Citizen science data reliability enhancing scientific research: insights from an 11-year study in the Mediterranean Sea

Court detection is an important part of sports video analysis, match statistics generation and automated broadcasting systems. Current methodologies often rely on general object detection methods, lacking domain specific knowledge for the correct identification of court borders. In this work, a new deep learning architecture for badminton court detection, CourtKeyNet, was introduced. CourtKeyNet comprises several novel components, including (1) an Octave Feature Extractor for analyzing visual data with various frequency bands, (2) a Polar Transform Attention mechanism to boost the boundary detection capability, (3) a Keypoint Localization module with a hybrid heatmap and regression-based approach to localize precise corners, (4) a Quadrilateral Constraint Module to achieve geometric consistency, (5) a novel Geometric Consistency Loss function. Extensive experiments show that CourtKeyNet has the best results that outperforms the general purpose keypoint detection methods in terms of mean keypoint localization accuracy and court detection Intersection over Union (IoU). • Proposed CourtKeyNet, a neural network tailored for badminton court detection. • Introduced Octave Feature Extractor with multi-frequency domain-specific modules. • Polar Transform Attention mechanism for boundary-focused detection. • Geometric Intelligence is built into the architecture for court detection. • High accuracy in a variety of lighting environment conditions.

Prior studies have demonstrated the feasibility of fluorescently labeled tilmanocept for sentinel lymph node biopsy (SLNB) in the oral cavity. We evaluated the added value of fluorescently labeled tilmanocept in sentinel lymph node biopsy (SLNB) of the oral cavity compared to gamma probe. Healthy male New Zealand white rabbits received oral cavity injections of radioactive (Technetium 99m) and fluorescently (IRDye800) conjugated tilmanocept followed by either fluorescence guided (n = 14) or gamma probe guided (n = 14) SLNB performed 1 h, 48 h, or 5 days postinjection. Duration of the SLNB performed by two individuals was measured and compared using the two methods. Fluorescence guidance resulted in a 1.8-fold reduction in time of SLN removal (median 104 vs. 191 s, p = 0.05). For the fluorescence guided SLNB, 7.1% (1 of 14) had nonsentinel node tissue removed prior to the correct identification of the SLN, whereas gamma probe/radioactivity guided SLNB had 28.6% (4 of 14). When comparing operation time between attending and resident surgeon, time to identification of first SLN was not significantly different for fluorescence guided surgery (82 vs. 107 s, respectively) or with gamma probe (158 vs. 204 s); however, median times using the gamma probe were nearly double for both operators. We additionally show the durability of fluorescence signal up to 5 days and clear visualization of proposed tracer with commercially available intraoperative imagers. The use of fluorescent labeled tilmanocept decreases operative time needed for SLNB as well as reduces the amount of nonsentinel tissue removed. N/A.

Bloodstream infections represent a major medical emergency, in which rapid and accurate pathogen identification is essential for optimizing therapy. This study evaluates the analytical performance of the Autof MS2600 and Biotyper Sirius (Bruker) for the identification of microorganisms directly from positive blood culture. A prospective study was conducted over 7 weeks at the Brussels University Hospital laboratory, including all new episodes of bacteremia. Practical aspects and three preparation protocols were compared: a commercial kit and an in-house pretreatment protocol on the Autof MS2600, and an in-house pretreatment protocol on the Sirius system. A total of 157 bloodstream infection episodes were analyzed, including 64 Gram-negative bacteria, 90 Gram-positive bacteria, and 3 yeasts. Correct species-level identification rates were 62.4% using the Autobio kit, 73.2% using the Autof MS2600 with the in-house protocol, and 72.6% using the Sirius system. In-house protocols performed significantly better than the commercial kit (p < 0.05), with no significant difference between the two in-house protocols (p = 1.000). Performance was higher for Gram-negative bacteria (85.9% with in-house vs. 75.0% with the kit) than for Gram-positive bacteria (66.7% vs. 55.6%). The Sirius system showed the highest non-identification rate (25.5%), compared with the Autof MS2600 (11.5% with the in-house and 16.6% with the kit). The in-house pretreatment was faster than the kit-based protocol (19 vs. 40min). In-house pretreatment protocols significantly improve and accelerate direct microbial identification from positive blood cultures, with comparable performance between platforms. Differences in non-identification rates mainly reflect system-specific scoring thresholds and algorithmic approaches. In this context, the choice of platform may be based on practical considerations, although this should be interpreted in light of the study design and sample size.

