ABSTRACT
PURPOSE: Differentiating pulmonary lymphoma from lung infections using CT images is challenging. Existing deep neural network-based lung CT classification models rely on 2D slices, lacking comprehensive information and requiring manual selection. 3D models that involve chunking compromise image information and struggle with parameter reduction, limiting performance. These limitations must be addressed to improve accuracy and practicality. METHODS: We propose a transformer sequential feature encoding structure to integrate multi-level information from complete CT images, inspired by the clinical practice of using a sequence of cross-sectional slices for diagnosis. We incorporate position encoding and cross-level long-range information fusion modules into the feature extraction CNN network for cross-sectional slices, ensuring high-precision feature extraction. RESULTS: We conducted comprehensive experiments on a dataset of 124 patients, with respective sizes of 64, 20 and 40 for training, validation and testing. The results of ablation experiments and comparative experiments demonstrated the effectiveness of our approach. Our method outperforms existing state-of-the-art methods in the 3D CT image classification problem of distinguishing between lung infections and pulmonary lymphoma, achieving an accuracy of 0.875, AUC of 0.953 and F1 score of 0.889. CONCLUSION: The experiments verified that our proposed position-enhanced transformer-based sequential feature encoding model is capable of effectively performing high-precision feature extraction and contextual feature fusion in the lungs. It enhances the ability of a standalone CNN network or transformer to extract features, thereby improving the classification performance. The source code is accessible at https://github.com/imchuyu/PTSFE .
ABSTRACT
The research was designed to use computed tomography (CT) with 3D-CT reconstruction imaging techniques and the various anatomical sections-plana transversalia, frontalis, and dorsalia-to describe the anatomical architecture of the Zebu cattle head. Our study used nine mature heads. The CT bone window created detailed images of cranial bones, mandibles, teeth, and hyoid bones. All of the head cavities were evaluated, including the cranial, orbital, oral, auricular, and nasal cavities with their paranasal and conchal sinuses. The septum nasi, attached to the vomer and maxillary bones, did not reach the nasal cavity floor caudally at the level of the second premolar teeth, resulting in a single median channel from the choanae to the nasopharynx. The positions, boundaries, and connections of the paranasal sinuses were clearly identified. There were four nasal conchal sinuses (that were named the dorsal, middle, ethmoidal, and ventral) and five paranasal sinuses that were described as the following: sinus frontalis, maxillaris, palatinorum, and lacrimalis, as defined in the different anatomical sections and computed tomographic images. The complicated sinus frontalis caused the pneumatization of all bones that surrounded the cranial cavity, with the exception of the ethmoidal and body of basisphenoid bones. The sinus maxillaris was connected to the sinus lacrimalis and palatinorum through the maxillolacrimal and palatomaxillary openings, and to the middle nasal meatus through the nasomaxillary opening. Our findings provide a detailed anatomical knowledge for disease diagnosis to internal medicine veterinarians and surgeons by offering a comprehensive atlas of the Zebu cattle anatomy.
