Inter-operator Variability Research Articles

Screening of liver cancer with abbreviated MRI.

Current recommendations for the surveillance of HCC are based on the semiannual liver ultrasound (with or without serum alpha-fetoprotein) in patients with cirrhosis and in subgroups with chronic hepatitis B infection. However, the sensitivity of this strategy is suboptimal for the detection of early-stage tumors, especially in obese patients, due to interoperator variability and poor adherence. The detection rate of focal liver lesions is excellent with MRI, making it the best alternative candidate for surveillance. However, performing a full contrast-enhanced MRI is unrealistic because of limited availability and health economics. Abbreviated MRI (AMRI) corresponds to the acquisition of a limited number of sequences with a high detection rate. The theoretical benefits of AMRI are a reduced acquisition time (≤10 min) with improved time-effectiveness and cost-effectiveness compared with conventional MRI, and greater accuracy than ultrasound. Numerous protocols may be performed, including T1-weighted, T2-weighted, and DWI sequences, with or without contrast administration. Although published studies report promising per-patient results, they should be interpreted with caution. Indeed, most studies were simulated, retrospectively reviewing a subset of sequences in relatively small populations who underwent a full MRI. They also included groups that were not representative of screening populations. In addition, most were published by Asian groups, with at-risk populations that were different from Western populations. There are no existing longitudinal studies that directly compare the different AMRI approaches or AMRI to ultrasound. Finally, it is possible that 1 approach will not fit all patients and that strategies should be tailored to the risk of HCC, in particular in relation to the cost and availability of AMRI. Several trials are ongoing to evaluate these questions.

On-site verification of the CBEU method using UriSelect4 chromogenic medium at the National Institute of Hygiene (INH) of Lomé.

The Cytobacteriological Examination of Urine (CBEU) is one of the most requested microbiological analyses in the medical bacteriology laboratory. According to the ISO 15189 standard, Medical Biology Laboratory (MBL) must validate the techniques before their use. It is in this perspective that, within the Medical Bacteriology Laboratory of the INH of Lome, we have carried out the on-site verification of the CBEU method using UriSelect4 chromogenic medium at the INH of Lomé. The biological material was composed of reference bacterial strains and clinical bacterial strains isolated at INH. Failure Modes Effects and Criticality Analysis (FMECA) with 5​M model was used to perform risk analysis and performances as repeatability, intermediate fidelity/reproducibility, sensitivity and inter-operator variability were evaluated using references strains and different operators. "Comité Français d'Accréditation" (COFRAC) SH GTA 04 was used as reference. Single-factor ANOVA was used to analyze the data. Operating procedures, Patients preparation, Samples and Preparation of culture media were identified as main critical points with a criticality index of 8, 9, 9 and 12 respectively. The use of uriselect4 to perform ECBU was shown to be repeatable and reproducible. The sensibility of Uriselect4 to detect urinary infection was 100% with negligible inter-operator variability. This study performed on-site verification of Uriselect4 at the bacteriology lab of INH of Lome and identified some critical points to master. The overall performance criteria from COFRAC SH GTA 04 were conform.

Democratization of deep learning for segmenting cartilage from MRIs of human knees: Application to data from the osteoarthritis initiative.

In this study, we aimed to democratize access to convolutional neural networks (CNN) for segmenting cartilage volumes, generating state-of-the-art results for specialized, real-world applications in hospitals and research. Segmentation of cross-sectional and/or longitudinal magnetic resonance (MR) imagesof articular cartilage facilitates both clinical management of joint damage/disease and fundamental research. Manual delineation of such images is a time-consuming task susceptible to high intra- and interoperator variability and prone to errors. Thus, enabling reliable and efficient analyses of MRIs of cartilage requires automated segmentation of cartilage volumes. Two main limitations arise in the development of hospital- or population-specific deep learning (DL) models for image segmentation: specialized knowledge and specialized hardware. We present a relatively easy and accessible implementation of a DL model to automatically segment MRIs of human knees with state-of-the-art accuracy. In representative examples, we trained CNN models in 6-8 hand obtained results quantitatively comparable to state-of-the-art for every anatomical structure. We established and evaluated our methods using two publicly available MRI data sets originating from the Osteoarthritis Initiative, Stryker Imorphics, and Zuse Institute Berlin (ZIB), as representative test cases. We use Google Colabfor editing and adapting the Python codes and selecting the runtime environment leveraging high-performance graphical processing units. We designed our solution for novice users to apply to any data set with relatively few adaptations requiring only basic programming skills. To facilitate the adoption of our methods, we provide a complete guideline for using our methods and software, as well as the software tools themselves.Clinical significance:We establish and detail methods that clinical personal can apply to create their own DLmodels without specialized knowledge of DL nor specialized hardware/infrastructure and obtain results comparable with the state-of-the-art to facilitate both clinical management of joint damage/disease and fundamental research.