Almonds (Prunus dulcis) (Mill.) are an economically important crop in California, with an estimated value of $8.6 billion dollars annually. The aim of this study was to identify the hemipteran pest and beneficial species in conventionally managed almond orchards when the developing almond crop was vulnerable to feeding damage. Insects were collected in 9 orchards by multiple methods and then Hemiptera were sorted to morphospecies. The DNA barcode of the mitochondrial DNA COI gene was sequenced for each specimen. For each species, the haplotype diversity was determined. A high diversity of Hemiptera was collected, including 26 species from 12 families. Pest species included Halyomorpha halys Stål (Hemiptera: Pentatomidae), Chinavia hilaris (Say) (Hemiptera: Pentatomidae), Lygus hesperus Knight, and Leptoglossus zonatus (Dallas) (Hemiptera: Coreidae); beneficials included Brochymena quadripustulata (Fabricius) (Hemiptera: Pentatomidae), Orius spp. Wolff (Hemiptera: Anthocoridae), Nabis spp. Latreille (Hemiptera: Nabidae) and Geocoris atricolor Montandon (Hemiptera: Geocoridae). Several orchards had high species diversity (13 species), others had only one species recovered. Haplotype diversity was low in most pest species except for L. hesperus (0.6) and was high in 2 beneficial groups, Nabis spp. and Orius tristicolor (White) (0.6 to 0.76). Hemipteran pests were detected through the time when developing almonds were susceptible to feeding damage, while beneficial Hemiptera were more abundant later. Correct identification of insect species is an integral component to Integrated Pest Management (IPM) to trace the origin of invasive species, uncover cryptic species, and to develop a relationship between species and crop damage.

Left ventricular hypertrophy (LVH) is a frequent clinical phenotype with diverse etiologies, including hypertensive heart disease, hypertrophic cardiomyopathy, and infiltrative or metabolic disorders. Despite the availability of disease-specific therapies, diagnostic delays and misclassifications remain prevalent. From a health services perspective, these diagnostic gaps are often driven by significant variability in specialist-level diagnostic reasoning and the inconsistent implementation of clinical pathways. To design and evaluate the feasibility of a competency-based curriculum aimed at standardizing specialist-level clinical pathways for LVH-related diseases, and to assess its immediate impact on physician knowledge acquisition, diagnostic reasoning, and clinical confidence. A structured curriculum was developed following principles of patient-centered care and multidisciplinary collaboration. The program, delivered as a continuing medical education (CME) initiative for 45 cardiologists, incorporated guideline reinforcement, genetics-oriented learning, case-based discussions, and pathway design exercises via a blended learning format. Outcomes were assessed through pre- and post-training knowledge tests, structured case-based evaluations, and self-reported confidence and satisfaction surveys. Forty-five cardiologists from secondary and tertiary centers completed the program. Mean knowledge test scores significantly improved from 62% ± 12% pre-training to 86% ± 8% post-training (p < 0.001). The correct identification of rare LVH etiologies (e.g., Fabry disease, cardiac amyloidosis) in structured cases rose from 35% to 78%. Participants reported significantly higher diagnostic confidence for both common and rare etiologies, and overall course satisfaction was high (92% rating ≥ 4/5). This competency-based curriculum was associated with short-term improvements in specialist-level knowledge and diagnostic reasoning for LVH-related diseases. This program shows potential to assist in reducing variability in specialist practice and establishing standardized clinical pathways, serving as a preliminary health services intervention that may enhance downstream referral consistency and inter-provider care coordination. Not applicable.

