Estimating organ dose with optimized peak dose index in cone-beam CT scans.

PURPOSE: Organ dose evaluation is important for optimizing cone beam computed tomography (CBCT) scan protocols. However, an evaluation method for various CBCT scanners is yet to be established. In this study, we developed scanner-independent conversion coefficients to estimate organ doses using appropriate peak dose (f(0)) indices. METHODS: This study included various scanners (angiography scanners and linear accelerators) and protocols for the head and body (thorax, abdomen, and pelvis) scan regions. f(0) was measured at five conventional positions (center position (f(0)c) and four peripheral positions (f(0)p) at 90° intervals) in the CT dose index (CTDI) phantom. To identify appropriate measurement positions for organ dose estimation, various f(0) indices were considered. Organ doses were measured by using optically stimulated luminescence dosimeters positioned in an anthropomorphic phantom. Thereafter, the conversion coefficients were calculated from each obtained f(0) value and organ or tissue dose using a linear fit for all scanners, and the coefficient of variation (CV) of the conversion coefficients was calculated for each organ or tissue. The f(0) index with the minimum CV value was proposed as the appropriate index. RESULTS: The appropriate f(0) index was determined as f(0)c for the body region and a maximum of four f(0)p values for the head region. Using the proposed conversion coefficients based on the appropriate f(0) index, the organ/tissue doses were well estimated with a mean error of 14.2% across all scanners and scan regions. CONCLUSIONS: The proposed scanner-independent coefficients are useful for organ dose evaluation using CBCT scanners.

Prediction of persistent occiput posterior position by sonographic assessment of fetal head attitude at start of second stage of labor: prospective study.

OBJECTIVES: To evaluate the relationship between the attitude of the fetal head quantified by means of the chin-to-chest angle (CCA) in fetuses in occiput posterior (OP) position at the beginning of the second stage of labor, and persistent OP position at birth. METHODS: This was a single-center, prospective observational study conducted at the University Hospital of Parma, Parma, Italy. We included singleton pregnancies at term with fetuses in the OP position at the beginning of the second stage of labor. The fetal head position, station by means of angle of progression and head-to-perineum distance, and attitude by means of CCA were assessed using transabdominal or transperineal ultrasound. The primary outcome was persistent OP position at birth. RESULTS: Between January and July 2022, 76 women were included in the study. There were 48 (63.2%) spontaneous rotations of the fetal head and spontaneous vaginal delivery occurred in all. Among the 28 (36.8%) fetuses that did not rotate spontaneously into an occiput anterior position, eight (28.6%) had a spontaneous vaginal delivery, while operative vaginal delivery and Cesarean delivery was performed in 11 (39.3%) and nine (32.1%) cases, respectively. Multivariable logistic regression analysis showed that the CCA (adjusted odds ratio (aOR), 2.15 (95% CI, 1.22-3.78); P = 0.008) and nulliparity (aOR, 0.20 (95% CI, 0.06-0.76); P = 0.02) were associated independently with persistent OP position at birth. Moreover, the CCA showed an area under the receiver-operating-characteristics curve of 0.69 (95% CI, 0.56-0.82); P = 0.005) for the prediction of persistent OP position. The optimal cut-off value of the CCA was 36.5°, and was associated with a sensitivity of 0.82 (95% CI, 0.63-0.94), specificity of 0.50 (95% CI, 0.35-0.65), positive predictive value of 0.49 (95% CI, 0.34-0.64), negative predictive value of 0.83 (95% CI, 0.64-0.94), positive likelihood ratio of 1.64 (95% CI, 1.18-2.29) and negative likelihood ratio of 0.36 (95% CI, 0.15-0.83). CONCLUSIONS: Our data show that, within a population of women with fetal OP position at the beginning of the second stage of labor, the sonographic fetal head attitude measured by means of the CCA might help in the identification of fetuses at risk of persistent OP position. Such findings can be useful for patient counseling when OP position is diagnosed at full cervical dilatation. Further studies should investigate if the CCA might select patients who may benefit from manual rotation of the fetal head. © 2023 The Authors. Ultrasound in Obstetrics & Gynecology published by John Wiley & Sons Ltd on behalf of International Society of Ultrasound in Obstetrics and Gynecology.

