Interobserver Bias Research Articles

Artificial intelligence–powered programmed death ligand 1 analyser reduces interobserver variation in tumour proportion score for non–small cell lung cancer with better prediction of immunotherapy response

BackgroundManual evaluation of programmed death ligand 1 (PD-L1) tumour proportion score (TPS) by pathologists is associated with interobserver bias. ObjectiveThis study explored the role of artificial intelligence (AI)-powered TPS analyser in minimisation of interobserver variation and enhancement of therapeutic response prediction. MethodsA prototype model of an AI-powered TPS analyser was developed with a total of 802 non–small cell lung cancer (NSCLC) whole-slide images. Three independent board-certified pathologists labelled PD-L1 TPS in an external cohort of 479 NSCLC slides. For cases of disagreement between each pathologist and the AI model, the pathologists were asked to revise the TPS grade (<1%, 1%–49% and ≥50%) with AI assistance. The concordance rates among the pathologists with or without AI assistance and the effect of the AI-assisted revision on clinical outcome upon immune checkpoint inhibitor (ICI) treatment were evaluated. ResultsWithout AI assistance, pathologists concordantly classified TPS in 81.4% of the cases. They revised their initial interpretation by using the AI model for the disagreement cases between the pathologist and the AI model (N = 91, 93 and 107 for each pathologist). The overall concordance rate among the pathologists was increased to 90.2% after the AI assistance (P < 0.001). A reduction in hazard ratio for overall survival and progression-free survival upon ICI treatment was identified in the TPS subgroups after the AI-assisted TPS revision. ConclusionThe AI-powered TPS analyser assistance improves the pathologists’ consensus of reading and prediction of the therapeutic response, raising a possibility of standardised approach for the accurate interpretation.

Italy’s Hidden Hillforts: A Large-Scale Lidar-Based Mapping of Samnium

ABSTRACT This article presents the first results of the Ancient Hillforts Survey, a large-scale lidar-based analysis and ground-truthing aimed at creating a representative and comparative dataset of hillforts in Italy unbiased by site location or vegetational canopy. An analysis of 15,300 km2 spanning Campania, Lazio, Abruzzo, Molise, and Apulia detected 305 new suspected hillforts. The area was visually interpreted using image blends of lidar visualizations (VAT method) based on in-house-generated digital terrain models. Interobserver and intraobserver interpretational biases were tested and CORINE Land Cover data used to evaluate the representativeness of the legacy data compared with the new dataset and to estimate the number of sites for no-data areas. The results from the Daunian mountains (Apulia) are presented to demonstrate the effectiveness of this approach in advancing knowledge of historically under-surveyed areas and in addressing long-term debates. Here, the data showed a novel hillfort system interpretable as Samnite, dating between the 6th and 3rd century b.c.

Automated Analysis of Blood Smear Images for Leukemia Detection: A Comprehensive Review

Leukemia, the malignancy of blood-forming tissues, becomes fatal if not detected in the early stages. It is detected through a blood smear test that involves the morphological analysis of the stained blood slide. The manual microscopic examination of slides is tedious, time-consuming, error-prone, and subject to inter-observer and intra-observer bias. Several computerized methods to automate this task have been developed to alleviate these problems during the past few years. However, no exclusive comprehensive review of these methods has been presented to date. Such a review shall be highly beneficial for novice readers interested in pursuing research in this domain. This article fills the void by presenting a comprehensive review of 149 papers detailing the methods used to analyze blood smear images and detect leukemia. The primary focus of the review is on presenting the underlying techniques used and their reported performance, along with their merits and demerits. It also enumerates the research issues that have been satisfactorily solved and open challenges still existing in the domain.