Cancer incidence in Brazil is projected to reach approximately 781,000 new cases annually in 2026-2028, with persistent regional and social inequalities. School-based interventions may represent an upstream strategy to improve cancer literacy and promote equity-oriented prevention. To evaluate the immediate impact of a structured cancer education intervention on knowledge, beliefs, emotional responses, and health system navigation among Brazilian public school adolescents. Methods A repeated cross-sectional pre-post study was conducted among 8th-grade students from public schools (pre-test n = 910; post-test n = 888). The educational intervention addressed cancer epidemiology, modifiable risk factors, warning signs, stigma, and navigation within the Brazilian Unified Health System (SUS). Differences between assessments were analyzed using Pearson's chi-square test. Effect sizes were estimated using Cramér's V. Significant improvements were observed in knowledge of cancer incidence (39.9% vs. 67.8%, p < 0.001, V = 0.59) and modifiable risk factors, particularly overweight (43.9% vs. 82.9%, V = 0.39), physical inactivity (48.9% vs. 86.6%, V = 0.39), and alcohol consumption (65.1% vs. 91.7%, V = 0.32). Recognition of cancer warning signs increased from 40.5% to 73.6% (p < 0.001, V = 0.33). Correct identification of the Basic Health Unit as the appropriate SUS entry point rose from 10.8% to 37.2% (p < 0.001, V = 0.33). When reflecting on cancer, participants reported a modest increase in negative emotional responses, including anxiety, sadness, and feelings of insecurity, whereas levels of fear remained relatively stable over time. Notably, motivation to engage in discussions about cancer with family members and to adopt healthier lifestyle behaviors increased significantly, suggesting a shift toward greater proactive engagement despite heightened emotional awareness. This school-based intervention produced meaningful short-term improvements in cancer literacy and system-oriented health knowledge. Integrating structured cancer education into public school curricula may contribute to long-term reductions in socially patterned disparities in cancer prevention and access to care.

Purpose/IntroductionProton pump inhibitors (PPIs) are commonly prescribed for stress ulcer prophylaxis (SUP) in non-critically ill hospitalized patients despite limited evidence supporting routine use. This leads to unnecessary medication exposure and possible adverse effects. This study evaluated medical residents’ knowledge of PPI-related adverse effects and the appropriateness of SUP indications, and examined differences across specialties and the relationship between safety knowledge and prescribing decisions.MethodsWe carried out a cross-sectional electronic survey of medical residents enrolled in Saudi Commission for Health Specialties (SCFHS) accredited programs. The questionnaire was sent by the SCFHS, which had nine Likert-scale items assessing PPI risk knowledge and SUP appropriateness. A response rate could not be calculated because the total number of survey recipients was not available. Composite scores were calculated for knowledge, appropriateness, and overall competency. Group comparisons and correlations were analyzed using standard parametric methods.ResultsA total of 119 residents completed the survey. The mean PPI risk knowledge score was 3.31 ± 0.65, with 47.1% achieving competency. The mean SUP appropriateness score was 2.85 ± 0.39, and only 5.9% met the competency threshold. Correct identification of inappropriate SUP use was particularly low for high-dose corticosteroids alone (5.0%), dual antiplatelet therapy alone (10.9%), and mechanical ventilation for less than 24 hours (19.3%). Internal medicine residents demonstrated higher knowledge and overall competency scores than other specialties, while appropriateness scores did not differ by specialty. No significant correlation was observed between knowledge and appropriateness (r = 0.10, p = 0.262).ConclusionResidents demonstrated moderate awareness of PPI-related risks but limited recognition of appropriate SUP indications in non-critically ill patients. These findings suggest that medication safety knowledge alone may not correspond to guideline-concordant prescribing decisions and highlight opportunities to strengthen targeted educational efforts.

