxygen species-sensitive), mechanically active contractile patches, complex electrospun architectures, and 3D bioprinting. Furthermore, by integrating emerging technologies such as handheld in situ 3D bioprinting, low-temperature plasma therapy, and artificial intelligence-driven design, this work provides a roadmap for the next generation of precision biomaterials capable of overcoming specific biological barriers to regeneration in chronic wounds.

This study aimed to compare lower-limb muscle strength and postural sway between individuals with KOA and healthy controls, and to examine the association between strength parameters and postural stability using clinically feasible assessment methods. A cross-sectional analysis was conducted among 90 participants (45 KOA, 45 matched controls). The participants were older adults aged 50-75&#xa0;years, and 40.0% of the KOA group and 42.22% of the control group were male. Participants with knee osteoarthritis were classified as having moderate disease severity based on Kellgren-Lawrence grades II-III. Strength and balance assessments were conducted in a counterbalanced order, with a standardized 10-min rest interval between testing sessions to minimize fatigue effects. Isometric quadriceps and hamstring strength were measured using handheld dynamometry. Postural sway metrics-sway area, sway velocity, and single-leg stance time-were assessed via static posturography. Limb symmetry index (LSI) and quadriceps-to-hamstrings (Q: H) ratio were calculated. Pearson correlation and multiple linear regression analyses examined associations between strength and postural control. KOA participants showed significantly greater sway area and velocity, reduced stance time, and lower quadriceps and hamstrings strength compared to controls (all p < 0.001). Quadriceps strength (&#x3b2; = -0.024, p < 0.001) and LSI (&#x3b2; = -0.062, p = 0.001) were independent predictors of sway area under eyes-closed conditions. LSI and quadriceps strength were strongly correlated with sway parameters and stance performance. Lower limb strength deficits and inter-limb asymmetry significantly contribute to postural instability in individuals with KOA. Objective, clinically feasible tools such as handheld dynamometry and posturography can support evaluation and inform rehabilitation strategies targeting strength and balance.

To compare the feasibility, image quality, acquisition time, and evaluator preference of four smartphone-based fundus imaging (SBFI) systems in dogs and cats using a handheld fundus camera as a reference device. Twenty client-owned animals, including 10 dogs and 10 cats. Fundus images were obtained from 40 eyes (20 mydriatic and 20 non-mydriatic) using four SBFI systems (iExaminer Panoptic, NUN WFE-02S, OQVet, and VistaView) and a handheld infrared fundus camera (Aurora). Image acquisition success and acquisition time were recorded for each device. Image quality was independently graded using a 10-point Likert scale by a masked panel of 32 veterinary ophthalmologists. Evaluators also selected their preferred image among the four SBFI systems. Image acquisition was successful in all animals for all devices under both mydriatic and non-mydriatic conditions (100%). Image quality differed significantly between devices. Among SBFI systems, OQVet achieved the highest median score (7; interquartile range [IQR] 6-8), followed by VistaView (6; IQR 5-8), NUN (5; IQR 4-7), and iExaminer (4; IQR 3-5). OQVet received approximately 50% of evaluator preference votes. The reference camera achieved the highest image quality in 77.5% of imaging sets but was outperformed by OQVet (15%) and VistaView (7.5%) in some examinations. Pupil dilation did not significantly influence image quality or acquisition time. Smartphone-based fundus imaging enables reliable retinal imaging in dogs and cats and represents an accessible option for retinal documentation and screening in veterinary practice.

To characterize the frequency-dependent bioimpedance properties of major ocular tissues in intact ex vivo porcine eyes under simulated surgical conditions and evaluate tissue separability at discrete frequencies. Bioimpedance spectra were acquired from sclera, corneal epithelium, iris, lens, vitreous, and retina in intact ex vivo porcine eyes using a two-electrode probe and a precision LCR meter over 5 kHz to 1 MHz. Measurements were obtained under balanced salt solution and ophthalmic viscosurgical device conditions. Probe-tissue contact was verified by microscope visualization and optical coherence tomography. Tissue separability at 5, 50, 100, and 900 kHz was evaluated using global and pairwise statistical comparisons, effect sizes, and ROC-based separability metrics. Robotic-stabilized and handheld measurements were also compared. Ocular tissues demonstrated distinct, frequency-dependent impedance magnitude distributions. Across sampled frequencies, 60% to 80% of tissue pairs showed significant differences after multiplicity correction. Median pairwise effect sizes ranged from Cohen's d = 0.48 at 5 kHz to 1.04 to 1.06 at 50 to 100 kHz. Median ROC-based separability was 0.91 at 5 kHz and 0.76 to 0.77 at 50 to 900 kHz. Robotic-stabilized measurements showed lower variance than handheld measurements, although tissue-specific impedance ranges and frequency-dependent trends were preserved across acquisition modes. Major ocular tissues exhibit reproducible, frequency-dependent bioimpedance signatures in intact ex vivo eyes under simulated surgical preparation. These findings establish a physiologically relevant ocular impedance reference dataset and support bioimpedance as a complementary modality for tissue differentiation in ophthalmic microsurgery.