Automated Volume Measurement of the Syndesmosis Using 3D Weightbearing CT

Category:Ankle; Trauma; OtherIntroduction/Purpose:Recent studies have shown that weight-bearing computed tomography (WBCT) that allows 3D volume measurement of the distal syndesmosis while under physiologic load brings about a higher accuracy in detection of syndesmotic instability, especially if subtle. Since the volume measurement method is complex, time-consuming, and has a noticeable interobserver bias, most clinicians might not still be interested in utilizing it in practice. Automatization of this measurement using computer-assisted methods will not only reduce the interobserver bias but also will be faster and more applicable in practice. This study aimed to develop an automated 3D syndesmosis volume measurement tool using image processing methods and compare the speed and the interobserver bias with human interpreters.Methods:We included 30 patients with subtle syndesmotic instability who had undergone WBCT and were diagnosed intraoperatively. Thirty individuals with otherwise healthy ankles who had WBCT images were allocated to the control group. The volume measurement up to 5cm proximal to the tibial plafond was considered as the most sensitive and specific method to assess syndesmotic stability based on the literature. An algorithm was developed using MATLAB software that could recognize and calculate the syndesmotic volume using WBCT images. The volume measurement method was used by two orthopaedic surgeons for the same population. The time spent by each surgeon and the algorithm was measured. Finally, the values were compared using the t-test; the interobserver correlation coefficient (ICC) was also calculated. P<0.05 was considered statistically significant.Results:The percentage of changes in 3D volume showed a significant difference at 5cm (25.4%; p<0.001), proximal to the tibial plafond. Based on our clinical study, the volume measurement up to 5cm showed a sensitivity of 95.8%, 95%CI:87.8-100, and specificity of 83.3%, 95% CI:68.4-98.2 for detection of syndesmotic instability. The ICC between the clinicians was 75% while using the automated measurement algorithm showed an ICC of 97%. The mean time spent by the clinicians was 268.4+-56.4 and by the algorithm was 2.9+-0.3 seconds (p<0.001, Figure 1).Conclusion:Given the fact that 3D volume measurement of the syndesmosis using WBCT is still a complex technique for most clinicians while it is one of the most accurate diagnostic methods based on recent literature, developing a faster and more accurate method for 3D volume measurement of the syndesmosis leads to a more reliable and easier to use the technique to be used in practice.

An objective assessment of microneedling therapy in atrophic facial acne scars.

Microneedling therapy is one of the recent advances in treating acne facial scars. To evaluate the efficacy of microneedling therapy in the treatment of atrophic acne facial scars. A total number of 10 patients between the age group of 20 and 40 years (8 women and 2 men) who complained of acne scars from November 2012 to August 2014 were selected for the study. Microneedling therapy was performed following the application of the Eutectic Mixture of Local Anesthetics in an interval of 2 months. All patients underwent three sessions. A follow-up of 2 months after the last sitting was carried out. Cochran's Q-paired test, Kendall's W-test, and Pearson's correlation were used for statistical analysis. Patients in our study had three forms of acne scars: box-scar, icepick form, and rolling scars. Patients were told to rate the outcome of the treatment at the end of the follow-up. Three observers randomly selected were shown the pre- and postoperative photographs of the patients to rate the treatment outcome. Statistically there was no interobserver bias. Ninety percent reduction in number of scars and depth of scars was noted at the end of three sittings. The improvement in pigmentation was insignificant. Seventy percent improvement in the skin texture was noted. Nine patients suffered from transient postinflammatory erythema and six patients had postoperative swelling. Microneedling therapy is a safe and effective method of treating acne scars.

Are Hounsfield Unit Measurements of Bony Absorption Changes in Cholesteatoma Helpful?