Rising nighttime temperature (Tnight) can reduce crop yields while low Tnight may restrict plant growth and development. Despite quantifiable effects of Tnight, genetic basis underlying plant responses to Tnight remains unclear. We investigated natural variation in long-term response of effective photosynthetic efficiency (Fq'/Fm') to Tnight among Arabidopsis accessions. Genome-wide association study (GWAS) was conducted for Fq'/Fm' of the plants grown under 15°C or 20°C Tnight. GWAS revealed highly polygenic architecture of Fq'/Fm', with associated single nucleotide polymorphisms (SNPs) varying across Tnight conditions and measurement days. Notably, 15°C Tnight stabilised the contributions of a subset of SNPs, whereas 20°C Tnight enhanced day-to-day variations in SNP-trait associations. We then incorporated the associated SNPs in genomic prediction (GP) models to assess the improvement of prediction accuracy. The GWAS-derived SNPs significantly improved the prediction ability of GP models, indicating collective influence of numerous small-effect SNPs. Finally, the model predictions were experimentally validated in an independent, genetically diverse population, which confirmed the correct identification of low-Fq'/Fm' accessions in 15°C Tnight. These results uncover the genetic underpinnings of long-term Fq'/Fm' response to cool vs warm nights and establish a framework for leveraging GWAS and GP to explore complex traits, such as photosynthesis, toward breeding climate-resilient crops.

Laser induced breakdown spectroscopy has demonstrated its high potential in material identification in many areas including field exploration, recycling industry and forensic science. A spectral LIBS database was built from more than 500 minerals, alloys, polymers and glass samples of known composition. Using a short and long delay for LIBS analysis, providing an elemental and a molecular fingerprint of each sample a spectral library was built. The unknown spectra are compared with the database. Similarity metrics are evaluated on repetitions and synthetic spectra, a Pearson correlation gave best results to provide accurate identification. Spectral comparison are then tested using mineral, metal (copper, nickel and iron/neodymium alloys) as examples of more and more complex spectra. The accuracy results are compared between similarity test and classification methods (k nearest neighbors, support vector machine and random forest). Results provided are different, for most materials, the developed protocol gave better identification results. The combined use of both molecular and elemental information allowed reducing the matching list to a restricted number of very close sample. It is especially useful in the case of rare earth elements (REE) alloyed with iron, and reliable identification is obtained in every case. • A LIBS spectral database containing more than 500 materials is employed to identify samples. • Spectra recorded to obtain elemental and molecular informations are employed to identify unknown samples. • Evaluation of different spectral similarity metrics led to selection of correlation to compare spectra. • Examples of minerals and alloys identification showed the accuracy of the protocol. • Correct identification is obtained even for complex iron alloys and rare-earth element containing materials.

The high noon view: The value of an additional radiographic projection along the path of penetrating foreign bodies.

Lactose intolerance is common, so accurate detection of lactose in supplements and pharmaceuticals is critical. We validated an ultra-high performance liquid chromatography-tandem mass spectrometry (U-HPLC-MS/MS) method with high sensitivity and specificity for trace lactose in products labeled "lactose-free." The workflow mitigates matrix effects through solid-phase extraction and filtration and explicitly accounts for α/β mutarotation to ensure correct identification and quantification. Validation per ISO and ICH Q2 confirmed robustness and performance superior to enzymatic/colorimetric assays for quality control and regulatory compliance. The method achieved excellent linearity (R2 > 0.995) over 0.05-10ppm, recoveries of 95-105%, precision with RSD < 2%, LOQ 0.05ppm, and matrix effect < 15%. These results enable reliable verification of "lactose-free" claims across diverse matrices, reducing false positives, enhancing reproducibility, and improving labeling transparency.