Inter- and intra-operator variations in manual segmentation of hippocampus from MRI

PurposeWe evaluate inter- and intra-operator variations in manual segmentation of hippocampus and present their potential sources. Hippocampal atrophy is common in mesial temporal lobe epilepsy (mTLE). Effective diagnosis and treatment of mTLE depends on accurate and efficient segmentation of hippocampus from magnetic resonance imaging (MRI) data. Manual segmentation by expert radiologists remains the gold standard, although automated segmentation methods exist. MethodsHippocampus was segmented in MRI of 118 unilateral mTLE patients and 25 non-epileptic subjects (65 males, 78 females; mean age 39 years) by three operators (M1, M2, M3) manually and by three software tools (FreeSurfer, LocalInfo, ABSS) automatically. Segmentation results were evaluated using 7 volume-, voxel-, and distance-based performance measures. Inter-operator variation was evaluated by comparing the segmentation results of the three operators. Intra-operator variation was evaluated by comparing manual and automatic segmentation results. Segmentation results were used to lateralize epileptogenicity in mTLE patients. ResultsRanking of performance measures differed when using M3 segmentations as ground truth rather than M1 or M2 segmentations. The standard deviation tended to be higher when using M3 as ground truth and M1 or M2 as test segmentation. Variation in performance measures and increased standard deviation when using M3 as ground truth are indicative of inter-operator variability. Standard deviations were larger when using M2 segmentations as ground truth and automated segmentations as test segmentation, rather than M1 or M3. Large standard deviations between M2 segmentations and automated segmentations are indicative of intra-operator variation. Among automated segmentation methods, ABSS produced segmentations most similar to manual segmentations, but FreeSurfer and LocalInfo lateralized epileptogenicity more accurately. ConclusionsPart of inter-operator variation might be due to temporal separation of M3 segmentations from M1 and M2 segmentations (M3 performed segmentations a few years before operators M1 and M2). Inter-operator variation will likely reduce if all operators segment within the same time frame, reducing discrepancies in training of operators. Intra-operator variation can likely be mitigated with additional oversight by a neuroradiologist. Inter- and intra-operator variability may generate inconsistencies in outcomes. Future automated segmentation techniques may integrate neural-network-based (ABSS) and atlas-based (FreeSurfer and LocalInfo) segmentation techniques for optimal performance.

Effect of pre-analytical variables on Raman and FTIR spectral content of lymphocytes.

The use of Fourier transform infrared (FTIR) and Raman spectroscopy (RS) for the analysis of lymphocytes in clinical applications is increasing in the field of biomedicine. The pre-analytical phase, which is the most vulnerable stage of the testing process, is where most errors and sample variance occur; however, it is unclear how pre-analytical variables affect the FTIR and Raman spectra of lymphocytes. In this study, we evaluated how pre-analytical procedures undertaken before spectroscopic analysis influence the spectral integrity of lymphocytes purified from the peripheral blood of male volunteers (n = 3). Pre-analytical variables investigated were associated with (i) sample preparation, (blood collection systems, anticoagulant, needle gauges), (ii) sample storage (fresh or frozen), and (iii) sample processing (inter-operator variability, time to lymphocyte isolation). Although many of these procedural pre-analytical variables did not alter the spectral signature of the lymphocytes, evidence of spectral effects due to the freeze-thaw cycle, in vitro culture inter-operator variability and the time to lymphocyte isolation was observed. Although FTIR and RS possess clinical potential, their translation into a clinical environment is impeded by a lack of standardisation and harmonisation of protocols related to the preparation, storage, and processing of samples, which hinders uniform, accurate, and reproducible analysis. Therefore, further development of protocols is required to successfully integrate these techniques into current clinical workflows.