Subject(s)
Head , Imaging, Three-Dimensional , Tomography, X-Ray Computed , Animals , Cattle/anatomy & histology , Tomography, X-Ray Computed/veterinary , Imaging, Three-Dimensional/veterinary , Head/anatomy & histology , Head/diagnostic imaging , Skull/anatomy & histology , Skull/diagnostic imaging , Male , FemaleABSTRACT
Background: Based on the results of JCOG0802 and CALGB studies, segmentectomy has considered to be a standard procedure for early-stage non-small cell lung cancer (NSCLC). After lobectomy, the residual cavity is filled with mediastinal and diaphragmatic deviations, and compensatory volume changes are present in the residual lungs. In this study, we examined the efficacy of segmentectomy, a surgical procedure, by focusing on its impact on postoperative lung volume and function. Methods: We enrolled 77 patients who underwent segmentectomy as their initial surgical procedure, excluding those with additional lung resections and those who lacked postoperative computed tomography imaging. The predicted residual volume (mL) was defined as the total lung volume before surgery minus the volume of the resected area. Using the predicted residual volume (mL) and postoperative total lung volume (mL), we calculated the rate of postoperative lung volume increase [(postoperative total lung volume/predicted residual volume) × 100] (%). We also classified 52 cases with a rate of postoperative lung volume increase of ≥100% into a compensatory group, while those with a rate of <100% were classified into a non-compensatory group. Results: The average postoperative lung volume increase was 104.6% among 77 cases. Age ≥65 years, pack year index ≥27.5, ≥3 resected segments, and use of electrocautery for intersegmental plane division were significantly associated with compensatory group classification. In 20 compensatory cases with preoperative and postoperative pulmonary function tests, postoperative vital capacity and forced expiratory volume in one second values exceeded the preoperative predictions. This study further examined the areas responsible for postoperative compensatory lung volume increase. In the compensatory group, significant expansion was observed in the ipsilateral lobes, excluding the resected segment and contralateral lung, while no significant changes were noted in the volume of the lobe, including the resected segment. Conversely, the non-compensatory group showed a significant volume decrease in the resected lobe, but no significant increase in other areas. Conclusions: This study emphasizes the importance of preserving lung segments in segmentectomy. The study demonstrates extensive compensatory volume changes in the ipsilateral lung and contralateral lung. There was no significant volume decrease in any residual segment. This underlines the potential of segmentectomy to maintain lung function and expand treatment options post-surgery. In addition, the compensated group included patients with a lower pack-year index and younger patients. These results suggest that postoperative compensatory lung expansion includes not only hyperinflation of the remaining lung, but also an increase in the functional lung parenchyma.
ABSTRACT
Introduction: Congenital bronchial atresia (CBA), as a rare developmental abnormality of the lung, is usually asymptomatic and is accidently discovered in most cases. Currently, no standardized guidelines for the treatment or management of CBA have been established. Case presentation: A 22-year-old male soldier was referred to Shanghai Changhai Hospital, The First Affiliated Hospital of Naval Medical University due to chest tightness and shortness of breath after repeated strenuous activities. Contrast-enhanced computed tomography (CT) revealed an 18mm × 11mm solitary, well-circumscribed, and solid nodule with no enhancement in the right upper lobe (RUL), and emphysematous changes distributed throughout the RUL. A flexible bronchoscopic examination showed extrinsic compression stenosis in the bronchial opening of the right middle lobe (RML). After three-dimensional (3D) reconstruction CT and a multidisciplinary consultation, a diagnosis of CBA in the anterior segment (B3) of RUL was established. Subsequently, thoracoscopic right upper lobectomy was performed and resulted in an improved respiratory capacity 6 months after surgery. To date, the patient has good quality of life without any complication. Conclusion: This study underscores the role of bronchoscopy, 3D reconstruction CT, and a multidisciplinary consultation in the diagnosis of CBA, and highlights that a thoracoscopic intervention should be considered in such case.
ABSTRACT
Objectives: Precise segmentation of Odontogenic Cystic Lesions (OCLs) from dental Cone-Beam Computed Tomography (CBCT) is critical for effective dental diagnosis. Although supervised learning methods have shown practical diagnostic results in segmenting various diseases, their ability to segment OCLs covering different sub-class varieties has not been extensively investigated. Methods: In this study, we propose a new supervised learning method termed OCL-Net that combines a Multi-Scaled U-Net model, along with an Auto-Adapting mechanism trained with a combined supervised loss. Anonymous CBCT images were collected retrospectively from one hospital. To assess the ability of our model to improve the diagnostic efficiency of maxillofacial surgeons, we conducted a diagnostic assessment where 7 clinicians were included to perform the diagnostic process with and without the assistance of auto-segmentation masks. Results: We collected 300 anonymous CBCT images which were manually annotated for segmentation masks. Extensive experiments demonstrate the effectiveness of our OCL-Net for CBCT OCLs segmentation, achieving an overall Dice score of 88.84%, an IoU score of 81.23%, and an AUC score of 92.37%. Through our diagnostic assessment, we found that when clinicians were assisted with segmentation labels from OCL-Net, their average diagnostic accuracy increased from 53.21% to 55.71%, while the average time spent significantly decreased from 101s to 47s (P<0.05). Conclusion: The findings demonstrate the potential of our approach as a robust auto-segmentation system on OCLs in CBCT images, while the segmented masks can be used to further improve OCLs dental diagnostic efficiency.
