Interobserver Differences Research Articles

Assessment of Inter-Observer Variability, Cardiac Cycle and Gender Influence on Carotid Intima-Media Thickness Measured Using Ultrasound in Healthy Adults

Background : Carotid intema-media thickness (cIMT) is a widely utilized noninvasive marker for assessing early atheroscelorosis and cardiovascular risk. cIMT is commonly measured using B-mode ultrasound; however, several factors can introduce variability, including the cardiac cycle phase (systole vs. diastole) and interobserver differences. This study aims to evaluate the inter-observer variability in cIMT measurements and to determine differences in cIMT across genders, cardiac cycle phases, and contralateral carotid arteries. Methods: cIMT measurements were obtained using B-mode ultrasound from 50 healthy adults (25 males, 25 females) bilaterally during both systolic and diastolic phases. Two independent operators performed the measurements to assess inter-observer variability using the intraclass correlation coefficient (ICC) and Bland-Altman analysis. Gender differences, cardiac cycle phase effects, and differences between contralateral sides were analyzed using paired t-tests. Results: A total of 200 cIMT measurements were analyzed. Inter-observer agreement was moderate (ICC = 0.68) with a small bias (-0.02 ± 0.05 mm). Males exhibited significantly higher cIMT than females (mean difference = 0.08 mm, 95% CI 0.07–0.10, p < 0.001). No significant differences were found in cIMT between systolic and diastolic phases (p = 0.51) or between right and left carotid arteries (p = 0.70). Conclusion: The study demonstrates moderate inter-observer variability in cIMT measurements using B-mode ultrasound. Gender significantly influences cIMT, with males having thicker cIMT compared to females. However, no significant differences were observed between cardiac cycle phases or between contralateral sides of the carotid arteries.

Canine medial retropharyngeal lymph node measurements on T2 spin-echo sequences at 3T

IntroductionThe objective of this study is to estimate reference values for medial retropharyngeal lymph nodes (MRLNs) measured in high-field (3T) MRI studies of the canine head/brain using transverse T2 spin-echo images and to determine if dogs with structural brain disease exhibit medial retropharyngeal lymph nodes that are larger than expected from estimated reference values.MethodsThe study population comprises 142 MRLNs from 71 dogs with no evidence of structural brain disease and normal CSF evaluation and 116 MRLNs from 58 dogs with structural brain disease confirmed by histopathology as of infectious or neoplastic origin, or to represent meningoencephalitis of unknown etiology.ResultsBased on this sample, MRLNs are expected to measure 2.9–12.4 mm in maximum short-axis transverse diameter. Interobserver measurement differences are ~1 mm in 95% of the sampled subjects. Lymph node size is correlated with body weight (R = 0.47–0.52) and age (R = −0.39 – −0.47).DiscussionNo difference was found between the lymph node size of dogs with structural brain disease of any type, or overall, compared to that of dogs without structural brain disease.

Interobserver and intraobserver reliability of skull base angles measured on magnetic resonance images.

Determination of platybasia and basilar kyphosis are significant parts of routine cranial magnetic resonance (MR) interpretations. These situations may explain a patient's symptoms or may be associated with other anomalies. To indicate the interobserver and intraobserver reliability of the skull base angles (SBA) (Koenigsberg standard) and modified SBA (mSBA) measurement techniques. In total, 391 patients who had undergone cranial MR imaging were re-assessed regarding the SBA measurements. The SBA and mSBA techniques were used on MR images. Two reviewers independently measured the same angles twice within a 15-day interval, using different monitors. Intraclass correlation coefficient (ICC) was calculated to reveal the intraobserver and interobserver agreements. There was an excellent agreement between reviewers regarding both angle measurements (ICC was 0.998 for SBA and mSBA). Excellent agreement levels were also observed for intraobserver measurements. ICC was 0.998 for SBA and 0.999 for mSBA for reviewer 1. ICC was 0.997 for SBA and 0.999 for mSBA according to the measurement results of reviewer 2. Higher SBA and mSBA values were observed for females compared to males. There was no correlation between SBA and age for SBA. However, a negative and low-level correlation was observed between mSBA values and age for both reviewers. SBA and mSBA measurements indicated excellent agreement regarding interobserver and intraobserver differences. The study results showed that SBA angles were reliable measurement techniques to be used on MR images.

Quantitative modeling of lenticulostriate arteries on 7-T TOF-MRA for cerebral small vessel disease.