Arthroscopic Bankart repair is commonly performed following anterior shoulder instability; however, the presence of a Buford complex introduces unique anatomical and rehabilitative considerations. Evidence guiding postoperative rehabilitation for collision sport athletes with this combined pathology is limited. Additionally, although blood flow restriction (BFR) training has demonstrated benefits for mitigating muscle atrophy during periods of restricted loading, its application in postoperative shoulder rehabilitation remains underexplored. The purpose of this case report is to describe the integration of early-phase BFR within a structured, criterion-based rehabilitation program following arthroscopic Bankart repair in the presence of a Buford complex in a high school collision-sport athlete, and to highlight considerations for return-to-sport testing in this athlete. A 17-year-old male high school football linebacker with recurrent anterior shoulder instability underwent arthroscopic Bankart repair with concomitant management of a Buford complex. Physical therapy began 12 days postoperatively and followed a standardized protocol adapted to sport-specific demands. Due to early postoperative restrictions, BFR was incorporated during protective and intermediate phases to support neuromuscular activation and strength development while respecting tissue-healing constraints. Rehabilitation progressed through phased mobility, strengthening, closed-chain loading, and sport-specific activities using objective criteria to guide advancement. Outcome measures included passive and active range of motion (PROM, AROM), shoulder strength via handheld dynamometry, grip strength, girth measurements, patient-reported outcomes (Western Ontario Shoulder Instability Index [WOSI], Shoulder Instability-Return to Sport After Injury [SIRSI]), and functional performance tests (Posterior Shoulder Endurance Test [PSET], Y Balance Test-Upper Quarter [YBT-UQ], Closed Kinetic Chain Upper Extremity Stability Test [CKCUEST]). The athlete completed 18 sessions over 17 weeks. By postoperative Week 17, the athlete demonstrated full, pain-free passive and active shoulder range of motion. Shoulder flexion strength improved from 21.8 lbs at Week 6 to 30.5 lbs at Week 15, and grip strength improved from 92 lbs to 114 lbs. PSET improved from 54.5 seconds to 135.2 seconds, and CKCUEST performance increased from 15 to 24 touches. WOSI scores improved from 38.6% to 25.7%, and SIRSI scores increased from 74.2% to 80.1%. The athlete returned to full football participation 19 weeks after surgery without reported symptoms. This case report illustrates the feasibility of incorporating early-phase BFR within a criterion-based rehabilitation program following Bankart repair in a collision sport athlete with a Buford complex. Further investigation is warranted regarding standardized upper extremity BFR protocols and return-to-sport testing strategies for contact athletes. 4.

This study aimed to examine variations in hip rotation range of motion (ROM) and strength when measured at 90&#xb0; or 0&#xb0; hip flexion in amateur soccer players, and to assess whether a standardized measurement protocol can be established. Additionally, it assessed the impact of normalizing strength values by body mass and lower leg length, and whether playing position affects hip rotator ROM and strength. A total of 56 amateur soccer players from the Valencian region were tested for maximal voluntary isometric contraction (MVIC) and ROM of hip rotator muscles, measured at both 0&#xb0; and 90&#xb0; hip flexion. ROM was measured using a digital inclinometer and strength was assessed using a handheld dynamometer. Statistical analyses included paired samples t-tests, effect size calculations, Pearson's correlation analysis, and one-way ANOVA to examine variations across playing positions. Significant differences were found in hip internal rotation (IR) ROM at 90&#xb0; vs. 0&#xb0; flexion for both right and left hips (p&#x2009;=&#x2009;0.001 and 0.011, respectively). However, the magnitude of these differences was limited, with higher ROM values at 90&#xb0;, while external rotation (ER) differences were not significant. Strength measurements showed significant increases at 90&#xb0; compared with 0&#xb0; for IR and ER in both right and left hips (IR: p&#x2009;<&#x2009;0.001; ER: p&#x2009;<&#x2009;0.001 and 0.005, respectively). Normalized strength values also revealed significant differences for both variables across angles (p&#x2009;<&#x2009;0.001-0.005). No significant differences were found between ROM and strength values when comparing playing positions. Findings support assessing strength in both positions, whereas ROM differences between positions were of limited magnitude and uncertain clinical relevance. These findings underscore the importance of context-specific assessment protocols. Normalizing strength values may not be necessary in this homogeneous population. No position-dependent differences in hip rotator ROM or strength were detected in this sample, suggesting that screening or normative values may not need to be position-specific for amateur male soccer players.