ABSTRACT
Traumatic upper extremity injuries are a common cause of emergency department visits, comprising between 10-30% of traumatic injury visits. Timely and accurate evaluation is important to prevent severe complications such as permanent deformities, ischemia, or even death. Computed tomography (CT) and CT angiography (CTA) are the favored non-invasive imaging techniques for assessing upper extremity trauma, playing a crucial role in both the treatment planning and decision-making processes for such injuries. In CT postprocessing, a novel 3D rendering method, cinematic rendering (CR), employs sophisticated lighting models that simulate the interaction of multiple photons with the volumetric dataset. This technique produces images with realistic shadows and improved surface detail, surpassing the capabilities of volume rendering (VR) or maximal intensity projection (MIP). Considering the benefits of CR, we demonstrate its use and ability to achieve photorealistic anatomic visualization in a series of 11 cases where patients presented with traumatic upper extremity injuries, including bone, vascular, and skin/soft tissue injuries, adding to diagnostic confidence and intervention planning.
ABSTRACT
INTRODUCTION: The right intersectional plane and the right hepatic hilum were noted too often exhibit anatomical variations, making difficult the laparoscopic right anterior sectionectomy (LRAS). METHODS: We analyzed the anatomical features employing 3D-CT images of 55 patients, and evaluated these features according to the course of ventral branches of segment VI of the portal vein (PV, P6a) relative to the right hepatic vein (RHV). RESULTS: P6a run on the dorsal side of RHV in 32 patients (58%, Dorsal-P6a) and the ventral side of RHV in 23 (42%, Ventral-P6a). Ventral-P6a had more patients with S6 partially drained by middle hepatic vein (MHV, 39% vs. 0%, P < 0001), the narrower angle between the anterior and posterior branches of PV (73.1° vs. 93.8°, P = 0.006), the wider angle between the RHV and inferior vena cava (54.3° vs. 44.3°, P < 0.001), and more steeply pitched angle between S6 and S7 along the RHV (140.6° vs. 162.0°, P < 0.001) compared to Dorsal-P6a. CONCLUSION: In LRAS for Dorsal-P6a patients, the transection surface was relatively flat. In LRAS for Ventral-P6a patients, the narrow space between anterior and posterior glissons makes difficult the glissonean approach. The transection plane was steeply pitched, and RHV was partially exposed. S6 was often partially drained to MHV in 39% of the Ventral-P6a patients, which triggers congestion during liver transection of a right intersectional plane after first splitting the confluence of this branch.
Subject(s)
Hepatectomy , Hepatic Veins , Imaging, Three-Dimensional , Laparoscopy , Portal Vein , Tomography, X-Ray Computed , Humans , Portal Vein/surgery , Portal Vein/anatomy & histology , Portal Vein/diagnostic imaging , Female , Hepatic Veins/diagnostic imaging , Hepatic Veins/anatomy & histology , Hepatic Veins/surgery , Male , Laparoscopy/methods , Middle Aged , Hepatectomy/methods , Aged , Adult , Retrospective StudiesABSTRACT
The inguinal region, specifically the femoral vasculature, is a commonly used site of injection for intravenous drug users (IVDU). Repeated puncture of the vessel wall results in breakdown and subsequent arterial pseudoaneurysm- dilatations or outpouching of blood vessels, which, if left untreated, can result in fatal complications such as rupture with hemorrhage, sepsis, or even limb loss. The current modalities for arterial pseudoaneurysms include Doppler ultrasound and computed tomography (CT) angiography, both of which play important roles in management and surgical planning. However, 3D cinematic rendering (CR), a novel CT post-processing technique, offers timely, highly detailed photorealistic images that more clearly display the relation of anatomical structures, allowing for greater diagnostic confidence and precise surgical planning, particularly useful in the emergency setting. In this pictorial review, we demonstrate role of 3D CR in diagnosis and management of femoral pseudoaneurysms in IVDU through 9 illustrative cases.