We developed a framework for segmenting and modeling lenticulostriate arteries (LSAs) on 7-T time-of-flight magnetic resonance angiography and tested its performance on cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) patients and controls. We prospectively included 29 CADASIL patients and 21 controls. The framework includes a small-patch convolutional neural network (SP-CNN) for fine segmentation, a random forest for modeling LSAs, and a screening model for removing wrong branches. The segmentation performance of our SP-CNN was compared to competitive networks. External validation with different resolution was performed on ten patients with aneurysms. Dice similarity coefficient (DSC) and Hausdorff distance (HD) between each network and manual segmentation were calculated. The modeling results of the centerlines, diameters, and lengths of LSAs were compared against manual labeling by four neurologists. The SP-CNN achieved higher DSC (92.741 ± 2.789, mean ± standard deviation) and lower HD (0.610 ± 0.141 mm) in the segmentation of LSAs. It also outperformed competitive networks in the external validation (DSC 82.6 ± 5.5, HD 0.829 ± 0.143 mm). The framework versus manual difference was lower than the manual inter-observer difference for the vessel length of primary branches (median -0.040 mm, interquartile range -0.209 to 0.059 mm) and secondary branches (0.202 mm, 0.016-0.537 mm), as well as for the offset of centerlines of primary branches (0.071 mm, 0.065-0.078 mm) and secondary branches (0.072, 0.064-0.080 mm), with p < 0.001 for all comparisons. Our framework for LSAs modeling/quantification demonstrated high reliability and accuracy when compared to manual labeling. NCT05902039 ( https://clinicaltrials.gov/study/NCT05902039?cond=NCT05902039 ). The proposed automatic segmentation and modeling framework offers precise quantification of the morphological parameters of lenticulostriate arteries. This innovative technology streamlines diagnosis and research of cerebral small vessel disease, eliminating the burden of manual labeling, facilitating cohort studies and clinical diagnosis. The morphology of LSAs is important in the diagnosis of CSVD but difficult to quantify. The proposed algorithm achieved the performance equivalent to manual labeling by neurologists. Our method can provide standardized quantitative results, reducing radiologists' workload in cohort studies.

Assessing the inter-observer and intra-observer reliability of radiographic measurements for size-specific dose estimates

BackgroundCalculating size-specific dose estimates (SSDEs) requires measurement of the patient’s anteroposterior (AP) and lateral thickness based on computed tomography (CT) images. However, these measurements can be subject to variation due to inter-observer and intra-observer differences. This study aimed to investigate the impact of these variations on the accuracy of the calculated SSDE.MethodsFour radiographers with 1–10 years of experience were invited to measure the AP and lateral thickness on 30 chest, abdomen, and pelvic CT images. The images were sourced from an internet-based database and anonymized for analysis. The observers were trained to perform the measurements using MicroDicom software and asked to repeat the measurements 1 week later. The study was approved by the institutional review board at Taibah University, and written informed consent was obtained from the observers. Statistical analyses were performed using Python libraries Pingouin (version 0.5.3), Seaborn (version 0.12.2), and Matplotlib (version 3.7.1).ResultsThe study revealed excellent inter-observer agreement for the calculated effective diameter and AP thickness measurements, with Intraclass correlation coefficients (ICC) values of 0.95 and 0.96, respectively. The agreement for lateral thickness measurements was lower, with an ICC value of 0.89. The second round of measurements yielded nearly the same levels of inter-observer agreement, with ICC values of 0.97 for the effective diameter, 1.0 for AP thickness, and 0.88 for lateral thickness. When the consistency of the observer was examined, excellent consistency was found for the calculated effective diameter, with ICC values ranging from 0.91 to 1.0 for all observers. This was observed despite the lower consistency in the lateral thickness measurements, which had ICC values ranging from 0.78 to 1.0.ConclusionsThe study’s findings suggest that the measurements required for calculating SSDEs are robust to inter-observer and intra-observer differences. This is important for the clinical use of SSDEs to set diagnostic reference levels for CT scans.

P-747 Can patient self-scans using a home vaginal ultrasound device be used to monitor the ovarian stimulation in ART?