Introduction High-resolution computed tomography (HRCT) scans of the temporal bone are used to assess the bony erosion of the middle-ear structures whenever cholesteatoma is suspected. Objective To study the differences in HRCT Hounsfield unit (HU) index measurements of middle-ear bony structures between an ears with and without cholesteatoma. Methods A retrospective study of 59 patients who underwent surgery due to unilateral cholesteatoma. The HRCT HU index of the scutum, of three middle-ear ossicles, of the lateral semicircular canal (LSCC), and of the fallopian canal was measured in both ears. A comparison was made between the cholesteatoma and the non-cholesteatomatous ear (control). All measurements were conducted by an otolaryngologist. To assess the interobserver bias, 10% of the samples were randomly and independently assessed by another otolaryngologist and a neuroradiologist who were blinded. Results The average HU index was lower in the ear with cholesteatoma when compared with the non-cholesteatomatous ear. While the differences were statistically significant regarding the measurements of the scutum (516.02 ± 311.693 versus 855.64 ± 389.999; p = 0.001), the malleus (1049.44 ± 481.765 versus 1413.47 ± 313.376; p = 0.01), and the incus (498.03 ± 264.184 versus 714.25 ± 405.631; p = 0.001), the differences in the measurements of the LSCC (1042.34 ± 301.066 versus 1154.53 ± 359.609; p = 0.69) and of the fallopian canal (467.19 ± 221.556 versus 543.51 ± 263.573; p = 0.108) were not significantly different between both groups. The stapes was immeasurable in both groups due to its small size. Conclusion Hounsfield unit index measurements are a useful tool that may aid in the diagnosis of early-stage cholesteatoma.

1805P Assistance with an artificial intelligence-powered PD-L1 analyzer reduces interobserver variation in pathologic reading of tumor proportion score in non-small cell lung cancer

Programmed death ligand 1 (PD-L1) expression is the standard biomarker in advanced non-small cell lung cancer (NSCLC). However, manual evaluation of PD-L1 tumor proportion score (TPS) by pathologists has practical limitations of interobserver bias, variation in subjectivity on the area of interest, and intensive labor. This study aimed to explore whether the artificial intelligence (AI)-powered TPS analyzer could reduce the human discrepancy.

Impact of Aortic Diameter Measurements at Three Anatomical Landmarks on Left Ventricular Output Calculation in Neonates.

To assess reproducibility and accuracy of left ventricular output (LVO) quantifications in neonates, when left ventricular outflow tract diameter (LVOTD) was measured at the hinges of the aortic valve (AV), at the aortic sinus (AS), and at the sinotubular junction (STJ). This was an observational study. In the first cohort of very preterm neonates, we assessed intraobserver and interobserver repeatability of LVOTD measured at the AV, AS, and STJ and of the corresponding LVO. In the second cohort of older neonates, we compared paired LVO measurements by echo and magnetic resonance imaging (MRI). In the first cohort of 48 neonates, mean (standard deviation) weight and age at scan were 1046 (302) g and 28.1 (2.7) weeks. Interobserver bias (95% limits of agreement [LOA]) for LVOTD at the AV, AS, and STJ was 0 (-0.3 to 0.3)mm, 0 (-0.7 to 0.7) mm, and 0 (-0.8 to 0.7)mm, respectively. Interobserver bias (95% LOA) for the corresponding LVO was -1.3 (-31 to 33)ml/kg/min, -0.5 (-88 to 87)ml/kg/min, and -7.2 (-83 to 69)ml/kg/min, respectively. In the second cohort of 10 neonates, median (range) weight and age at scan were 1942 (970-3640)g and 37.2 (31.7-39.8)weeks. LVO measured at the AV showed stronger agreement with MRI: bias (LOA) -10.6 (-74 to 52) ml/kg/min, compared to LVO measured at AS and STJ: 194 (-0.5 to 388)ml/kg/min and 43 (-72 to 159)ml/kg/min respectively. Reproducibility and accuracy of LVO quantification by echo were better when aortic diameter was measured at AV.

Investigation of the clinical inter-observer bias in prostate fiducial marker image registration between CT and MR images