Automated tumor segmentation and brain tissue extraction from multiparametric MRI of pediatric brain tumors: A multi-institutional study.

Brain tumors are the most common solid tumors and the leading cause of cancer-related death among all childhood cancers. Tumor segmentation is essential in surgical and treatment planning, and response assessment and monitoring. However, manual segmentation is time-consuming and has high interoperator variability. We present a multi-institutional deep learning-based method for automated brain extraction and segmentation of pediatric brain tumors based on multi-parametric MRI scans. Multi-parametric scans (T1w, T1w-CE, T2, and T2-FLAIR) of 244 pediatric patients ( n = 215 internal and n = 29 external cohorts) with de novo brain tumors, including a variety of tumor subtypes, were preprocessed and manually segmented to identify the brain tissue and tumor subregions into four tumor subregions, i.e., enhancing tumor (ET), non-enhancing tumor (NET), cystic components (CC), and peritumoral edema (ED). The internal cohort was split into training ( n = 151), validation ( n = 43), and withheld internal test ( n = 21) subsets. DeepMedic, a three-dimensional convolutional neural network, was trained and the model parameters were tuned. Finally, the network was evaluated on the withheld internal and external test cohorts. Dice similarity score (median ± SD) was 0.91 ± 0.10/0.88 ± 0.16 for the whole tumor, 0.73 ± 0.27/0.84 ± 0.29 for ET, 0.79 ± 19/0.74 ± 0.27 for union of all non-enhancing components (i.e., NET, CC, ED), and 0.98 ± 0.02 for brain tissue in both internal/external test sets. Our proposed automated brain extraction and tumor subregion segmentation models demonstrated accurate performance on segmentation of the brain tissue and whole tumor regions in pediatric brain tumors and can facilitate detection of abnormal regions for further clinical measurements.

Le malattie infiammatorie croniche intestinali: aspetti clinici

Inflammatory bowel diseases (IBD) are chronic diseases of the gastrointestinal tract that have been steadily increasing in recent years, particularly in pediatric age. The clinical manifestations of IBD differ according to the forms (Crohn’s m., ulcerative rectocolitis, undifferentiated colitis), and can be distinguished into intestinal and extra-intestinal. The laboratory contributes at various levels in the diagnosis of IBD, from the use of inflammation markers, to the most commonly used serologic markers (ASCA and p-ANCA), to fecal markers (calprotectin). Ultrasonography of the bowel loops is the most recommended noninvasive instrumental investigation to date because of its noninvasiveness, safety, and low cost. The major limitation of ultrasonography is operator-dependent variability, although recently, the use of new technique, SICUS (small intestine contrast ultrasonography), has been shown to greatly improve the sensitivity of the examination and reduce inter-operator variability. Endoscopic investigation coupled with histologic analysis of biopsy specimens is, to date, the gold standard for making a diagnosis of IBD, playing a key role in the diagnosis and therapy of these conditions. The primary goals of pediatric IBD therapy are to induce and maintain a sustained remission with prevention of clinical relapse, optimize growth and pubertal development, and improve the patient’s quality of life. The importance of the genetic and epigenetic component in the phenotypic expression of IBD is increasingly widely demonstrated. These aspects will be the subject of a forthcoming contribution.

Effect of Segmentation Uncertainty on the ECGI Inverse Problem Solution and Source Localization.

Electrocardiographic Imaging (ECGI) is a promising tool to non-invasively map the electrical activity of the heart using body surface potentials (BSPs) and the patient specific anatomical data. One of the first steps of ECGI is the segmentation of the heart and torso geometries. In the clinical practice, the segmentation procedure is not fully-automated yet and is in consequence operator-dependent. We expect that the inter-operator variation in cardiac segmentation would influence the ECGI solution. This effect remains however non quantified. In the present work, we study the effect of segmentation variability on the ECGI estimation of the cardiac activity with 262 shape models generated from fifteen different segmentations. Therefore, we designed two test cases: with and without shape model uncertainty. Moreover, we used four cases for ectopic ventricular excitation and compared the ECGI results in terms of reconstructed activation times and excitation origins. The preliminary results indicate that a small variation of the activation maps can be observed with a model uncertainty but no significant effect on the source localization is observed.