Abstract Study question Can patient self-scans with a home ultrasound device, with or without professional remote online guidance, be used for ovarian and endometrial monitoring during for ART? Summary answer Follicle number/size and endometrial thickness recorded from patients’ home ultrasound self-acquired scans, especially when guided online, corresponded with standard sonographies and aspirated oocyte yield. What is known already Follicle number and size are a key parameter for monitoring the ovarian stimulation and determining the time for triggering and OPU. The endometrial thickness (EMT) and pattern are important parameters associated with implantation and pregnancy. Both are measured repeatedly in serial sonographies during the ART cycles, imposing a significant burden of time, travel and cost on patients, while adding to the clinic’s workload. Patient self-scans with a home ultrasound device can potentially relief patient travel discomfort and the clinic’s patient congestion. Study design, size, duration Open label study performed during one IVF cycle. Participants were trained to visualize uterus and ovaries using the Pulsenmore FC device. After training the patients were randomized to two groups: independent use (IU-scans) or remote online guidance by a clinical sonographer (CG-scans). Before each in-clinic scan patients obtained and recorded their pelvic scan, which were read blindly and compared to the in-clinic scans for ovaries and uterus visualization, follicles number and size, and EMT. Participants/materials, setting, methods Patients without Mullerian or pelvic pathologies undergoing stimulation for IVF. Clinical decisions and triggering were based on individual considerations and the standard in-clinic ultrasound measurements. OPU was performed 36-38h after triggering. All ultrasound scans were analyzed for image quality, number and size of follicles and EMT and compared to the in-clinic scans as the gold standard (GS). Correlation was calculated based on the Spearman Correlation Factor (SCF). Patient satisfaction was measured using a questionnaire. Main results and the role of chance Twenty-four patients completed the study, 11-CG and 13-IU. Image quality score was&amp;gt;3 and adequate uterine visualization was achieved in 100% of the CG-scans and in 87% of the IU-scans. The right ovary was imaged in 100% of the CG-scans and 85.7% of the IU-scans. The left ovary in 97.4% of the CG-scans and 85.7% of the IU-scans. CG-scans AFC correlated better with the GS than IU-scans AFC (SCF 0.85vs.0.12). Correct identification of the first follicle&amp;gt;14mm was achieved in 100% of the CG-scans and in 92% of the IU-patients. Maximal EMT from CG-scans was better correlated with the GS than the IU-scans (SCF 0.64vs.0.46), with a clinically insignificant difference (0.4mm). The number of stimulated follicles obtained from CG-scans was better correlated with the GS than IU-scans (SCF 0.94 vs.0.43). The oocytes/stimulated follicles ratio was 1.19 for the GS and 1.28 for the self scans (1.06 for CG-scans, 1.9 for IU-scans). The MII/follicles&amp;gt;14mm ratio was 1.12 for the GS and 1.49 for the self scans (1.27 for CG-scans,1.69 for IU-scans). Patients’ questionnaire revealed that 89% of patients felt comfortable to self-scan themselfs, 86.5% were interested to use the device and 75% will recommend it to their friends. Limitations, reasons for caution Self vaginal ultrasound scanning is not suitable for all patients and online support by a sonographer is required. Inter-observer differences in scan measurements should be taken into consideration. Since participation in the study was limited to one cycle, no patient learning curve for IU-scans could be evaluated. Wider implications of the findings Reliable folliculometry and endometrial measurements can be obtained from patient self-use vaginal ultrasound scans. Online assistance enhances scanning accuracy. Together with near/home lab tests, safe remote management of ovarian stimulation can be realized, reducing costs, discomfort and travel time, ultimately reducing workload for fertility clinics, and improving workforce flexibility. Trial registration number NCT05485623

The strength of the patellar and Achilles tendon reflexes is not affected in uncomplicated pregnancy.

Research on healthy pregnant women's reflexes is almost unheard of, even if an estimation of reflexes is a conventional part of the physical examination of preeclamptic patients, and hyperreflexia is generally considered to be a warning sign of eclampsia. The aim of the present study was to investigate the possible impact of gestation on tendon reflexes and the Babinski response, as examined in a clinical bed-side manner. A prospective, single-blind, cross-sectional study. Three study groups comprised non-pregnant (n = 20), 14-18 weeks (n = 16), and 34-38 weeks (n = 15) healthy pregnant women respectively. Two experienced neurologists examined separately each participant's patellar reflexes, Achilles reflexes, and Babinski response and rated them on two different scales, National Institute of Neurological Disorders and Stroke (NINDS) scale and Mayo Clinic Scale for Tendon Reflex Assessment. Inter-observer estimations between the patient groups and the neurologists were made by mixed effect model methodology. The patellar and Achilles reflexes´ strengths were rated similarly in all three groups (p > 0.05). The inter-observer difference was non-significant between the neurologists´ estimations (p > 0.05). The neurologists estimated the Babinski response in a few cases as neutral (0), otherwise down (normal). The patellar and Achilles reflex strengths are not affected in uncomplicated pregnancy. There is good concordance between neurologists´ estimations of tendon reflex strengths rated on the NINDS and Mayo scales.

A new bicornuate model of rat uterus transplantation.

Uterus transplantation has revolutionized reproductive medicine for women with absolute uterine factor infertility, resulting in more than 40 reported successful live births worldwide to date. Small animal models are pivotal to refine this surgical and immunological challenging procedure aiming to enhance safety for both the mother and the child. We established a syngeneic bicornuate uterus transplantation model in young female Lewis rats. All surgical procedures were conducted by an experienced and skilled microsurgeon who organized the learning process into multiple structured steps. Animals underwent meticulous preoperative preparation and postoperative care. Transplant success was monitored by sequential biopsies, monitoring graft viability and documenting histological changes long-term. Bicornuate uterus transplantation were successfully established achieving an over 70% graft survival rate with the passage of time. The bicornuate model demonstrated safety and feasibility, yielding outcomes comparable to the unicornuate model in terms of ischemia times and complications. Longitudinal biopsies were well-tolerated, enabling comprehensive monitoring throughout the study. Our novel bicornuate rat uterus transplantation model provides a distinctive opportunity for sequential biopsies at various intervals after transplantation and, therefore, comprehensive monitoring of graft health, viability, and identification of potential signs of rejection. Furthermore, this model allows for different interventions in each horn for comparative studies without interobserver differences contrary to the established unicornuate model. By closely replicating the clinical setting, this model stands as a valuable tool for ongoing research in the field of uterus transplantation, promoting further innovation and deeper insights into the intricacies of the uterus transplant procedure.