Background and purposeInter-modality image registration between computed tomography (CT) and magnetic resonance (MR) images is associated with systematic uncertainties and the magnitude of these uncertainties is not well documented. The purpose of this study was to investigate the potential uncertainty of gold fiducial marker (GFM) registration for localized prostate cancer and to estimate the inter-observer bias in a clinical setting.Methods Four experienced observers registered CT and MR images for 42 prostate cancer patients. Manual GFM identification was followed by a landmark-based registration. The absolute difference between observers in GFM identification and the displacement of the clinical target volume (CTV) was investigated. The CTV center of mass (CoM) vector displacements, DICE-index and Hausdorff distances for the observer registrations were compared against a clinical baseline registration. The time allocated for the manual registrations was compared.ResultsAbsolute difference in GFM identification between observers ranged from 0.0 to 3.0 mm. The maximum CTV CoM displacement from the clinical baseline was 3.1 mm. Displacements larger than or equal to 1 mm, 2 mm and 3 mm were 46%, 18% and 4%, respectively. No statistically significant difference was detected between observers in terms of CTV displacement. Median DICE-index and Hausdorff distance for the CTV, with their respective ranges were 0.94 [0.70–1.00] and 2.5 mm [0.7–8.7].ConclusionsRegistration of CT and MR images using GFMs for localized prostate cancer patients was subject to inter-observer bias on an individual patient level. A CTV displacement as large as 3 mm occurred for individual patients. These results show that GFM registration in a clinical setting is associated with uncertainties, which motivates the removal of inter-modality registrations in the radiotherapy workflow and a transition to an MRI-only workflow for localized prostate cancer.

Machine learning-based rapid diagnosis of human borderline ovarian cancer on second-harmonic generation images.

Regarding growth pattern and cytological characteristics, borderline ovarian tumors fall between benign and malignant, but they tend to develop malignancy. Currently, it is difficult to accurately diagnose ovarian cancer using common medical imaging methods, and histopathological examination is routinely used to obtain a definitive diagnosis. However, such examination requires experienced pathologists, being labor-intensive, time-consuming, and possibly leading to interobserver bias. By using second-harmonic generation imaging and k-nearest neighbors classifier in conjunction with automated machine learning tree-based pipeline optimization tool, we developed a computer-aided diagnosis method to classify ovarian tissues as being malignant, benign, borderline, and normal, obtaining areas under the receiver operating characteristic curve of 1.00, 0.99, 0.98, and 0.97, respectively. These results suggest that diagnosis based on second-harmonic generation images and machine learning can support the rapid and accurate detection of ovarian cancer in clinical practice.

English

Oral cancer is the sixth most common malignant tumour, and it is the leading cause of morbidity and mortality due to its capacity to spread and invade. Oral cancer occurs at a different rate in different areas of the world, ranging from 2 to 10 per 100,000 people each year. Oral cancer is prevalent in South Asian nations such as Sri Lanka, India, Pakistan, and Bangladesh. In India, the frequency is 7-17 per 100,000 people each year, with 75,000 - 80,000 new cases per year. Identifying oral cancer in its early stages has a significant impact on survival rates when compared to detecting it later. Despite this, almost half of all diagnosed patients die within five years. A variety of well-established cancer screening programmes have been demonstrated to lower the patient morbidity and mortality dramatically. Regular check-ups, which include a thorough inspection of the whole mouth, are critical for detecting malignant and pre-cancerous problems early on. Unfortunately, early detection of oral precancerous and cancerous lesions has proved difficult due to the lesions' asymptomatic nature, doctors' casual approach to benign lesions, and the fact that 50 % of patients had regional or distant metastases at the time of diagnosis. Oral cancer is one of the most common cancers that leads to defacement and death. Despite recent advancements in therapeutic modalities, the prognosis has not improved. Patient’s mortality rates are positively associated with the point of presentation, with 60 % of people diagnosed with late-stage illness. Early diagnosis is important for oral cancer patient’s survival rate, as it decreases morbidity and mortality. According to the World Health Organization, the bulk of oral cancer patients are diagnosed late in the disease's progression, with a mediocre 5 - year survival rate of 50 %. As a result, careful treatment of oral cancer necessitates early diagnosis and intervention. Surgical biopsy is the gold standard for medical purposes, but it requires clinical assistance. Other screening methods that are simple to use, non-invasive, and expensive are the norms for any test to be accepted as a histopathology choice. The older cancer diagnosis modalities took longer, had more inter-observer bias, and were less descriptive. A standard oral examination with digital palpation is used in traditional techniques of screening for oral possibly malignant illnesses and oral cancers. Conventional inspection has been shown to be a poor discriminator of oral mucosal lesions. A variety of visual aids have been developed to help clinicians spot anomalies in the oral mucosa and in recent years, scientific and clinical developments have aided in the early detection and treatment of this disease. This review reflects on some of the older diagnostic modalities and screening methods for oral cancer diagnosis, as well as some of the recent more sophisticated techniques. KEY WORDS Diagnostic Aids, Oral Cancer, Premalignant Lesions