CAL-Tutor: A HoloLens 2 Application for Training in Obstetric Sonography and User Motion Data Recording.

Obstetric ultrasound (US) training teaches the relationship between foetal anatomy and the viewed US slice to enable navigation to standardised anatomical planes (head, abdomen and femur) where diagnostic measurements are taken. This process is difficult to learn, and results in considerable inter-operator variability. We propose the CAL-Tutor system for US training based on a US scanner and phantom, where a model of both the baby and the US slice are displayed to the trainee in its physical location using the HoloLens 2. The intention is that AR guidance will shorten the learning curve for US trainees and improve spatial awareness. In addition to the AR guidance, we also record many data streams to assess user motion and the learning process. The HoloLens 2 provides eye gaze, head and hand position, ARToolkit and NDI Aurora tracking gives the US probe positions and an external camera records the overall scene. These data can provide a rich source for further analysis, such as distinguishing expert from novice motion. We have demonstrated the system in a sample of engineers. Feedback suggests that the system helps novice users navigate the US probe to the standard plane. The data capture is successful and initial data visualisations show that meaningful information about user behaviour can be captured. Initial feedback is encouraging and shows improved user assessment where AR guidance is provided.

Deconstructing body axis morphogenesis in zebrafish embryos using robot-assisted tissue micromanipulation

Classic microsurgical techniques, such as those used in the early 1900s by Mangold and Spemann, have been instrumental in advancing our understanding of embryonic development. However, these techniques are highly specialized, leading to issues of inter-operator variability. Here we introduce a user-friendly robotic microsurgery platform that allows precise mechanical manipulation of soft tissues in zebrafish embryos. Using our platform, we reproducibly targeted precise regions of tail explants, and quantified the response in real-time by following notochord and presomitic mesoderm (PSM) morphogenesis and segmentation clock dynamics during vertebrate anteroposterior axis elongation. We find an extension force generated through the posterior notochord that is strong enough to buckle the structure. Our data suggest that this force generates a unidirectional notochord extension towards the tailbud because PSM tissue around the posterior notochord does not let it slide anteriorly. These results complement existing biomechanical models of axis elongation, revealing a critical coupling between the posterior notochord, the tailbud, and the PSM, and show that somite patterning is robust against structural perturbations.

Intra- and inter-operator variability in MRI-based manual segmentation of HCC lesions and its impact on dosimetry

PurposeThe aim was to quantify inter- and intra-observer variability in manually delineated hepatocellular carcinoma (HCC) lesion contours and the resulting impact on radioembolization (RE) dosimetry.MethodsTen patients with HCC lesions treated with Y-90 RE and imaged with post-therapy Y-90 PET/CT were selected for retrospective analysis. Three radiologists contoured 20 lesions manually on baseline multiphase contrast-enhanced MRIs, and two of the radiologists re-contoured at two additional sessions. Contours were transferred to co-registered PET/CT-based Y-90 dose maps. Volume-dependent recovery coefficients were applied for partial volume correction (PVC) when reporting mean absorbed dose. To understand how uncertainty varies with tumor size, we fit power models regressing relative uncertainty in volume and in mean absorbed dose on contour volume. Finally, we determined effects of segmentation uncertainty on tumor control probability (TCP), as calculated using logistic models developed in a previous RE study.ResultsThe average lesion volume ranged from 1.8 to 194.5 mL, and the mean absorbed dose ranged from 23.4 to 1629.0 Gy. The mean inter-observer Dice coefficient for lesion contours was significantly less than the mean intra-observer Dice coefficient (0.79 vs. 0.85, p < 0.001). Uncertainty in segmented volume, as measured by the Coefficient of Variation (CV), ranged from 4.2 to 34.7% with an average of 17.2%. The CV in mean absorbed dose had an average value of 5.4% (range 1.2–13.1%) without PVC while it was 15.1% (range 1.5–55.2%) with PVC. Using the fitted models for uncertainty as a function of volume on our prior data, the mean change in TCP due to segmentation uncertainty alone was estimated as 16.2% (maximum 48.5%).ConclusionsThough we find relatively high inter- and intra-observer reliability overall, uncertainty in tumor contouring propagates into non-negligible uncertainty in dose metrics and outcome prediction for individual cases that should be considered in dosimetry-guided treatment.