Subject(s)
Aneurysm, False , Femoral Artery , Imaging, Three-Dimensional , Substance Abuse, Intravenous , Humans , Aneurysm, False/diagnostic imaging , Femoral Artery/diagnostic imaging , Substance Abuse, Intravenous/complications , Substance Abuse, Intravenous/diagnostic imaging , Computed Tomography Angiography/methods , Contrast Media , Adult , Male , FemaleABSTRACT
Several variations of pulmonary vein (PV) branching patterns exist. Since robot-assisted thoracoscopic surgery (RATS) is performed with magnified vision, it is crucial to carefully identify the running pattern of blood vessels before and during surgery. We present a case of a 77-year-old male patient with right lower lobe lung cancer. Right lower lobectomy via RATS was scheduled. Chest CT before surgery confirmed that the middle lobe PV (V4+5) merged with the inferior PV. Three-dimensional multidetector CT (3D-MDCT) subsequently confirmed that not only V4+5 but also the posterior segmental vein of the upper lobe (V2) merged with the inferior PV. We should have taped the lower lobe PV only, but we also taped the V2 and the middle lobe vein. However, since the oblique fissure was separated before cutting the taped blood vessel, the cutting of the blood vessel to be preserved was avoided. Surgeons should have a detailed understanding of the running patterns of pulmonary blood vessels before surgery to perform the procedure safely. Preoperative 3D-MDCT is useful for identifying the running pattern of blood vessels. An abnormality involving V2 and V4+5 merging into the inferior PV can also occur; hence, during right lower lobe resection, by dividing the lower lobe PV after the oblique fissure division, the surgeon can avoid unexpected transection of anomalous PVs that should be preserved.
ABSTRACT
In our case, the malformation was diagnosed prenatally at 40 weeks of gestation, and at the age of 14 days, the malformation was removed combined with a segmentectomy of the sixth segment of the left lung. Preoperative diagnostics focus on 3D-CT reconstruction and detailing of the anatomical variations of all arterial and venous vessels, as evident from our case. Treatment includes surgical removal or a minimally invasive interventional approach through the embolization of the vessel afferent to the malformation. After the operation, the child was discharged on the 30th day after birth in good condition and is developing normally. Early operative intervention is of great importance for the favorable outcome of the condition. In our case, this was hypoxemia with a saturation of 70-75%. The rare and often missed prenatal diagnosis of fetal AV malformation is significant for the adequate postnatal treatment and development of affected children.
ABSTRACT
Depressed skull fractures occur when broken bones displace inward, meaning that a portion of the outer table of the fracture line lies below the normal anatomical position of the inner table. They typically result from force trauma, when the skull is struck by an object with a moderately large amount of kinetic energy but a small surface area, or when an object with a large amount of kinetic energy impacts only a small area of the skull. In the present case, a depressed fracture of the frontal bone was detected at the autopsy of a 52-year-old man who, according to the belated confession of the assailant, was kicked in the head. The assailant was wearing sneakers. Could such a fracture be caused "just" by a kick? In this case it was possible due to an extraordinarily thin cranial vault (0.2 cm frontal, 0.3 cm occipital), which allowed the fractures to occur from a kinetic force that might not have been sufficient with a normal cranial vault thickness. An important role in the forensic analysis of the case was played by the 3D CT reconstruction.