Tear meniscus height agreement and reproducibility between two corneal topographers and spectral-domain optical coherence tomography

ABSTRACT Clinical relevance Tear meniscus height (TMH) is an important clinical marker in dry eye diagnosis and management. Purpose To evaluate the reproducibility and agreement of TMH measurements in non-clinical participants using the Oculus Keratograph 5 M, Medmont Meridia, and Spectral-domain optical coherence tomography (Spectralis SD-OCT). Methods Fifty-six participants (mean 43.8 ± 22.4 years) were recruited for this cross-sectional study. Image acquisitions were performed on the three devices, sequentially and randomized. The repeatability and reproducibility of inter-observer and inter-device analysis were performed. Repeated measures ANOVA and Bland-Altman Plots were used to evaluate the agreement between devices. Results The mean TMH with the Oculus Keratograph 5 M, Medmont Meridia and Spectralis SD-OCT were 0.29 ± 0.16 mm, 0.24 ± 0.09 mm and 0.27 ± 0.16 mm, respectively. There were no significant inter-observer differences (paired t-tests, p < 0.001). All the devices exhibited good inter-observer reliability (ICC ≥ 0.877), and good repeatability (CV ≤ 16.53%). Inter-device reliability is moderate (ICC = 0.621, p < 0.001). Repeated measures ANOVA revealed that TMH measurements given by the Spectralis SD-OCT are not significantly different from the Oculus Keratograph 5 M (p = 0.19) and the Medmont Meridia (p = 0.38). TMH measurements from Oculus Keratograph 5 M were significantly higher than those from Medmont Meridia (p = 0.02). Correlations between the mean TMH and the difference in the TMH measurements were positive for Oculus Keratograph 5 M and Medmont Meridia (r2 = 0.62, p < 0.001), negative for Medmont Meridia and Spectralis SD-OCT (r2 = -0.59, p < 0.001), and not significant for Oculus Keratograph 5 M and Spectralis SD-OCT (r2 = 0.05, p = 0.74). A strong correlation was found for TMH measured with all devices (r2 = 0.55 to 0.81, p < 0.001). Conclusions The Oculus Keratograph 5 M, Medmont Meridia, and Spectralis SD-OCT provide reliable and reproducible inter-observer TMH measurements. Inter-device reliability is moderate, with a close correlation between Spectralis SD-OCT and the Oculus Keratograph 5 M. Oculus Keratograph 5 M and Medmont Meridia are repeatable devices appropriate for the measurement of TMH, but they are not interchangeable in clinical practice.

Automated Landmark Annotation for Morphometric Analysis of Distal Femur and Proximal Tibia.

Manual anatomical landmarking for morphometric knee bone characterization in orthopedics is highly time-consuming and shows high operator variability. Therefore, automation could be a substantial improvement for diagnostics and personalized treatments relying on landmark-based methods. Applications include implant sizing and planning, meniscal allograft sizing, and morphological risk factor assessment. For twenty MRI-based 3D bone and cartilage models, anatomical landmarks were manually applied by three experts, and morphometric measurements for 3D characterization of the distal femur and proximal tibia were calculated from all observations. One expert performed the landmark annotations three times. Intra- and inter-observer variations were assessed for landmark position and measurements. The mean of the three expert annotations served as the ground truth. Next, automated landmark annotation was performed by elastic deformation of a template shape, followed by landmark optimization at extreme positions (highest/lowest/most medial/lateral point). The results of our automated annotation method were compared with ground truth, and percentages of landmarks and measurements adhering to different tolerances were calculated. Reliability was evaluated by the intraclass correlation coefficient (ICC). For the manual annotations, the inter-observer absolute difference was 1.53 ± 1.22 mm (mean ± SD) for the landmark positions and 0.56 ± 0.55 mm (mean ± SD) for the morphometric measurements. Automated versus manual landmark extraction differed by an average of 2.05 mm. The automated measurements demonstrated an absolute difference of 0.78 ± 0.60 mm (mean ± SD) from their manual counterparts. Overall, 92% of the automated landmarks were within 4 mm of the expert mean position, and 95% of all morphometric measurements were within 2 mm of the expert mean measurements. The ICC (manual versus automated) for automated morphometric measurements was between 0.926 and 1. Manual annotations required on average 18 min of operator interaction time, while automated annotations only needed 7 min of operator-independent computing time. Considering the time consumption and variability among observers, there is a clear need for a more efficient, standardized, and operator-independent algorithm. Our automated method demonstrated excellent accuracy and reliability for landmark positioning and morphometric measurements. Above all, this automated method will lead to a faster, scalable, and operator-independent morphometric analysis of the knee.