Translating Fecobionics Into a Technique That Addresses Clinical Needs for Objective Perineal Descent Measurements.

INTRODUCTION:Perineal descent is a phenomenon associated with anorectal dysfunction. It is diagnosed by defecography but subjected to manual measurements on the images/videos and interobserver bias. Fecobionics is a simulated feces for assessing important physiological parameters during defecation. Here, we translate Fecobionics into a new method for estimation of perineal descent based on electronic signals from the embedded inertial measurement units (IMUs).METHODS:A displacement measurement method by a combined zero-velocity update and gravity compensation algorithm from IMUs was developed. The method was verified in a robot model, which mimicked perineal descent motion.RESULTS:The method correlated well with the reference (R2 = 0.9789) and had a deviation from the peak displacement (range 0.25–2.5 cm) of −0.04 ± 0.498 cm. The method was further validated in 5 human experiments with comparison to the benchmark defecography technology (R2 = 0.79).DISCUSSION:The proposed technology is objective, i.e., electronic measurements rather than by fluoroscopy or MRI. The development may impact clinical practice by providing a resource-saving and objective technology for diagnosing perineal descent in the many patients suffering from anorectal disorders. The technology may also be used in colon experiments with Fecobionics and for other gastrointestinal devices containing IMUs such as ingestible capsules like the Smartpill.

Differential diagnosis of renal oncocytoma and chromophobe renal cell carcinoma using CT features: a central scar-matched retrospective study.

Renal oncocytoma (RO) and chromophobe renal cell carcinoma (chRCC) have a common cellular origin and different clinical management and prognosis. To explore the utility of computed tomography (CT) in the differentiation of RO and chRCC. Twenty-five patients with RO and 73 patients with chRCC presenting with the central scar were included retrospectively. Two experienced radiologists independently reviewed the CT imaging features, including location, tumor size, relative density ratio, segmental enhancement inversion (SEI), necrosis, and perirenal fascia thickening, among others. Interclass correlation coefficient (ICC, for continuous variables) or Kappa coefficient test (for categorical variables) was used to determine intra-observer and inter-observer bias between the two radiologists. The inter- and intra-reader reproducibility of the other CT imaging parameters were nearly perfect (>0.81) except for the measurements of fat (0.662). RO differed from chRCC in the cortical or medullary side (P = 0.005), relative density ratio (P = 0.020), SEI (P < 0.001), and necrosis (P = 0.045). The logistic regression model showed that location (right kidney), hypo-density on non-enhanced CT, SEI, and perirenal fascia thickening were highly predictive of RO. The combined indicators from logistic regression model were used for ROC analysis. The area under the ROC curve was 0.923 (P < 0.001). The sensitivity and specificity of the four factors combined for diagnosing RO were 88% and 86.3%, respectively. The correlation coefficient between necrosis and tumor size in all tumors including both of RO and chRCC was 0.584, indicating a positive correlation (P < 0.001). The CT imaging features of location (right kidney), hypo-density on non-enhanced CT, SEI, and perirenal fascia thickening were valuable indicators in distinguishing RO from chRCC.