Subject(s)
Imaging, Three-Dimensional , Skull Fracture, Depressed , Tomography, X-Ray Computed , Humans , Male , Middle Aged , Skull Fracture, Depressed/diagnostic imaging , Skull Fracture, Depressed/pathology , Autopsy/methods , Skull/diagnostic imaging , Skull/injuries , Skull/pathology , Skull Fractures/diagnostic imaging , Skull Fractures/pathology , Forensic Pathology/methodsABSTRACT
Accurate multi-organ segmentation in 3D CT images is imperative for enhancing computer-aided diagnosis and radiotherapy planning. However, current deep learning-based methods for 3D multi-organ segmentation face challenges such as the need for labor-intensive manual pixel-level annotations and high hardware resource demands, especially regarding GPU resources. To address these issues, we propose a 3D proxy-bridged region-growing framework specifically designed for the segmentation of the liver and spleen. Specifically, a key slice is selected from each 3D volume according to the corresponding intensity histogram. Subsequently, a deep learning model is employed to pinpoint the semantic central patch on this key slice, to calculate the growing seed. To counteract the impact of noise, segmentation of the liver and spleen is conducted on superpixel images created through proxy-bridging strategy. The segmentation process is then extended to adjacent slices by applying the same methodology iteratively, culminating in the comprehensive segmentation results. Experimental results demonstrate that the proposed framework accomplishes segmentation of the liver and spleen with an average Dice Similarity Coefficient of approximately 0.93 and a Jaccard Similarity Coefficient of around 0.88. These outcomes substantiate the framework's capability to achieve performance on par with that of deep learning methods, albeit requiring less guidance information and lower GPU resources.
Subject(s)
Deep Learning , Imaging, Three-Dimensional , Liver , Spleen , Tomography, X-Ray Computed , Liver/diagnostic imaging , Spleen/diagnostic imaging , Spleen/anatomy & histology , Humans , Imaging, Three-Dimensional/methods , Tomography, X-Ray Computed/methods , Image Processing, Computer-Assisted/methods , AlgorithmsABSTRACT
Surgical procedures to correct hip dysplasia associated with subluxation or dislocation of the femoral head are complex. The 3D geometric abnormalities of the acetabulum and proximal femur vary across patients. We, therefore, suggest a patient-specific surgical treatment involving computer-assisted 3D planning of the peri-acetabular osteotomies, taking into account the femoral head position; 3D printing of patient-specific guides for the cuts, repositioning, and fixation; and intra-operative application of the simulated displacements with their fixation. LEVEL OF EVIDENCE: IV.
Subject(s)
Acetabulum , Imaging, Three-Dimensional , Osteotomy , Printing, Three-Dimensional , Surgery, Computer-Assisted , Humans , Osteotomy/methods , Acetabulum/surgery , Acetabulum/diagnostic imaging , Child , Adolescent , Surgery, Computer-Assisted/methods , Male , Female , Preoperative Care/methods , Hip Dislocation/surgery , Hip Dislocation/diagnostic imaging , Hip Dislocation, Congenital/surgery , Hip Dislocation, Congenital/diagnostic imaging , Tomography, X-Ray ComputedABSTRACT
Several code-based methods have been created for comparing the frontal sinus in skeletal identification scenarios. However, little is known regarding matched-pair accuracy rates of these methods or how varying image modalities may affect these rates. The goals of this study were to validate the exclusion rates and to establish matched-pair accuracy rates of two well-cited coding methods, Cameriere et al. [23] and Tatlisumak et al. [24]. Additionally, individual variables were assessed for consistency in scoring between image modalities. Using a sample of U.S. African American, Native American, and European American females and males (n = 225), we examined individual variable scoring and string codes between two different image modalities (radiographs and CT-based 3D models). Arcades showed poor scoring consistency between modalities (p < 0.001). Although exclusion rates were similar to those reported in the original studies (93%-96%), matched-pair accuracy rates were low (13%-18%). None of the demographics (collection, sex, age, ancestry, and orientation) had an effect on the odds of a match. Interobserver and intraobserver analyses showed moderate to near-perfect agreement for all variables except supraorbital cells, which had minimal to no agreement. Currently, we do not recommend the application of these frontal sinus coding methods independent of other supporting identification methods given low variable consistency and accuracy rates. Visual identification should still be used to include or exclude an identification when using the frontal sinus.