Does Head and Neck Posture Affect Cone-Beam Computed Tomography Assessment of the Upper Airway?

PURPOSE: Radiographic analysis is often used as a screening tool to assess for risk of sleep-related breathing disorders. This study aimed to address 2 questions: (1) Does head posture significantly affect the minimum cross-sectional area (MCA)? and (2) Is the NBC3 (nasion-basion-C3) angle a reliable measurement to control for alteration of head position in cone-beam computed tomography (CBCT) scans? METHODS: Study design: prospective cohort study. SETTING: Private practices affiliated with a research institution. PARTICIPANTS: convenience sample of adult volunteers. VARIABLES: CBCT scans were taken in 2 conditions: baseline (natural head position, NHP) and 1 of 5 experimental conditions (head tilted up, head tilted down, sitting vs standing, use of chin-rest, and swallow variation). For the primary aim of our study, the primary predictor variable was head posture and the main outcome variable was percentage change in MCA. For the secondary aim, the primary predictor variable was head posture and the main outcome variable was degree change in NBC3 angle. RESULTS: Ninety subjects were included (age 40.7 ± 13.7 years, 70% female). Mean NBC3 at baseline was 112.4 ± 8.3°. Head tilted down significantly decreased (-41.4 ± 18.5 mm2, P = .03∗) and head tilted up significantly increased MCA (+147.4 ± 43.3 mm2, P = .0018∗). Head tilted down significantly reduced the NBC3 angle measurement (-10.5 ± 6.8°, P = .006) and head tilted up significantly increased the NBC3 angle measurement (+14.4 ± 5.8°, P = .0004). A quadratic regression model was fitted with moderately strong correlation (R2 = 0.54) showing an exponential effect of small changes in the NBC3 angle on MCA, P < .0001. The model predicts that increasing NBC3 by +5 and + 10° resulted in MCA changes of +25% and +88%, whereas a decrease in NBC3 by -5 and -10° results in MCA changes of -21% and -23%, respectively. CONCLUSION: Alterations in head posture significantly affect the MCA of the upper airway on CBCT. The NBC3 angle can be used to reliably assess changes in cranio-cervical extension and validate comparisons of MCA between CBCT scans for the same patient. A standardized protocol for CBCT acquisition is proposed.

Technical note: Flat panel proton radiography with a patient specific imaging field for accurate WEPL assessment.

BACKGROUND: Proton radiography (PR) uses highly energetic proton beams to create images where energy loss is the main contrast mechanism. Water-equivalent path length (WEPL) measurements using flat panel PR (FP-PR) have potential for in vivo range verification. However, an accurate WEPL measurement via FP-PR requires irradiation with multiple energy layers, imposing high imaging doses. PURPOSE: A FP-PR method is proposed for accurate WEPL determination based on a patient-specific imaging field with a reduced number of energies (n) to minimize imaging dose. METHODS: Patient-specific FP-PRs were simulated and measured for a head and neck (HN) phantom. An energy selection algorithm estimated spot-wise the lowest energy required to cross the anatomy (Emin) using a water-equivalent thickness map. Starting from Emin, n was restricted to certain values (n = 26, 24, 22, …, 2 for simulations, n = 10 for measurements), resulting in patient-specific FP-PRs. A reference FP-PR with a complete set of energies was compared against patient-specific FP-PRs covering the whole anatomy via mean absolute WEPL differences (MAD), to evaluate the impact of the developed algorithm. WEPL accuracy of patient-specific FP-PRs was assessed using mean relative WEPL errors (MRE) with respect to measured multi-layer ionization chamber PRs (MLIC-PR) in the base of skull, brain, and neck regions. RESULTS: MADs ranged from 2.1 mm (n = 26) to 21.0 mm (n = 2) for simulated FP-PRs, and 7.2 mm for measured FP-PRs (n = 10). WEPL differences below 1 mm were observed across the whole anatomy, except at the phantom surfaces. Measured patient-specific FP-PRs showed good agreement against MLIC-PRs, with MREs of 1.3 ± 2.0%, -0.1 ± 1.0%, and -0.1 ± 0.4% in the three regions of the phantom. CONCLUSION: A method to obtain accurate WEPL measurements using FP-PR with a reduced number of energies selected for the individual patient anatomy was established in silico and validated experimentally. Patient-specific FP-PRs could provide means of in vivo range verification.