Bioimpedance method for monitoring venous ulcers: Clinical proof-of-concept study

Evaluation of wound status is typically based on means which require the removal of dressings. These procedures are often also subjective and prone to inter-observer bias. To overcome aforementioned issues a bioimpedance measurement-based method and measurement system has been developed to evaluate the state of wound healing. The measurement system incorporated a purpose-built bioimpedance device, a measurement software and a screen-printed electrode array. The feasibility and the performance of the system and method were assessed in an open non-randomized follow-up study of seven venous ulcers. Healing of ulcers was monitored until the complete re-epithelialization was achieved. The duration of follow-up was from 19 to 106 days (mean 55.8 ± 25.2 days). A variable designated as the Wound Status Index (WSI), derived from the bioimpedance data, was used for describing the state of wound healing. The wound surface area was measured using acetate tracing for the reference. A strong correlation was found between the WSI and the acetate tracing data, r(93) = - 0.84, p < 0.001. The results indicate that the bioimpedance measurement-based method is a promising quantitative tool for the evaluation of the status of venous ulcers.

The use of magnetic resonance imaging technique and 3D printing in order to develop a three-dimensional fistula model for patients with Crohn's disease: personalised medicine.

IntroductionPreoperative evaluation of magnetic resonance (MR) images may not be sufficient for the precise planning of anal fistula surgery or for stem cell therapy. Three-dimensional (3D) printing allows one to obtain spatial structures in a 1 : 1 scale with unprecedented precision.AimTo combine magnetic resonance imaging (MRI) with 3D printing for more precise visualisation of perianal Crohn’s disease.Material and methodsMRI at 1.5T and a 3D printer were used. DICOM (Digital imaging and communications in medicine) images were imported into 3D Slicer v.4.8.0. Firstly, anal fistula was modelled on the basis of axial images. Fistula locations, the anus and anal canal, were marked with different coloured markers. The last step was to mark the skin that was connected to the anus and contact areas of the fistula with the skin. The prepared models were then exported to an STL format file. The anal fistula model was printed using a 3D printer. The development of the model, including printing, took approximately 6 h.Results and conclusionsThe accessibility of a rotatable 3D model before surgery allows for a more precise detection of the location and the degree of perianal disease. Moreover, this may also lower the inter-observer bias connected with interpretation of complex MR imaging before planned surgery. Development of MRI image transfer to 3D printing and the decreasing cost of 3D printers suggests a promising future of this technology in medical applications.

Feasibility of rapid magnetic resonance imaging (rMRI) for the emergency evaluation of suspected pediatric orbital cellulitis

To determine the feasibility of noncontrast rapid magnetic resonance imaging (rMRI), compared with traditional contrast-enhanced computed tomography (CT) in assessing pediatric emergency department patients with suspected orbital cellulitis or orbital abscess. All subjects <19years of age who presented emergently with suspected orbital cellulitis from July 1, 2017, to July 31, 2019, were included. Participants received both the standard contrast orbital CT, if deemed necessary, with the addition of the noncontrast rMRI after informed consent was obtained. No sedation was used for either examination. All clinical decisions were based on CT findings; rMRI was interpreted within 24hours of the visit. Three pediatric radiologists, with 8-21years' experience of pediatric neuroradiology, interpreted the rMRI, masked to the CT and clinical results. Results were analyzed for interobserver bias. A total of 14 patients were enrolled during the study period. Mean age was 5.9years (range, 0.33-13). Of the 14 patients, 13 (93%) were able to complete the rMRI at 1.5 and 3T; 1 patient (1.67years of age) was unable to complete the rMRI (no images obtained). Of the 26 unilateral orbital units assessed, 3 were positive for retroseptal orbital cellulitis by CT and were diagnosed correctly by rMRI. Interobserver agreement was 100% in detecting presence or absence of retroseptal cellulitis. CT and rMRI findings were concordant in 100% of cases in differentiating preseptal vs orbital cellulitis. Kappa statistics for three-category ratings by three raters for right eye/orbit was 0.921 and for left eye/orbit was 0.9288, suggesting almost perfect agreement. Concordance correlation coefficients were 0.938 for the right eye and 0.955 for the left eye. Noncontrast rMRI orbits showed findings concordant in all cases with contrast-enhanced CT for differentiating preseptal cellulitis from orbital cellulitis.