Effects of post-mortem inspection techniques change on the detection capability of low public health impact diseases of slaughtered pigs: A quasi-experimental study.

Slaughtered animals are regularly submitted to post-mortem inspection to ensure that all the edible parts are fit for human consumption. According to Regulation (EU) No 219/2014, pig carcasses inspection is exclusively visual as palpation and incision could lead to cross-contamination and spread of relevant zoonotic agents. However, when compared to incision and palpation, the visual method is characterized by low sensitivity; thus, the omission of incision and palpation could lead to a reduced detection capability of organic lesions. This study aims to assess the effectiveness of exclusively visual inspection to mark pulmonary and hepatic lesions associated with low public health impact diseases in pork carcasses. A quasiexperimental- before/after research protocol has been used. All the post-mortem inspections have been carried out in a slaughterhouse located in the province of Teramo (IT), on 7,764 swine from 2011 to 2017. Carcasses undergone the only visual inspection have shown a statistically significant reduction (pvalue <0.0001) in the diagnosis of hepatic (decrease of 59%) and pulmonary diseases (decrease of 38. 5%). To overcome the limits of the low sensitivity of the visual inspection, as well as the inter-operator diagnostic variability, the high number of carcasses examined is proposed as a factor conferring external validity to the study, which provides quantitative evidence in support of the causal association between the modified inspection technique and the reduced diagnostic capacity. A further support derives from the assessment of the prevalence of hepatic and pulmonary diseases in species for whom the inspection technique is not changed.

Reproducibility of Biomolecular Corona Experiments: A Primer for Reliable Results

AbstractThe biomolecular corona is a key component controlling the identity of nanomaterials in physiological environments. Studies aimed at identifying factors shaping the biomolecular corona have proliferated in the last decade but have been performed by research groups with different backgrounds. Efforts made within the scientific community to guarantee the reproducibility of experimental data have identified protocol standardization as an indispensable step for advancing knowledge in this arena. To contribute to fulfill this gap, here the relevance of interoperator variability in biomolecular corona studies and the benefits arising from automated systems usage are explored. Moreover, the role of molecular crowding during nanoparticle‐biofluid incubation and the effect of washing the pellet during corona isolation are thoroughly investigated. It is believed that the findings will help researchers enhance the accuracy of experimental design and reporting.

Hybrid convolutional neural networks with SVM classifier for classification of skin cancer

Background: The dermatologist widely uses digital dermoscopy for the detection of melanoma. The accurate detection of melanoma by clinicians is subjective and further depends on their experience. Fully automated computer-aided diagnosis systems are necessary to eliminate the inter-operator variability inherent in the personal analysis of dermoscopy images.Objective: Automated skin lesion classification is challenging because of the fine-grained difference in the appearance of these lesions on the skin surface. Deep convolutional neural network (CNN) have shown great separability across many fine-grained object classes.Methods: This article presents two novel hybrid CNN models with an SVM classifier at the output layer for classifying dermoscopy images into either benign or melanoma lesions. The features extracted by the first CNN and second CNN models are concatenated and fed to the SVM classifier for classification. The labels obtained from an expert dermatologist are used as a reference to evaluate the performance of the proposed model.Results: The proposed models displayed better results over the state-of-the-art CNN models on the publicly available ISBI 2016 dataset. The proposed models achieved 88.02% and 87.43% accuracy, which remain higher than the traditional CNN models.Conclusions: The proposed framework could attain considerable improvement in accuracy to classify dermoscopy images.

Comparing Variability in Measurement of Subcutaneous Tumors in Mice Using 3D Thermal Imaging and Calipers.