Cephalometric assessment regarding craniocervical posture in orthodontic patients.

A major factor that contributes to dental malocclusions is represented by the positioning of the mandible. Considering the existing interconnections between the craniocervical and craniomandibular systems it is interesting to assess how changes in one system can influence the other, thus establishing a pattern in terms of certain cephalometric landmarks that orthodontists could consider when diagnosing and evaluating an orthodontic case. Therefore, the aim of this study was to investigate the connections between cervical posture, head position, hyoid bone position in orthodontic patients with different skeletal patterns. 45 lateral cephalometric radiographs were analyzed. Skeletal class and vertical growth were the main elements that were considered when classifying patients. Craniofacial and Cervical landmarks were determined on the cephalograms, from which lines and angles resulted which were considered relevant in our study. Correlations between cephalometric variables of the patients were determined. there were some statistically significant changes identified concerning craniocervical posture and hyoid bone position between the patients in the following parameters: H-Rgn, OPT/HOR, CVT/HOR, OPT/SN, CVT/SN, H-SN. The results obtained allowed us to conclude that there were some differences at the skeletal level of the sample of patients studied. The findings are indicating that there is a close relationship between, mandible position, cervical- and head posture and the hyoid bone. The information obtained in this study could help to better understand the development of malocclusions, and to improve the orthodontic diagnostic and treatment plan.

Cost-efficient anthropomorphic head phantom for quantitative image quality assessment in cone beam CT.

In this study, a novel anthropomorphic head phantom for quantitative image quality assessment in cone beam computed tomography (CBCT) is proposed. The phantom is composed of tissue equivalent materials (TEMs) which are suitable for cost-efficient fabrication methods such as silicone casting and 3D printing. A monocalcium phosphate/gypsum mixture (MCPHG), nylon and a silyl modified polymer gel (SMP) are proposed as bone, muscle and brain equivalent materials respectively. The TEMs were evaluated for their radiodensity in terms of Hounsfield Units (HU) and their x-ray scatter characteristics. The median radiodensity and inter quartile range (IQR) of the MCPHG and SMP were found to be within the range of the theoretical radiodensity for bone and brain tissue: 922 (IQR = 156) and 47 (IQR = 7) HU respectively. The median radiodensity of nylon was slightly outside of the HU range of muscle tissue, but within the HU range of a combination of muscle and adipose tissue: -18 (IQR = 40) HU. The median ratios between the measured scatter characteristics and simulated tissues were between 0.84 and 1.13 (IQR between 0.05 and 0.14). The preliminary results of this study show that the proposed design and TEMs are potentially suitable for the fabrication of a cost-efficient anthropomorphic head phantom for quantitative image quality assessment in CT or CBCT.

Craniofacial Growth and Asymmetry in Newborns: A Longitudinal 3D Assessment.

OBJECTIVE: To evaluate the development of the craniofacial region in healthy infants and analyze the asymmetry pattern in the first year of life. METHODS: The participants were grouped by sex and age (1, 2, 4, 6, 9, and 12 months) to receive three-dimensional (3D) photographs. Stereoscopic craniofacial photos were captured and transformed into a series of craniofacial meshes in each group. The growth patterns of the anthropometric indices and the degree of craniofacial asymmetry were measured, and average craniofacial meshes and color-asymmetry maps with craniofacial asymmetry scores were calculated. RESULTS: A total of 373 photographs from 66 infants were obtained. In both genders, the highest and lowest growth rates for all anthropometric indices were noted between 1 and 2 months and between 9 and 12 months, respectively. Overall, male infants had higher anthropometric indices, head volume, and head circumference than female infants. The craniofacial asymmetry score was presented with a descending pattern from 1 to 12 months of age in both sex groups. Both sex groups showed decreased left-sided laterality in the temporal-parietal-occipital region between 1 and 4 months of age and increased right frontal-temporal prominence between 6 and 12 months of age. CONCLUSIONS: A longitudinal evaluation of the craniofacial growth of healthy infants during their first year of life was presented.

Image quality assessment of three cone beam computed tomography scanners-an analysis of the visibility of anatomical landmarks.