Subject(s)
Forensic Anthropology , Frontal Sinus , Imaging, Three-Dimensional , Tomography, X-Ray Computed , Humans , Frontal Sinus/diagnostic imaging , Frontal Sinus/anatomy & histology , Male , Forensic Anthropology/methods , Female , Adult , Young Adult , Racial Groups , Middle Aged , Adolescent , Observer Variation , AgedSubject(s)
Calcinosis , Hysterectomy , Female , Humans , Calcinosis/diagnostic imaging , Postoperative Complications/diagnosisABSTRACT
PURPOSE: The aim of the study reveals a new intuitive method for preoperatively assessing defect ratio in glenoid deficiency based on the native glenoid width and the bare spot. METHODS: A linear relationship, i.e. the rh formula, between the native glenoid width (2r) and height (h) was revealed by a cadaver cohort (n = 204). To validate the reliability of the rh formula, 280 3D-CT images of intact glenoids were recruited. To evaluate the accuracy of rh formula in estimating glenoid defect, the 65 anterior-inferior defect models were artificially established based on the 3D-CT images of intact glenoids. Moreover, a clinically common anterior-posterior (AP) method was compared with the rh formula, to verify the technical superiority of rh formula. RESULTS: The regression analysis indicated a linear relationship between the width and height of intact glenoid: 2r = 0.768 × h - 1.222 mm (R2 = 0.820, p < 0.001). An excellent reliability was found between the formula prediction and model width (ICC = 0.911, p = 0.266). An excellent agreement was found between the predicted values and model parameters (glenoid width, ICCrh = 0.967, prh = 0.778; defect ratio, prh = 0.572, ICCrh = 0.997). And, it is of higher accuracy compared to the AP method (glenoid width, ICCAP = 0.933, pAP = 0.001; defect ratio, ICCAP = 0.911, pAP = 0.033). CONCLUSION: Applying the cadaver-based formula on 3D-CT scans accurately predicts native glenoid width and redefines bare spot for preoperatively determining glenoid bone loss.
Subject(s)
Bone Diseases, Metabolic , Glenoid Cavity , Joint Instability , Shoulder Joint , Humans , Shoulder Joint/diagnostic imaging , Shoulder Joint/surgery , Reproducibility of Results , Tomography, X-Ray Computed/methods , Scapula/diagnostic imaging , Cadaver , Glenoid Cavity/diagnostic imagingABSTRACT
The present case involved a 78-year-old woman with repeated recurrences of allergic bronchopulmonary mycosis (ABPM) who presented to our outpatient clinic with a chief complaint of dyspnoea with respiratory failure. Computed tomography (CT) of the chest showed atelectasis of the lower lobes due to mucus plugs. Blood and biochemical tests showed a high peripheral blood eosinophil count (1330/µL) and elevated immunoglobulin E (15,041 IU/mL; normal, < 361 IU/mL). Recurrent ABPM was diagnosed. The patient also showed chronic lower respiratory tract infection associated with Mycobacterium avium complex and Pseudomonas aeruginosa. First, we removed the mucus plug with a cryoprobe to avoid administering corticosteroids. However, subsequent 3-dimensional CT showed residual mucus plugs, so we administered dupilumab as an additional treatment. After initiating dupilumab, mucus plugs disappeared and respiratory failure resolved. We were able to implement multidisciplinary treatment that did not rely on corticosteroids.