Anogenital Distance in Healthy Infants: Method-, Age- and Sex-related Reference Ranges.

The use of anogenital distance (AGD) in clinical and epidemiological settings is increasing; however, sex-specific reference data on AGD and data on longitudinal changes in AGD in children is scarce. To create age-, sex-, and method-related reference ranges of AGD in healthy boys and girls aged 0-24 months, to assess the age-related changes in AGD and to evaluate the 2 predominantly used methods of AGD measurement. The International AGD consortium comprising 4 centers compiled data from 1 cross-sectional and 3 longitudinal cohort studies (clinicaltrials.gov [NCT02497209]). All data were collected from population-based studies, recruiting from 4 maternity or obstetric centers (United States, Cambridge [United Kingdom], Odense, and Copenhagen [Denmark]). This study included a total of 3705 healthy, mainly Caucasian children aged 0-24 months on whom 7295 measurements were recorded. AGDAS (ano-scrotal), AGDAF (ano-fourchette), AGDAP (ano-penile), AGDAC (ano-clitoral), AGD body size indices (weight, body mass index [BMI], body surface area, and length), and intra- and interobserver biases. We created age-specific reference ranges by centers. We found that AGD increased from birth to 6 months of age and thereafter reached a plateau. Changes in AGD/BMI during the first year of life were minor (0-6% and 0-11% in boys and girls, respectively). Reference ranges for AGD can be used in future epidemiological research and may be utilized clinically to evaluate prenatal androgen action in differences-in-sex-development patients. The increase in AGD during the first year of life was age-related, while AGD/BMI was fairly stable. The TIDES and Cambridge methods were equally reproducible.

Continuous monitoring of mobility performance trajectory in patients receiving chemotherapy.

e14104 Background: Accurate performance status (PS) assessment is important for chemotherapy decisions, but current tools such as the Eastern Cooperative Oncology Group (ECOG) PS assessment are subjective, static, and prone to inter-observer and recall biases. Continuous patient monitoring technology can enhance PS assessment by providing an objective and dynamic measurement of mobility performance (MP), defined as mobility in real-life situations. This study examined the feasibility of a pendant sensor to continuously track MP during a 4-week period of chemotherapy. Methods: Patients on chemotherapy were recruited from a single center. Patients were asked to continuously wear a water-resistant pendant sensor (PAMSys, Biosensics, USA) for 4-weeks during the day and night. Patients completed psychosocial surveys (cognitive, depression, and fatigue) and a technology acceptance survey measured on a Likert scale from 1 (strongly disagree) to 5 (strongly agree). Chart review was completed for demographics, medical history, and baseline ECOG PS. Using a validated algorithm, postures (lying and sitting postures: LSP) and walking parameters (cadence, longest walking bout, total steps per day) were extracted from the pendant. We also calculated percent change from baseline activity to the lowest activity level during the 4-week period. To determine clinically meaningful digital biomarkers, we compared these MP measures between two groups: ECOG 0 or ECOG ≥1. Results: Fifteen patients (median age 68), with either a solid (n = 14) or hematologic (n = 1) malignancy, stages II-IV, with ECOG 0 (n = 7) and ECOG ≥1 (n = 8) were evaluated. Baseline psychosocial parameters were not significantly different. ECOG ≥1 tend to have poorer baseline walking compared to ECOG 0 with the largest effect size observed for cadence (79.2 ± 10.0 vs 73.3±3.7, Cohen’s d effect size d = 0.78). The percentage of MP decline was higher in ECOG 0 compared to ECOG ≥1 (89.0±22.9% vs 79.9±34.5%). None of the observed trends reached statistical significance. Patients reported that the pendant was comfortable to wear (4.9±0.30) and did not interfere with activities (4.9±0.25). Conclusions: This study showed the feasibility of a pendant sensor to continuously monitor MP and its trajectory during chemotherapy. Results suggest continuous MP monitoring may provide objective information to enhance PS assessment. Larger studies are needed to confirm statistical significance of digital MP metrics and its potential effect on clinical decisions such as dose reduction, rehabilitation, and palliative care.