Repeatable tumor measurements are key to accurately assessing tumor growth and treatment efficacy. A preliminary study that we conducted showed that a novel 3D and thermal imaging system (3D-TI) for measuring subcutaneous tumors in rodents significantly reduced interoperator variability across 3 in vivo efficacy studies. Here we further studied this reduction in interoperator variability across a much larger dataset. A dataset consisting of 6,532 paired 3D-TI and caliper interoperator measurements was obtained from tumor scans and measurements in 27 laboratories across 289 studies, 153 operators, over 20 mouse strains, and 100 cell lines. Interoperator variability in both measurement methods was analyzed using coefficient of variation (CV), intraclass correlation (ICC) analysis, and significance testing. The median 3D-TI CV was significantly lower than the median caliper CV. The effects of large interoperator variability at critical points in the study were also investigated. At stratified randomization, changing the operator performing caliper measurements resulted in a 59% probability that a mouse would be reassigned to a different group. The probability that this would occur when using 3D-TI was significantly lower at 29%. In studies in which a tumor was expected to regress, changing the operator during the study was associated with a tumor volume increase of approximately 500mm³ when using calipers. This change did not occur when using 3D-TI. We conclude that 3D-TI significantly reduces interoperator variability as compared with calipers and can improve reproducibility of in vivo studies across a wide range of mouse strains and cell lines.

An ultra-low-dose protocol for computed tomography-guided lung radiofrequency ablations

To evaluate the safety and efficacy of ultra-low-dose (ULD) protocol for computed tomography (CT)-guided lung radiofrequency ablation (RFA). Patients who had undergone lung RFA between November 2017 and January 2021 were consecutively and retrospectively included. Thirty patients were treated using a conventional standard protocol (SP), including helical acquisitions with mA automatic adjustment and sequential CT at 80 kVp; and 31, with a ULD protocol defined with helical acquisitions with fixed mA and sequential series at 100 kVp. These parameters were selected from those used for a diagnostic lung low-dose CT scanner. Patient characteristics, dose indicators, technical efficacy (minimal margin [MM], recurrence during follow-up), and complications (pneumothorax, alveolar haemorrhage, and haemoptysis) were recorded. We included 61 patients (median age, 65 [54–73] and 33 women), with no significant differences according to the type of protocol, except for the type of anaesthesia. Even if the number of helical acquisitions did not significantly change, all dose indicators significantly decreased by 1.5-fold–3-fold. The median dose-length-product and effective dose, with their ranges, respectively, were 465 mGy cm (315–554) and 6.5 mSv (4.4–7.8) in the SP group versus 178 mGy cm (154–267) and 2.5 mSv (2.2–3.7) in the ULD group, (p < 001). The ULD group exhibited lower intraoperator variability and better interoperator alignment than those of the SP group. The MM was not significantly different between the two groups (4.6 mm versus 5 mm, p = 16). One local recurrence was observed in each group at 8 months in the SP and at one year in the ULD group (p = 1). The complication rates did not differ significantly. Implementing an ULD protocol during lung RFA may provide similar efficacy, a reduction of dose indicators, and intra- and interoperator variability, without increasing complication rates, compared to those associated with an SP.

18F]FDG PET/CT in Large Vessel Vasculitis: The Impact of Expertise and Confounders on Image Analysis.

Diagnosis of vasculitis is challenging. To avoid invasive approaches, clinical guidelines recommend the use of diagnostic imaging. This study aimed at evaluating the diagnostic accuracy of [18F]-fluorodeoxyglucose ([18F]FDG) position emission tomography/computed tomography (PET/CT) in large vessel vasculitis (LVV) and how this is affected by inter-operator variability. A total of 279 patients who performed [18F]-FDG PET/CT for suspicion of LVV were retrospectively analyzed. We tested the qualitative and semi-quantitative analysis and parameters influencing image quality and interpretation. Exams were evaluated by two readers with different experience and their performance was compared. LVV diagnosis was confirmed in 81 patients. [18F]-FDG PET/CT accuracy was 73% and 67% for the expert reader and less experienced reader, respectively. The expert reader overall performed better than the less experienced one, with higher accuracy in patients with normal BMI (77.3 vs. 63.8%), normal level of glycemia (73.3 vs. 65%), younger age (76.6 vs. 68.2%), and when no therapy was in course at time of imaging (76.7 vs. 66.7%). The diagnostic performance of both readers did not improve using semi-quantitative parameters. We confirmed the appropriateness of the recommended criteria for image acquisition and interpretation, underlining the importance of experience in image interpretation for the optimal diagnostic performance of [18F]FDG PET/CT in vasculitis.

Ab.ai - A Novel Automated AI Tool for Reporting Antibiograms.