BACKGROUND/OBJECTIVES: The aim of this study was to evaluate large field of view (FoV) subjective image quality of three cone-beam computed tomography (CBCT) machines to assess the threshold for reliable diagnostic perceptibility when lowering exposure settings. MATERIALS AND METHODS: One entire cadaver's head was scanned using three CBCT scanners. The largest available FoV of each device, imaging orthodontic relevant structures, was applied. CBCT datasets with different image quality based on standard to minimum scanning parameters were acquired. Five dentists evaluated the visibility of selected anatomical structures of the upper and lower jaw using a 5-point rating scale. RESULTS: Image quality depends on parameters such as a minimum voxel size of 0.2 to 0.4mm or a minimum of 4mA. A reduction in number of images (scanning protocol) reduces image quality. Visualization of different anatomical structures for orthodontic treatment planning requires distinct scanning protocols to support adequate perception of these structures. LIMITATIONS: This study does not account for the evaluation of paediatric anatomical structures due to the availability of cadaver's head. CONCLUSIONS: CBCT scans performed for orthodontic purposes using a large FoV with reduced parameters (400 µm, 2 to 4 mA and low dose protocols) are acceptable for visualization of large anatomical structures. Further lowering these parameters will not be sufficient to view small anatomical structures. Orthodontic indications will have to define specific anatomical structures to choose adequate scanning protocols to reduce dose and ensure reliable diagnostic visibility.

Quantitative assessment and localization of the hollowing of the temple after craniectomy and cranioplasty-The frontozygomatic shadow.

BACKGROUND: After cranioplasty, in many cases a not negligible soft tissue defect remains in the temporozygomatical area, also referred to as a hollowing defect of the temple. OBJECTIVE: To assess the precise localization and volume of the hollowing defect, to optimize future cranioplasties. METHODS: CT data of patients who received craniectomy and conventional CAD cranioplasty in our institution between 2012 and 2018 were analyzed. CT datasets prior to craniectomy and after cranioplasty were subtracted to quantify the volume and localization of the defect. RESULTS: Out of 91 patients, 21 had suitable datasets. Five cases had good cosmetic results with no defect visible, 16 patients had an apparent hollowing defect. Their average defect volume was 5.0 cm3 ± 4.5 cm3. The defect localizations were in the area behind the zygomatic process and just below the superior temporal line, covering an area of app. 3x3 cm2. Surgical attempts of temporal muscle restoration were more often found in reports of good results (p<0.01), but also in 50% of reports, whose surgeries resulted in hollowing of the temple. Mean time between the two surgeries was 112 ± 43 days. No significant differences between patients with and without hollowing defect were detected regarding time between the two surgeries, age or performing surgeon. CONCLUSION: This work supplies evidence for the indication of a surgical corrective during cranioplasty in the small but cosmetically relevant area of the "frontozygomatic shadow". Based on our 3D data analysis, future focused surgical strategies may obtain better aesthetical results here.

Stereophotogrammetric head shape assessment in neonates is feasible and can identify distinct differences between term-born and very preterm infants at term equivalent age.

The development of head shape and volume may reflect neurodevelopmental outcome and therefore is of paramount importance in neonatal care. Here, we compare head morphology in 25 very preterm infants with a birth weight of below 1500 g and / or a gestational age (GA) before 32 completed weeks to 25 term infants with a GA of 37-42 weeks at term equivalent age (TEA) and identify possible risk factors for non-synostotic head shape deformities. For three-dimensional head assessments, a portable stereophotogrammetric device was used. The most common and distinct head shape deformity in preterm infants was dolichocephaly. Severity of dolichocephaly correlated with GA and body weight at TEA but not with other factors such as neonatal morbidity, sex or total duration of respiratory support. Head circumference (HC) and cranial volume (CV) were not significantly different between the preterm and term infant group. Digitally measured HC and the CV significantly correlated even in infants with head shape deformities. Our study shows that stereophotogrammetric head assessment is feasible in all preterm and term infants and provides valuable information on volumetry and comprehensive head shape characteristics. In a small sample of preterm infants, body weight at TEA was identified as a specific risk factor for the development of dolichocephaly.

Assessment of the global noise algorithm for automatic noise measurement in head CT examinations.