Introduction: Nasopharyngeal swab collection is considered one of the most sensitive methods for obtaining upper respiratory tract specimens; however, it requires adequate anatomical understanding and proper technique. Inadequate training may lead to patient discomfort, procedural complications, and falsenegative results. Aim: To assess the effectiveness of a modified skull model in improving knowledge, confidence, and technical skills related to nasopharyngeal swab collection among undergraduate medical students. Materials and Methods: The present prospective interventional study was conducted among 75 Phase III MBBS students by the Department of Otorhinolaryngology at Varun Arjun Medical College &amp; Rohilkhand Hospital, Uttar Pradesh, India, a tertiary care teaching hospital in Northern India between November 2024 and mid-February 2025. A cadaveric skull model was modified using silicone, cotton pads, and coloured inks to simulate nasal and nasopharyngeal anatomy. Participants received a structured demonstration followed by supervised hands-on practice until three consecutive successful swabs were achieved. Knowledge, confidence, and procedural skills were assessed before and after training using a structured questionnaire and an observational checklist. Results: The mean knowledge score improved significantly from 1.85±1.06 before training to 8.10±1.20 after training (p&lt;0.0001). Confidence levels increased markedly, with 64 participants (85.3%) reporting high confidence (Likert 4-5) after training. Significant improvement was also observed in procedural skills, including swab direction, insertion depth, and correct identification of the nasopharyngeal sampling site (p&lt;0.0001). Conclusion: The modified skull model proved to be an effective, affordable, and reproducible tool for improving knowledge, confidence, and technical proficiency in nasopharyngeal swab collection among undergraduate medical students.

Misuse of performance-enhancing substances in sport remains a persistent threat to the values of fair competition, with estimated doping prevalence far exceeding adverse analytical findings in routine testing. This highlights the need for improved surveillance tools capable of rapid and simple, on-site screening methods. This proof-of-concept study evaluated atmospheric solids analysis probe-mass spectrometry (ASAP-MS) as a minimally complex, high-throughput, and potentially deployable approach for detecting prohibited substances in black market products. Sixteen pharmaceutical products previously confirmed to contain prohibited substances were blind-analysed using ASAP-MS, alongside eleven third-party batch-tested sports supplements. Resultant multi-channel spectra were interrogated manually for the presence of known precursor and product ions. In addition, a compound library was developed and applied using software-driven matching. At least one prohibited substance was detected in all products without generating false positives in certified supplements. Manual spectral review identified only 68% of individual substances correctly. Initial issues with false positive detection of trenbolone in testosterone ester products using the software-driven approach were iteratively optimised through alteration of match score thresholds. This achieved 96% correct substance identification and 100% detection for an adverse finding. Moreover, investigation of diagnostic precursor ions via heatmaps offered a complementary and additional reporting tool to reduce false negatives. This work demonstrated that ASAP-MS, with tailored software-driven analysis, provides reliable near-real-time qualitative screening of prohibited substances. More broadly, this supports the potential for portable, rapid mass spectrometry-based screening platforms to strengthen forensic intelligence, accelerate anti-doping investigations, and enhance athlete protection by complementing laboratory-based confirmatory testing in sport.