Anti-Microbial Resistance is one of the greatest threats that mankind faces right now due to the inappropriate use of antibiotics. Institution of appropriate antibiotics in right dose for the right patient at right time is the "gamechanger" in fighting AMR. Antibiotic Sensitivity Testing (AST) or antibiogram is done to ascertain the sensitivity profile of the organism. The most widely used method in laboratory practice in India is the Kirby-Bauer's disk diffusion test. There are few shortcomings in the manual interpretation of antibiograms in the form of high inter-operator variability, mandatory requirement of trained microbiologists - which is difficult in low-resource settings and high degree of interpersonal bias due to various factors like stress, workload, and visual acuity. We propose the Ab.ai tool for automating the AST procedures in laboratory. The Ab.ai tool comprises of 3 phases: first for data collection, second for data processing and the third for generation of antibiotic sensitivity reports. Various software packages like OpenCV and EasyOCR are used for the development of the Ab.ai tool. A total of 50 antibiograms of both GPC and GNB are interpreted both by manual and automated method. The manual method is considered the "gold-standard" and the performance of Ab.ai tool was compared against the manual method. The Ab.ai tool achieved an agreement of 98.4% on susceptibility categorization of GPC antibiotics and 97.6% on that of GNB antibiotics against the gold standard manual method. The proposed Ab.ai tool serves as a perfect candidate for automating AST procedures and would prove to be a "game-changer" in battling AMR.

Intra- and inter-operator variability of refractometric total proteins measurement of canine plasma.

Refractometric total proteins are commonly used in practice as a quick and inexpensive way to measure total protein concentration in bodily fluids. Little information is available about how the operator performing the measurement affects the results. The aim of our study was to determine the inter- and intra-operator variability of refractometric total proteins measured on canine plasma using a temperature-compensated handheld refractometer. A pooled sample of canine lithium-heparin plasma was created using leftover samples from dogs presented to our hospital. The sample was then divided into three aliquots. Total proteins of these aliquots were measured by veterinary nurses, interns, residents and specialists working at our hospital. Statistical analysis revealed excellent inter-operator (ICC 0.99, CI 95%: 0.971-1.00) and intra-operator (ICC 0.997, CI 95%: 0.990- 0.999) variability. Having different operators measuring refractometric total plasma proteins in practice should not affect the results. This suggests different operators can be used when monitoring total plasma proteins of a patient over time and when designing a study that involves this test.

Can machine learning of post-procedural cone-beam CT images in acute ischemic stroke improve the detection of 24-h hemorrhagic transformation? A preliminary study.

Hemorrhagic transformation (HT) is an independent predictor of unfavorable outcome in acute ischemic stroke (AIS) patients undergoing endovascular thrombectomy (EVT). Its early identification could help tailor AIS management. We hypothesize that machine learning (ML) applied to cone-beam computed tomography (CB-CT), immediately after EVT, improves performance in 24-h HT prediction. We prospectively enrolled AIS patients undergoing EVT, post-procedural CB-CT, and 24-h non-contrast CT (NCCT). Three raters independently analyzed imaging at four anatomic levels qualitatively and quantitatively selecting a region of interest (ROI) < 5 mm2. Each ROI was labeled as "hemorrhagic" or "non-hemorrhagic" depending on 24-h NCCT. For each level of CB-CT, Mean Hounsfield Unit (HU), minimum HU, maximum HU, and signal- and contrast-to-noise ratios were calculated, and the differential HU-ROI value was compared between both hemispheres. The number of anatomic levels affected was computed for lesion volume estimation. ML with the best validation performance for 24-h HT prediction was selected. One hundred seventy-two ROIs from affected hemispheres of 43 patients were extracted. Ninety-two ROIs were classified as unremarkable, whereas 5 as parenchymal contrast staining, 29 as ischemia, 7 as subarachnoid hemorrhages, and 39 as HT. The Bernoulli Naïve Bayes was the best ML classifier with a good performance for 24-h HT prediction (sensitivity = 1.00; specificity = 0.75; accuracy = 0.82), though precision was 0.60. ML demonstrates high-sensitivity but low-accuracy 24-h HT prediction in AIS. The automated CB-CT imaging evaluation resizes sensitivity, specificity, and accuracy rates of visual interpretation reported in the literature so far. A standardized quantitative interpretation of CB-CT may be warranted to overcome the inter-operator variability.