PURPOSE: The global noise (GN) algorithm has been previously introduced as a method for automatic noise measurement in clinical CT images. The accuracy of the GN algorithm has been assessed in abdomen CT examinations, but not in any other body part until now. This work assesses the GN algorithm accuracy in automatic noise measurement in head CT examinations. METHODS: A publicly available image dataset of 99 head CT examinations was used to evaluate the accuracy of the GN algorithm in comparison to reference noise values. Reference noise values were acquired using a manual noise measurement procedure. The procedure used a consistent instruction protocol and multiple observers to mitigate the influence of intra- and interobserver variation, resulting in precise reference values. Optimal GN algorithm parameter values were determined. The GN algorithm accuracy and the corresponding statistical confidence interval were determined. The GN measurements were compared across the six different scan protocols used in this dataset. The correlation of GN to patient head size was also assessed using a linear regression model, and the CT scanner's X-ray beam quality was inferred from the model fit parameters. RESULTS: Across all head CT examinations in the dataset, the range of reference noise was 2.9-10.2 HU. A precision of ±0.33 HU was achieved in the reference noise measurements. After optimization, the GN algorithm had a RMS error 0.34 HU corresponding to a percent RMS error of 6.6%. The GN algorithm had a bias of +3.9%. Statistically significant differences in GN were detected in 11 out of the 15 different pairs of scan protocols. The GN measurements were correlated with head size with a statistically significant regression slope parameter (p < 10-7 ). The CT scanner X-ray beam quality estimated from the slope parameter was 3.5 cm water HVL (2.8-4.8 cm 95% CI). CONCLUSION: The GN algorithm was validated for application in head CT examinations. The GN algorithm was accurate in comparison to reference manual measurement, with errors comparable to interobserver variation in manual measurement. The GN algorithm can detect noise differences in examinations performed on different scanner models or using different scan protocols. The trend in GN across patients of different head sizes closely follows that predicted by a physical model of X-ray attenuation.

Intrapartum sonographic assessment of the fetal head flexion in protracted active phase of labor and association with labor outcome: a multicenter, prospective study.

BACKGROUND: To date, no research has focused on the sonographic quantification of the degree of flexion of the fetal head in relation to the labor outcome in women with protracted active phase of labor. OBJECTIVE: This study aimed to assess the relationship between the transabdominal sonographic indices of fetal head flexion and the mode of delivery in women with protracted active phase of labor. STUDY DESIGN: Prospective evaluation of women with protracted active phase of labor recruited across 3 tertiary maternity units. Eligible cases were submitted to transabdominal ultrasound for the evaluation of the fetal head position and flexion, which was measured by means of the occiput-spine angle in fetuses in nonocciput posterior position and by means of the chin-to-chest angle in fetuses in occiput posterior position. The occiput-spine angle and the chin-to-chest angle were compared between women who had vaginal delivery and those who had cesarean delivery. Cases where obstetrical intervention was performed solely based on suspected fetal distress were excluded. RESULTS: A total of 129 women were included, of whom 43 (33.3%) had occiput posterior position. Spontaneous vaginal delivery, instrumental delivery, and cesarean delivery were recorded in 66 (51.2%), 17 (13.1%), and 46 (35.7%) cases, respectively. A wider occiput-spine angle was measured in women who had vaginal delivery compared with those submitted to cesarean delivery owing to labor dystocia (126±14 vs 115±24; P<.01). At the receiver operating characteristic curve, the area under the curve was 0.675 (95% confidence interval, 0.538-0.812; P<.01), and the optimal occiput-spine angle cutoff value discriminating between cases of vaginal delivery and those delivered by cesarean delivery was 109°. A narrower chin-to-chest angle was measured in cases who had vaginal delivery compared with those undergoing cesarean delivery (27±33 vs 56±28 degrees; P<.01). The area under the curve of the chin-to-chest angle in relation to the mode of delivery was 0.758 (95% confidence interval, 0.612-0.904; P<.01), and the optimal cutoff value discriminating between vaginal delivery and cesarean delivery was 33.0°. CONCLUSION: In women with protracted active phase of labor, the sonographic demonstration of fetal head deflexion in occiput posterior and in nonocciput posterior fetuses is associated with an increased incidence of cesarean delivery owing to labor dystocia. Such findings suggest that intrapartum ultrasound may contribute in the categorization of the etiology of labor dystocia.