Effect of the ‘Pilates stance’ and Pilates-based matwork training on measurements of height, waist circumference, and interscapular distance

BackgroundStandardised guidelines for stance are used to improve interobserver reliability in anthropometric measurements in clinical practice. A key feature of the stance in Pilates is the ‘drawing in and up’ of the abdomen. The aim of this study was to study the impact of the Pilates stance on height, waist circumference and interscapular distance, compared to that recommended in clinical practice. Methods48 healthy females (median age 60 years) were assessed before and after 10-week Pilates-based matwork training. One Pilates expert and one novice took independent measurements of weight, height, waist circumference and interscapular distance (ISD). ResultsPilates stance, compared to Normal, increased height by up to 2.7 cm and decreased waist up to 5.2 cm (each P < 0.001, repeated measures ANOVA). ISD decreased up to 14 mm (P < 0.001) and this decrease was greater after training (P < 0.001). After controlling for age and length of time learning Pilates, greater baseline ISD predicted a greater change in ISD after the intervention. Effect of Pilates stance was greater when the expert took the measurements (each P ≤ 0.001). ConclusionsActivation of trunk muscles in the Pilates stance increases height and decreases waist circumference, compared to the stance recommended in UK healthcare settings. A decrease in ISD was observed, which was greater after a Pilates-based matwork programme. There are significant inter-observer differences, therefore current clinical guidelines for stance are recommended for repeated anthropometry. The value of the Pilates stance in improving posture and the role of ISD as a marker, should be further studied in various contexts, including clinical settings.

DeepPGSegNet: MRI-based pituitary gland segmentation using deep learning.

In clinical research on pituitary disorders, pituitary gland (PG) segmentation plays a pivotal role, which impacts the diagnosis and treatment of conditions such as endocrine dysfunctions and visual impairments. Manual segmentation, which is the traditional method, is tedious and susceptible to inter-observer differences. Thus, this study introduces an automated solution, utilizing deep learning, for PG segmentation from magnetic resonance imaging (MRI). A total of 153 university students were enrolled, and their MRI images were used to build a training dataset and ground truth data through manual segmentation of the PGs. A model was trained employing data augmentation and a three-dimensional U-Net architecture with a five-fold cross-validation. A predefined field of view was applied to highlight the PG region to optimize memory usage. The model's performance was tested on an independent dataset. The model's performance was tested on an independent dataset for evaluating accuracy, precision, recall, and an F1 score. The model achieved a training accuracy, precision, recall, and an F1 score of 92.7%, 0.87, 0.91, and 0.89, respectively. Moreover, the study explored the relationship between PG morphology and age using the model. The results indicated a significant association between PG volume and midsagittal area with age. These findings suggest that a precise volumetric PG analysis through an automated segmentation can greatly enhance diagnostic accuracy and surveillance of pituitary disorders.

Understanding overall shoulder function and health: the value of specific quantitative vs. qualitative shoulder range of motion on patient-reported outcome measures

Hypothesis and BackgroundIt is known that, though widely used, shoulder range of motion (ROM) measurements are not standardized and have a high rate of intra- and interobserver differences. Particularly, the inconsistency in quantitative and qualitative measurements and their relationship to patient-reported outcome measures (PROMs) make shoulder health difficult to determine. MethodsThis was a prospective study of 147 patients who presented with a chief complaint of shoulder pain to the orthopedic sports medicine and shoulder clinic of a single fellowship-trained surgeon. Measured by one examiner, quantitative ROM measurements were taken with a goniometer and qualitative ROM measured by the anatomic level that the patient could reach. The following PROMs were used as well: American Shoulder and Elbow Surgeons Shoulder Score (ASES), Single Assessment Numeric Evaluation (SANE), Shoulder Pain and Disability Index (SPADI), Oxford Shoulder Score (OSS), Disabilities of the Arm, Shoulder, and Hand questionnaire (DASH), 12-Item Short Form Health Survey (SF-12), and Patient-Reported Outcomes Measurement Information System pain interference short form 6a (PROMIS 6a). Statistical analysis was performed with SPSS using Pearson correlation and two-sample t-test. Benjamini-Hochberg correction was used to determine p-value at which statistical significance was reached to correct for multiple comparisons. ResultsQualitative internal rotation (IR), (the hand behind back reach test) and qualitative forward flexion (FF) correlated with all goniometer angle measurements and PROMs (both shoulder and general health scores). These qualitative measures proved to be an appropriate proxy for IR and FF goniometer measurements. Qualitative external rotation (ER) was not a good substitute for quantitative ER measurement. Quantitative ER correlated with all PROMs. As ROM increased when measured by quantitative ER, qualitative IR, and qualitative FF, shoulder and general health PROMs incrementally increased as well. Discussion/ConclusionQualitative IR measurement, the hand-behind-back reach test, is an adequate substitution for IR goniometer angle as well as a strong representation of global shoulder ROM, shoulder health, and general health while factoring in patient age. Qualitative FF measurement is also an appropriate proxy for quantitative FF and represents global shoulder and general health without factoring in age. Quantitative ER via goniometer angle, is a better assessment of the shoulder than qualitative ER and is a representation of overall shoulder and general health. We recommend the use of quantitative ER, qualitative IR, and qualitative FF measurements to best understand a patient’s overall shoulder health and its impact on their overall health.