Accurate measurement and reporting of pulmonary arterial wedge pressure (PAWP) and cardiac output (CO) are essential for diagnosing and managing pulmonary hypertension (PH) and pulmonary arterial hypertension (PAH). Errors in these parameters can lead to misclassification and inappropriate treatment. Recent updates from the 2023 European Respiratory Society/European Society of Cardiology guidelines and the 2024 7th World Symposium on Pulmonary Hypertension emphasize the need for standardized right heart catheterization (RHC) practices. To identify knowledge gaps in RHC practices among interprofessional team members and implement a targeted blended learning program developed for pulmonary critical care and cardiology fellows, with a focus on accurate acquisition and interpretation of PAWP and CO. Using the SQUIRE framework, we conducted a baseline survey assessing RHC practices and knowledge among pulmonary critical care and cardiology attending physicians and fellows, and critical care nurses at an academic medical center. Respondents were actively involved in RHC procedures across the cardiac catheterization laboratory, medical intensive care unit, and cardiac care unit. Survey items with < 70% correct response were used to guide development of a 120-minute blended learning program incorporating lectures, video instruction, and hands-on-simulation. Pre- and post-training assessments were conducted. Of 141 invited participants, 85 responded to the survey (60.3% response rate). Baseline data revealed knowledge gaps and therefore opportunities for learning in CO measurement methods and PAWP interpretation. Assessment post-blended learning program demonstrated knowledge improvements in key areas including identification of indirect FicK (iFick) as the most common CO method, confirmation of zero-reference level by the proceduralist-nurse team, acquisition of PAWP, and obtaining at least three measurements of CO via thermodilution. However, correct identification of end-expiratory PAWP declined, highlighting the need for greater emphasis on this skill in future training sessions. This is the first reported interprofessional training initiative to target CO and PAWP measurement accuracy in RHC. This study identified key knowledge gaps and therefore opportunities for learning. Knowledge improvements measured after the educational intervention support the need for ongoing, structured, and interprofessional training to enhance procedural competency and interpretation of RHC hemodynamic data to optimize PH and PAH care.

ABSTRACT Background Gordonia spp. are gram-positive aerobic bacilli whose infections in humans are rare and typically affect immunocompromised hosts and those with indwelling medical devices. Conventional identification systems frequently misidentify Gordonia species. Here, we present a bloodstream infection caused by Gordonia hongkongensis . Case Summary A 76-year-old man with metastatic colorectal adenocarcinoma receiving oral chemotherapy presented with febrile syndrome and acute diarrhea, unresponsive to prior antimicrobial therapy. Blood cultures were positive for gram-positive rods, which grew producing pink-to-orange pigmented colonies. Matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS) and 16S rRNA gene sequencing could not distinguish the isolate at the species level, identifying it as belonging to the genus Gordonia . Owing to the limitations of these methods, whole-genome sequencing (WGS) was performed, which revealed G. hongkongensis as the causative pathogen. Updating the MALDI-TOF MS database with the main spectral profiles enabled the correct identification of the isolate upon reanalysis. Empirical antimicrobial therapy with piperacillin–tazobactam was initiated before organism identification and susceptibility testing. After a 7-day course of treatment, the patient showed clinical improvement and resolution of fever. Conclusion Human infections with G. hongkongensis are rare. This case underscores the diagnostic challenges associated with Gordonia species and emphasizes the role of WGS in accurately identifying emerging pathogens. Increased awareness and improved diagnostic tools are essential for a better understanding and management of Gordonia infections.

Background/Objectives: The human brain is responsible for controlling various physiological functions, and hence, the presence of tumors in the brain is a major concern in the medical field. The correct identification and categorization of tumors in the brain using Magnetic Resonance Imaging (MRI) is a major requirement for the diagnosis and treatment of a tumor. The proposed research will focus on designing a CNN model that is optimized for tumor image classification. Methods: This research proposes an optimized CNN model featuring strategically placed dropout layers and hyperparameter optimization. This study uses a dataset of 640 MRI scans (320 tumor and 320 non-tumor) collected from a private hospital in Saudi Arabia. The proposed method utilizes a learning rate of 0.001 in combination with the Adam optimizer to ensure stable and efficient convergence. Its performance was benchmarked against established architectures, including VGG-19, Inception V3, ResNet-10, and ResNet-50, with evaluation based on classification accuracy and computational cost. Results: The experimental results show that the optimized CNN proposed in this work performs much better than the deeper architectures. The network reached a maximum training accuracy of 97.77% and a final test accuracy of 95.35% with a small test loss of 0.2223. The test accuracy of the optimized VGG-19 and Inception V3 networks was much lower, with a training time per epoch that was several orders of magnitude higher. The validation stability of the proposed network was high (92.25% to 95.35%) during the final stages of training. Conclusions: The conclusion drawn from this study is that hyperparameter optimization and strategic regularization are more advantageous for tumor classification using MRI images than the mere depth of the model. The accuracy of 95.35% with low computational complexity makes this lightweight CNN model a feasible solution for real-time applications.