MON-LB86 The Sensitivity and Specificity of Various Thyroid Nodule Ultrasound Characteristics and the Diagnostic Accuracy of the ATA Guidelines and ACR TI-RADS for Predicting Thyroid Cancer at an Urban Endocrinology Clinic

Introduction: Several current guidelines assess sonographic features to guide management of thyroid nodules. The ACR uses an additive point system to assign the level of risk to various sonographic features, whereas the ATA groups sonographic features together to determine the level of risk. The purpose of this study is to compare the performance of the ATA guidelines and ACR TI-RADS at an urban endocrinology clinic in risk stratifying thyroid nodules by their specific sonographic features. Methods: This retrospective, chart-review study includes adult patients who met sonographic criteria for fine needle aspiration (FNA) biopsy based on ATA or ACR TI-RADS at an outpatient endocrinology practice in San Francisco, CA between December 2011 and August 2019. Patients with a prior history of thyroid malignancy (anaplastic and medullary thyroid carcinoma or thyroid lymphoma) were excluded. The reference standard for the diagnosis of malignancy was surgical pathology or FNA cytology Bethesda category V or VI when surgical pathology was unavailable. Analysis of guideline performances and specific sonographic features included: sensitivities (Sn), specificities (Sp), positive predictive values (PPV), negative predictive values (NPV), and area-under-the-curve (AUC) using Fisher’s exact test. Results: Two hundred seventy-five nodules among 195 adults (86.2% were women) were included in the analysis. Twelve nodules were malignant, with an associated malignancy rate of 4.4%. TI-RADS had higher accuracy based on AUC of 0.710 compared to 0.623 using ATA guidelines. TI-RADS also had a higher PPV of 21.4% among nodules with 9 points, versus 5.8% among nodules in the ATA “high suspicion” category. Ultrasound characteristics with the highest Sp, relative PPV and NPV were: microcalcifications (84.5%, 4.3%, 96.0%, respectively), taller-than-wide (81.7%, 7.1%, 96.7%), irregular margins (77.7%, 6.0%, 96.5%); the characteristic with the highest Sn was hypoechogenicity (83.3%), however this had relatively low Sp (25.4%) and PPV (4.5%). Conclusions: TI-RADS performed better with higher overall accuracy and PPV when applied to nodules classified as having the highest malignancy risk. Taller-than-wide shape, irregular margins, and microcalcifications were the characteristics most useful for malignancy risk stratification. Limitations of this study include: interobserver bias, small sample size, referred patient population (which may differ from other institutions), and inability in some cases to confirm malignant FNA cytology with surgical pathology.

Age estimation in an Indian subpopulation by cone-beam computed tomographic analysis of spheno-occipital synchondrosis fusion

Age estimation has been an ever-evolving aspect of forensic science. There is no dearth of markers in estimating age around 14–15 years; however, there is paucity for adolescent- adult transition markers. Various studies have cited the late closure of spheno-occipital synchondrosis (SOS). This study is an attempt at ascertaining the possibility of age estimation in an Indian subpopulation by cone-beam computed tomographic (CBCT) analysis of SOS.271 (145 males, 126 females; 10–25 years) full volume scans were assessed on the Planmeca Promax, 3D-mid CBCT unit. On the basis of the staging system proposed by Lottering et al (2015), the SOS was analyzed in coronal and sagittal planes and its closure was graded in six stages (1–6). The intra-class correlation (ICC) coefficient for the intra and inter-observer bias was 0.98 and 0.95 respectively, depicting an excellent agreement.This study depicts that closure of SOS has a linear relation with age. Females tend to attain each stage earlier than males. However, this sexual dimorphism exists till 16 years of age, thereafter, 100 % population shows complete closure of SOS.The closure of SOS can be used to determine age in the adolescent-adult transition phase with more reliability in males as compared to females. In the central Indian population, complete fusion of SOS (Stage 6) was noted in individuals aged 16 years and above, irrespective of sex.