BRAFV600E immunohistochemistry can reliably substitute BRAF molecular testing in the Lynch syndrome screening algorithm in colorectal cancer.

The Lynch syndrome (LS) screening algorithm requires BRAF testing as a fundamental step to distinguish sporadic from LS-associated colorectal carcinomas (CRC). BRAF testing by immunohistochemistry (IHC) has shown variable results in the literature. Our aim was to analyse concordance between BRAFV600E IHC and BRAF molecular analysis in a large, mono-institutional CRC whole-slide, case series with laboratory validation. MisMatch repair (MMR) protein (hMLH1, hPMS2, hMSH2, and hMSH6) and BRAFV600E IHC were performed on all unselected cases of surgically resected CRCs (2018-2023). An in-house validation study for BRAFV600E IHC was performed in order to obtain optimal IHC stains. BRAFVV600E IHC was considered negative (score 0), positive (scores 2-3), and equivocal (score 1). Interobserver differences in BRAFV600E IHC scoring were noted in the first 150 cases prospectively collected. Nine-hundred and ninety CRCs cases (830 proficient (p)MMR/160 deficient (d)MMR) were included and all cases performed BRAFV600E IHC (BRAFV600E IHC-positive 13.5% of all series; 66.3% dMMR cases; 3.4% pMMR cases), while 333 also went to BRAF mutation analysis. Optimal agreement in IHC scoring between pathologists (P < 0.0001) was seen; concordance between BRAFV600E IHC and BRAF molecular analysis was extremely high (sensitivity 99.1%, specificity 99.5%; PPV 99.1%, and NPV 99.5%). Discordant cases were reevaluated; 1 score 3 + IHC/wildtype case was an interpretation error and one score 0 IHC/mutated case was related to heterogenous BRAFV600E IHC expression. Among the 12 IHC-equivocal score 1+ cases (which require BRAF molecular analysis), three were BRAF-mutated and nine BRAF-wildtype. BRAFV600E IHC can be used as a reliable surrogate of molecular testing after stringent in-house validation.

Influence of threshold selection and image sequence in in-vivo segmentation of enlarged perivascular spaces

BackgroundGrowing interest surrounds perivascular spaces (PVS) as a clinical biomarker of brain dysfunction given their association with cerebrovascular risk factors and disease. Neuroimaging techniques allowing quick and reliable quantification are being developed, but, in practice, they require optimisation as their limits of validity are usually unspecified. New methodWe evaluate modifications and alternatives to a state-of-the-art (SOTA) PVS segmentation method that uses a vesselness filter to enhance PVS discrimination, followed by thresholding of its response, applied to brain magnetic resonance images (MRI) from patients with sporadic small vessel disease acquired at 3 T. ResultsThe method is robust against inter-observer differences in threshold selection, but separate thresholds for each region of interest (i.e., basal ganglia, centrum semiovale, and midbrain) are required. Noise needs to be assessed prior to selecting these thresholds, as effect of noise and imaging artefacts can be mitigated with a careful optimisation of these thresholds. PVS segmentation from T1-weighted images alone, misses small PVS, therefore, underestimates PVS count, may overestimate individual PVS volume especially in the basal ganglia, and is susceptible to the inclusion of calcified vessels and mineral deposits. Visual analyses indicated the incomplete and fragmented detection of long and thin PVS as the primary cause of errors, with the Frangi filter coping better than the Jerman filter. Comparison with existing methodsLimits of validity to a SOTA PVS segmentation method applied to 3 T MRI with confounding pathology are given. ConclusionsEvidence presented reinforces the STRIVE-2 recommendation of using T2-weighted images for PVS assessment wherever possible. The Frangi filter is recommended for PVS segmentation from MRI, offering robust output against variations in threshold selection and pathology presentation.