Introduction: Mass preventive chest imaging examinations (fluorography and X-ray) play a critical role in identifying diseases of public health significance, yet place an additional burden on the healthcare system. Automated sorting of imaging studies using medical devices (MD) powered by artificial intelligence (AI) technologies facilitates optimization by relieving radiologists from having to review studies without pathological findings. However, despite high AI accuracy, rare false-negative findings remain possible, which can be critical in medical screening. Objective: The purpose of this retrospective study was to evaluate an automated dual reading method for autonomous sorting of preventive imaging studies. Materials and methods: The study included 411 clinical cases previously misclassified as «normal.» Each study was reanalyzed by a different AI model. Results: As a result, the re-classification yielded 31.6% correct pathology identification and reduced clinically important discrepancies by 25.5%. When simulating the automated dual reading, the proportion of false-negative findings decreased from 0.071% to 0.052%. Discussion: The simulation revealed significant performance variability across the AI models, emphasizing the need for careful post-authorization monitoring to replace underperforming applications. Conclusion: Thus, automated dual reading improves the safety of autonomous sorting by minimizing the number of missed pathological findings. At the same time, the method requires a strategic approach to AI model selection as it poses a risk of reducing the system’s operating efficiency. The optimal benefit-risk ratio should be determined against contribution of preventive programs to public health interest.

High cardiovascular risk is one of the main determinants of morbidity and mortality in different populations. Systemic arterial hypertension (SAH), a chronic condition with high prevalence in Brazil, has well-established diagnosis and treatment, and the correct identification of its diagnostic criteria is essential for proper risk stratification and clinical management. This study consists of a narrative literature review aimed at synthesizing evidence on SAH, with emphasis on diagnostic methods and their different clinical presentations. The methodology was based on a search of scientific databases, followed by a critical analysis of the literature. The results highlight the relevance of ambulatory blood pressure monitoring (ABPM) and home blood pressure monitoring (HBPM) as fundamental tools for the diagnosis and follow-up of SAH. In addition, resistant hypertension is identified as a complex and often underdiagnosed condition, whose detection can be improved through the proper application of these methods.

Accurate and rapid identification of Cryptococcus spp. is critical for patient care. Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) is increasingly used for yeast identification; however, its performance varies with instrument and database coverage. We compared the performance of three MALDI-TOF MS systems-Bruker Biotyper Sirius (Sirius; Bruker Daltonics), bioMérieux VITEK MS PRIME (PRIME; bioMérieux), and ASTA MicroIDSys (ASTA; ASTA Inc.)-for detecting Cryptococcus spp. across protein-extraction protocols. We analyzed 51 C. neoformans, two C. deuterogattii, five Vanrija humicola, and one Naganishia liquefaciens clinical isolates. Each isolate was tested after on-plate formic acid extraction; isolates failing identification were retested after repeat on-plate and intube extractions. The results were categorized as correct, incomplete, incorrect, or no identification against ribosomal DNA sequencing results. After initial on-plate extraction, correct identification rates were 86.4%, 88.1%, and 69.5% for Sirius, PRIME, and ASTA, respectively. After repeat on-plate extraction, ASTA accuracy increased to 81.4% (P =0.016). Additional in-tube extraction did not significantly improve identification across platforms, although PRIME outperformed ASTA (P =0.012). All platforms reliably classified C. neoformans (correct identification rates: 94%-96%) after additional on-plate extraction; discrepancies occurred mainly in uncommon species, possibly owing to gaps in database coverage. Sirius and PRIME demonstrated robust performance for C. neoformans identification and C. gattii complex differentiation, outperforming ASTA for relatively rare species. Single on-plate extraction is sufficient in clinical practice. However, continued database expansion is essential for maximizing MALDI-TOF MS performance and accurately identifying Cryptococcus spp., thereby informing antifungal therapy and improving patient outcomes.