Open-source, cost-effective, portable, 3D-printed digital lensless holographic microscope.

In this work, the design, construction, and testing of the most cost-effective digital lensless holographic microscope to date are presented. The architecture of digital lensless holographic microscopy (DLHM) is built by means of a 3D-printed setup and utilizing off-the-shelf materials to produce a DLHM microscope costing US$52.82. For the processing of the recorded in-line holograms, an open-source software specifically developed to process this type of recordings is utilized. The presented DLHM setup has all the degrees of freedom needed to achieve different fields of view, levels of spatial resolution, and 2D scanning of the sample. The feasibility of the presented platform is tested by imaging non-bio and bio samples; the resolution test targets, a section of the head of a Drosophila melanogaster fly, red blood cells, and cheek cells are imaged on the built microscope.

Decreased neonatal morbidity in 'stomach-down' left congenital diaphragmatic hernia: implications of prenatal ultrasound diagnosis for counseling and postnatal management.

OBJECTIVE: To evaluate the influence of stomach position on postnatal outcome in cases of left congenital diaphragmatic hernia (CDH) without liver herniation, diagnosed and characterized on prenatal ultrasound (US), by comparing those with ('stomach-up' CDH) to those without ('stomach-down' CDH) intrathoracic stomach herniation. METHODS: Infants with left CDH who underwent prenatal US and postnatal repair at our institution between January 2008 and March 2017 were eligible for inclusion in this retrospective study. Detailed prenatal US examinations, fetal magnetic resonance imaging (MRI) studies, operative reports and medical records of infants enrolled in the pulmonary hypoplasia program at our institution were reviewed. Cases with liver herniation and those with an additional anomaly were excluded. Cases in which bowel loops were identified within the fetal chest on US while the stomach was intra-abdominal were categorized as having stomach-down CDH. Cases in which bowel loops and the stomach were visualized within the fetal chest on US were categorized as having stomach-up CDH. Prenatal imaging findings and postnatal outcomes were compared between the two groups. RESULTS: In total, 152 patients with left CDH were initially eligible for inclusion. Seventy-eight patients had surgically confirmed liver herniation and were excluded. Of the 74 included CDH cases without liver herniation, 28 (37.8%) had stomach-down CDH and 46 (62.2%) had stomach-up CDH. Of the 28 stomach-down CDH cases, 10 (35.7%) were referred for a suspected lung lesion. Sixty-eight (91.9%) cases had postnatal outcome data available for analysis. There was no significant difference in median observed-to-expected (o/e) lung-area-to-head-circumference ratio (LHR) between cases with stomach-down CDH and those with stomach-up CDH (41.5% vs 38.4%; P = 0.41). Furthermore, there was no difference in median MRI o/e total lung volume (TLV) between the two groups (49.5% vs 44.0%; P = 0.22). Compared with stomach-up CDH patients, stomach-down CDH patients demonstrated lower median duration of intubation (18 days vs 9.5 days; P < 0.01), median duration of extracorporeal membrane oxygenation (495 h vs 223.5 h; P < 0.05), rate of supplemental oxygen requirement at 30 days of age (20/42 (47.6%) vs 3/26 (11.5%); P < 0.01) and rate of pulmonary hypertension at initial postnatal echocardiography (28/42 (66.7%) vs 9/26 (34.6%); P = 0.01). No neonatal death occurred in stomach-down CDH patients and one neonatal death was seen in a patient with intrathoracic stomach herniation. CONCLUSIONS: In infants with left CDH without liver herniation, despite similar o/e-LHR and o/e-TLV, those with stomach-down CDH have decreased neonatal morbidity compared to those with stomach herniation. Progressive or variable physiological distension of the stomach over the course of gestation may explain these findings. Stomach-down left CDH is mistaken for a lung mass in a substantial proportion of cases. Accurate prenatal US characterization of CDH is crucial for appropriate prenatal counseling and patient management. © 2021 International Society of Ultrasound in Obstetrics and Gynecology.

Anatomic, computed tomographic, and ultrasonographic assessment of the lymph nodes in presumed healthy adult cats: The head, neck, thorax, and forelimb.