Current European guidelines for primary atherosclerotic cardiovascular disease (ASCVD) prevention recommend using Systematic Coronary Risk Evaluation 2 (SCORE2) algorithms for risk classification and decision-making. For the Greek population, an updated model - HellenicSCORE II+ - has been developed. This cross-sectional study compared SCORE2 versus HellenicSCORE II+ in detecting preclinical carotid atherosclerosis. Middle-aged (40-69 years) individuals from the general population without ASCVD, were invited to participate on a voluntary basis in screening programs in 3 municipalities of Attica, Greece (2023-2025). Handheld carotid ultrasonography was performed and carotid plaque score (CPS) was calculated by summing points allocated to the number/height of plaques. A total of 965 individuals were analyzed [mean age 57.1&#xb1;8.0 (SD) years, men 43.2%, body mass index 27.6&#xb1;4.7 kg/m2, smokers 27.8%, diabetes 7%, antihypertensive/lipid-lowering drug treatment 41.6%/46.5% respectively, SCORE2 5.2&#xb1;3.4%, HellenicSCORE II+ 3.7&#xb1;2.4%]. Participants classified as low-moderate/high/very-high ASCVD risk were 50.9%/43.3%/5.8% according to SCORE2, 74.4%/23%/2.6% with HellenicSCORE II+ and 55.6%/36.4%/8% with CPS. The agreement between SCORE2 and HellenicSCORE II+ was 67.2% (kappa 0.37, P<0.01), whereas between CPS and SCORE2/HellenicSCORE II+ 57.6%/56.2% (kappa 0.24/0.13, P<0.01 for each, P<0.01 for comparison). Receiver operating characteristic curve analysis demonstrated similar discrimination of SCORE2/HellenicSCORE II+ for detecting carotid atherosclerosis (AUC 0.74, 95% confidence intervals 0.71-0.78 and 0.71, 0.68-0.74 respectively, P=NS for comparison). SCORE2 classified a higher proportion of participants as high/very-high ASCVD risk compared with HellenicSCORE II+. Both models demonstrated moderate discrimination for detecting carotid plaque burden, highlighting the need for carotid imaging in refining ASCVD risk.

Rapid and accurate detection of respiratory viruses is essential for controlling disease transmission and enabling effective public health responses, particularly in resource-limited settings. In this study, we present an electrochemical-sensor-assisted lab-in-a-cartridge (EC-LIC) platform for on-site detection of SARS-CoV-2 featuring a self-contained chemical heating system. The device incorporates rotational and gravity-driven fluid handling along with exothermic heating using calcium oxide and a flameless ration heater to generate controlled temperature gradients. Coupled with a CRISPR-Cas13a-based electrochemical sensor, the system enables direct detection of the SARS-CoV-2 N gene without nucleic acid amplification, achieving high sensitivity and specificity. Integrated with a handheld electrochemical reader, the EC-LIC operates as a fully automated sample-to-answer system, completing the assay within 40&#x202f;min over a wide dynamic range from 1.0&#x202f;&#xd7;&#x202f;10&#xb0; to 1.0&#x202f;&#xd7;&#x202f;105&#x202f;fg/mL with a limit of detection as low as 1.21&#x202f;&#xd7;&#x202f;10-1&#x202f;fg/mL. Clinical validation using samples from 102 individuals (60 positive and 42 negative) demonstrated a sensitivity of 98% and a specificity of 90%. These results establish the EC-LIC as a robust nucleic acid detection platform for rapid clinical screening and early epidemic response.

A dataset of two spectral lighting simulation reference models - one office and one factory hall - is presented. It aims to demonstrate and support full-spectral daylight and electric lighting simulations and facilitate evaluation of non-visual effects of light. The dataset includes Rhino CAD geometry, comprehensive spectral material and light source data and window system BSDF data. Example implementations in the two software tools, Radiance and OWL, enable reproducible workflows and support adoption in other software. The dataset is openly available on Zenodo. The office model reproduces Room 518 at the University of Innsbruck, including a west-facing fa&#xe7;ade and interior furnishings. The factory hall model follows the proposed geometry in the European standard 15193 for building energy performance. Interior reflectances in the office were measured in-situ using a handheld spectrometer. Exterior spectra and factory hall materials matching specified reflectances were obtained from an online spectral materials database. Glazing transmittance was derived from IGDB data using LBNL Optics/WINDOW. BSDFs for venetian blinds at various tilt angles, and for a diffusing pane adapted from the Complex Glazing Database, were generated in WINDOW. Luminaires in both models are specified with photometric files (Eulumdat/IES) and lamp spectra (Fluorescent 840, 4000 K LED). The provided example implementations (Radiance, OWL) include prepared input data and scripts to run first spectral simulations; example results are also included. The dataset is prepared to support reuse by researchers, designers and software developers for method validation, software engineering and comparison, and development of spectral metrics and controls.