BORS/BJR TRAVELLING FELLOWSHIP ABSTRACT: MOTION CAPTURE OF NEONATAL INFANT KICKING MOVEMENTS CAN PROVIDE AN EARLY PREDICTION OF CEREBRAL PALSY

AbstractObjectivesNeonatal motor development transitions from initially spontaneous to later increasingly complex voluntary movements. A delay in transitioning may indicate cerebral palsy (CP). The general movement optimality score (GMOS) evaluates infant movement variety and is used to diagnose CP, but depends on specialized physiotherapists, is time-consuming, and is subject to inter-observer differences. We hypothesised that an objective means of quantifying movements in young infants using motion tracking data may provide a more consistent early diagnosis of CP and reduce the burden on healthcare systems. This study assessed lower limb kinematic and muscle force variances during neonatal infant kicking movements, and determined that movement variances were associated with GMOS scores, and therefore CP.MethodsElectromagnetic motion tracking data (Polhemus) was collected from neonatal infants performing kicking movements (min 50° knee extension-flexion, &lt;2 seconds) in the supine position over 7 minutes. Tracking data from lower limb anatomical landmarks (midfoot inferior, lateral malleolus, lateral knee epicondyle, ASIS, sacrum) were applied to subject-scaled musculoskeletal models (Gait2354_simbody, OpenSim). Inverse kinematics and static optimisation were applied to estimate lower limb kinematics (knee flexion, hip flexion, hip adduction) and muscle forces (quadriceps femoris, biceps femoris) for isolated kicks. Functional principal component analysis (fPCA) was carried out to reduce kicking kinematic and muscle force waveforms to PC scores capturing ‘modes’ of variance. GMOS scores (lower scores = reduced variety of movement) were collected in parallel with motion capture by a trained operator and specialised physiotherapist. Pearson's correlations were performed to assess if the standard deviation (SD) of kinematic and muscle force waveform PC scores, representing the intra-subject variance of movement or muscle activation, were associated with the GMOS scores.ResultsThe study compared GMOS scores, kinematics, and muscle force variances from a total of 26 infants with a mean corrected gestational age of 39.7 (±3.34) weeks and GMOS scores between 21 and 40. There was a significant association between the SD of the PC scores for knee flexion and the GMOS scores (PC1: R = 0.59, p = 0.002; PC2: R = 0.49, p = 0.011; PC3: R = 0.56, p = 0.003). The three PCs captured variances of the overall flexion magnitude (66% variance explained), early-to-late kick knee extension (20%), and continual to biphasic kicking (6%). For hip flexion, only the SD of PC1 correlated with GMOS scores (PC1: R = 0.52, p = 0.0068), which captured the variance of the overall flexion magnitude (81%). For the biceps femoris, the SD of PC1 and PC3 associated with GMOS scores (PC1: R = 0.50, p = 0.002; PC3: R = 0.45, p = 0.03), which captured the variance of the overall bicep force magnitude (79%) and early-to-late kick bicep activation (8%).ConclusionsInfants with reduced motor development as scored in the GMOS displayed reduced variances of knee and hip flexion and biceps femoris activation across kicking cycles. These findings suggest that combining objectively measured movement variances with existing classification methods could facilitate the development of more consistent and accurate diagnostic tools for early detection of CP.Declaration of Interest(b) declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported:I declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research project.

AI-Based Foot X-Ray Reading in Real-World: Evaluating the Accuracy of Assistive Decisions for Diagnosing Foot &amp; Ankle Disorders

Category: Midfoot/Forefoot; Hindfoot Introduction/Purpose: Evaluation of plain radiographs is of fundamental importance in the diagnosis and treatment planning of foot and ankle disorders. These assessments are essential, but can be difficult for the following reasons: First, many angles must be measured using various reference lines. Second, there may be inter-observer differences in the measurements. Consistent and standardized measurement methods are required to increase the efficiency and accuracy of radiographic readings. Several recent studies have reported good results using deep learning (DL) models for medical image analysis. It is expected that efficiency and accuracy can be increased through radiographic readings using DL. This study aims to develop a DL model that recognizes the bones of the foot and ankle in plain radiographs and automatically measures angles frequently used for diagnosis. Methods: The DL model was developed using two architectures: segmenting objects by locations (SOLO) for bone segmentation and Global Universal U-Net (GU2Net) for reference line detection. The angles measured were hallux valgus angle (HVA) and 1st- 2nd intermetatarsal angle (IMA) on anterior-posterior (AP) X-ray; calcaneal pitch angle (CPA) and lateral Meary's angle (LMA) on lateral X-ray. Only data obtained from external test set were used for statistical analysis to evaluate the model's performance. The test set used a total of 558 images collected from January 2022 to September 2022. Angles measured by the DL model were compared with angles measured manually by human researchers. The intraclass correlation coefficient (ICC) was used to evaluate the accuracy of angle measurements. The error between the angle measured by the DL model and the angle measured by human researchers was evaluated using root mean square error (RMSE) and mean absolute difference (MAE). Results: As shown in the attached table, the ICCs of HVA were 0.957 (right) and 0.933 (left). The ICCs of IMA were 0.881 (right) and 0.902 (left). The ICCs of LMA were 0.895 (right) and 0.937 (left). The ICCs of CPA were 0.918 (right) and 0.868 (left). All ICCs were greater than 0.8, indicating excellent correlations between angles measured with the DL model and angles measured manually (ground truth). For RMSE and MAE, the left IMA was the smallest (1.50 and 1.16, respectively), and the right LMA was the largest (4.50 and 2.47, respectively). Conclusion: The DL model in our study was able to perform various angle measurements in both AP and lateral X-rays with considerable accuracy. Through additional model training and development, it is expected that it will be useful for analyzing various foot and ankle disorders such as hallux valgus, flatfoot, pes cavus and other alignment deformities.