Assessment of the lymph nodes is key in staging cancer patients. Descriptions of normal features of the feline lymph nodes using computed tomography (CT) and ultrasound (US) are limited. A prospective anatomic and comparative imaging study was performed. The frequency of identification and the size of the lymph nodes during gross pathology from 6 feline cadavers were compared to the images of lymph nodes from 30 presumed healthy adult cats obtained by CT and US. Measurements (length, width, and height) were compared among techniques. The CT and US features of the identified lymph nodes were also recorded. The frequency of identification of the lymph centers varied among techniques and individually. The mandibular lymph nodes were identified in 100% of the cadavers and in 100% of the healthy cats using CT and US. The medial retropharyngeal lymph nodes were identified in 100% of the cats using CT and US. The deep cervical lymph nodes were not visualized in the cadavers. The cranial mediastinal and tracheobronchial lymph nodes were not visualized using US. Lymph nodes showed a higher length on CT and higher width on US. The height was the most statistically significant variable measurement among techniques. On CT, lymph nodes were most frequently isoattenuating or slightly hypoattenuating to surrounding musculature, with homogeneous contrast enhancement. On US, most lymph nodes were isoechoic or hypoechoic to surrounding fat tissue. The lymph nodes were most frequently elongated or rounded.

Calibrated uncertainty estimation for interpretable proton computed tomography image correction using Bayesian deep learning.

Integrated-type proton computed tomography (pCT) measures proton stopping power ratio (SPR) images for proton therapy treatment planning, but its image quality is degraded due to noise and scatter. Although several correction methods have been proposed, techniques that include estimation of uncertainty are limited. This study proposes a novel uncertainty-aware pCT image correction method using a Bayesian convolutional neural network (BCNN). A DenseNet-based BCNN was constructed to predict both a corrected SPR image and its uncertainty from a noisy SPR image. A total 432 noisy SPR images of 6 non-anthropomorphic and 3 head phantoms were collected with Monte Carlo simulations, while true noise-free images were calculated with known geometric and chemical components. Heteroscedastic loss and deep ensemble techniques were performed to estimate aleatoric and epistemic uncertainties by training 25 unique BCNN models. 200-epoch end-to-end training was performed for each model independently. Feasibility of the predicted uncertainty was demonstrated after applying two post-hoc calibrations and calculating spot-specific path length uncertainty distribution. For evaluation, accuracy of head SPR images and water-equivalent thickness (WET) corrected by the trained BCNN models was compared with a conventional method and non-Bayesian CNN model. BCNN-corrected SPR images represent noise-free images with high accuracy. Mean absolute error in test data was improved from 0.263 for uncorrected images to 0.0538 for BCNN-corrected images. Moreover, the calibrated uncertainty represents accurate confidence levels, and the BCNN-corrected calibrated WET was more accurate than non-Bayesian CNN with high statistical significance. Computation time for calculating one image and its uncertainties with 25 BCNN models is 0.7 s with a consumer grade GPU. Our model is able to predict accurate pCT images as well as two types of uncertainty. These uncertainties will be useful to identify potential cause of SPR errors and develop a spot-specific range margin criterion, toward elaboration of uncertainty-guided proton therapy.

Automatic quality assessment for 2D fetal sonographic standard plane based on multitask learning.

ABSTRACT: The quality control of fetal sonographic (FS) images is essential for the correct biometric measurements and fetal anomaly diagnosis. However, quality control requires professional sonographers to perform and is often labor-intensive. To solve this problem, we propose an automatic image quality assessment scheme based on multitask learning to assist in FS image quality control. An essential criterion for FS image quality control is that all the essential anatomical structures in the section should appear full and remarkable with a clear boundary. Therefore, our scheme aims to identify those essential anatomical structures to judge whether an FS image is the standard image, which is achieved by 3 convolutional neural networks. The Feature Extraction Network aims to extract deep level features of FS images. Based on the extracted features, the Class Prediction Network determines whether the structure meets the standard and Region Proposal Network identifies its position. The scheme has been applied to 3 types of fetal sections, which are the head, abdominal, and heart. The experimental results show that our method can make a quality assessment of an FS image within less a second. Also, our method achieves competitive performance in both the segmentation and diagnosis compared with state-of-the-art methods.