Evaluation of Interobserver and Intraobserver Differences of Graf Method in Developmental Hip Dysplasia

Evaluation of Interobserver and Intraobserver Differences of Graf Method in Developmental Hip Dysplasia

Dosimetric Quality of Artificial Intelligence Based Organ at Risk Segmentation

to compare dosimetric parameters between Artificial intelligence (AI) generated organ at risks (OAR) and Radiation Oncologist approved OARs and evaluation of appropriateness unedited AI- OARs in routine clinical plan optimization and evaluation. The OARs (lung, spinal cord and heart) for twenty SBRT (stereotactic body radiotherapy) lung CT simulation datasets were derived by AI based segmentation algorithms. These AI- OARs were edited by a staff Radiation Oncologist and then subjected to our SBRT peer-review process at our institution. A SBRT plan based on the approved contours was created. Dosimetric parameters for the unedited AI-OARs and edited physician-approved OARs were then compared. Lung V20 differences between AI- OAR and physician- OAR varied from 0.01% - 0.7% with a mean value of 0.1% difference (p-value 0.004). Spinal cord D0.03cc varied from 0.02 Gy - 0.9 Gy with a mean value of 0.3 Gy difference (p-value 0.002). Heart D0.03cc varied from 0.01 Gy - 4.3 Gy with mean value 0.9 Gy difference (p-value 0.02). Dosimetric parameters for AI-based lung, spinal cord and heart OARs vs physician approved OARs were different, overall, the differences were generally small. These differences are likely on par with inter-observer differences seen between individual radiation oncologists. Unedited OARs have the promise for routine use in plan optimization and evaluation to further improve efficiency.

Accuracy and reproducibility analysis of different reference axes for femoral prosthesis rotation alignment in TKA based on 3D CT femoral model

BackgroundThere are many reference axes to determine the rotational positioning of the femoral prosthesis in total knee arthroplasty (TKA), mainly including the surgical transepicondylar axis (sTEA), anatomical transepicondylar axis (aTEA), Whiteside line, and the posterior condylar line (PCL), etc., but there is still no definite conclusion on which is the most accurate reference axis.ObjectiveTo explore the reproducibility of each reference axis of femoral external osteotomy based on the 3D CT femoral model, compare the deviation of the simulated femoral prosthesis rotation alignment, positioned based on each reference axis, with the gold standard sTEA, and analyze the accuracy of each reference axis.MethodsThe imaging data of 120 patients with knee osteoarthritis who underwent a 3D CT examination of the knee in our hospital from June 2018 to December 2021 were retrospectively collected. The 3D model of the femur was established by Mimics software. The line relative to PCL externally rotated 3° (PCL + 3°), aTEA, and the vertical line of the Whiteside line were constructed and compared with the gold standard sTEA. Intra-observer, as well as inter-observer reproducibility analysis, was performed by the intra-group correlation coefficient (ICC) and Bland-Altman method.ResultsThe angle ∠WS, between the vertical line of Whiteside and sTEA, was 2.54 ± 2.30°, with an outlier of 54.2%; the angle ∠aTEA, between aTEA and sTEA, was 4.21 ± 1.01°, with an outlier of 99.1%; the angle ∠PCL, between PCL + 3° external rotation and sTEA, was 0.50 ± 1.06°, with the highest accuracy and an outlier of 5.8%, and the differences among all three were statistically significant, P < 0.05. The intra-observer ICC values of ∠WS, ∠aTEA, and ∠PCL were 0.975 (0.964–0.982), 0.926 (0.896–0.948), and 0.924(0.892,0.946), respectively, and the reproducibility levels were excellent; the inter-observer ICC values of ∠WS, ∠aTEA, and ∠PCL were 0.968(0.955–0.978), 0.906 (0.868–0.934) and 0.970 (0.957,0.979), respectively, with excellent reproducibility levels; Bland-Altman plots suggested that the scatter points of intra-observer and inter-observer measurement differences more than 95% were within the limits of agreement.ConclusionThe reference axis for locating the distal femoral external rotation osteotomy based on the 3D CT femoral model has good reproducibility. The PCL is easy to operate, has the highest precision, and the lowest outliers among the reference axes is therefore recommended.