ABSTRACT
PURPOSE: Accurately classifying displaced intraarticular calcaneal fractures (DIACFs) is essential for orthopedic surgeons to choose optimal treatment methods and provide results evaluation and communication. Many authors studying used Sanders classification reported moderate intra- and interobserver reliability. Taking the software opportunity of 3D virtual exarticulation, Goldzak updated French tri-dimensional Utheza classification, providing an alternative framework for classifying DIACFs. The aim of this study was to compare the intra- and interobserver reliability of Sanders versus Goldzak classification systems. METHODS: The CT scans of 30 patients with displaced intraarticular calcaneal fractures, treated in the same trauma center between 2014-2018, were analyzed by 16 medical doctors (specialists and residents in orthopedic surgery, specialists and residents in radiology), and classified according to Sanders and Goldzak classifications. The same images were sent on two separate sessions, in a randomized order. Interobserver reliability and intraobserver reproducibility were assessed using Kappa statistics and Gwet's AC1 coefficient. RESULTS: Interobserver reliability using Gwet reported a value of 0.36 for Goldzak classification and 0.30 for Sanders classification (corresponding to "fair assessment" in both cases). In absence of subclasses, "substantial assessment" was reported for Goldzak classification (Gwet of 0.61) and "moderate assessment" for Sanders classification (Gwet of 0.46). Goldzak system had a greater interobserver reliability in the group of radiology residents. Intraobserver reliability coefficient was 0.60 for Goldzak classification and 0.69 for Sanders classification, indicating a substantial agreement for both classifications. CONCLUSION: Despite the better view of the fracture lines provided by 3D reconstructions, this study failed to prove the superiority of Goldzak classification compared to Sanders classification for DIACFs.
Subject(s)
Ankle Injuries , Calcaneus , Foot Injuries , Fractures, Bone , Intra-Articular Fractures , Humans , Reproducibility of Results , Calcaneus/diagnostic imaging , Calcaneus/injuries , Observer Variation , Fractures, Bone/diagnostic imaging , Fractures, Bone/surgery , Intra-Articular Fractures/diagnostic imaging , Intra-Articular Fractures/surgeryABSTRACT
BACKGROUND: Reproducible methods for determining adequate bone densities for stemless anatomic total shoulder arthroplasty (aTSA) are currently lacking. The purpose of this study was to evaluate the utility of preoperative computed tomography (CT) imaging for assessing the bone density of the proximal humerus for supportive differentiation in the decision making for stemless humeral component implantation. It was hypothesized that preoperative 3-dimensional (3-D) CT bone density measures provide objective classifications of the bone quality for stemless aTSA. METHODS: A 3-part study was performed that included the analysis of cadaveric humerus CT scans followed by retrospective application to a clinical cohort and classification with a machine learning model. Thirty cadaveric humeri were evaluated with clinical CT and micro-CT (µCT) imaging. Phantom-calibrated CT data were used to extract 3-D regions of interest and defined radiographic scores. The final image processing script was applied retrospectively to a clinical cohort (n = 150) that had a preoperative CT and intraoperative bone density assessment using the "thumb test," followed by placement of an anatomic stemmed or stemless humeral component. Postscan patient-specific calibration was used to improve the functionality and accuracy of the density analysis. A machine learning model (Support vector machine [SVM]) was utilized to improve the classification of bone densities for a stemless humeral component. RESULTS: The image processing of clinical CT images demonstrated good to excellent accuracy for cylindrical cancellous bone densities (metaphysis [ICC = 0.986] and epiphysis [ICC = 0.883]). Patient-specific internal calibration significantly reduced biases and unwanted variance compared with standard HU CT scans (P < .0001). The SVM showed optimized prediction accuracy compared with conventional statistics with an accuracy of 73.9% and an AUC of 0.83 based on the intraoperative decision of the surgeon. The SVM model based on density clusters increased the accuracy of the bone quality classification to 87.3% with an AUC of 0.93. CONCLUSIONS: Preoperative CT imaging allows accurate evaluation of the bone densities in the proximal humerus. Three-dimensional regions of interest, rescaling using patient-specific calibration, and a machine learning model resulted in good to excellent prediction for objective bone quality classification. This approach may provide an objective tool extending preoperative selection criteria for stemless humeral component implantation.
Subject(s)
Arthroplasty, Replacement, Shoulder , Bone Density , Humerus , Imaging, Three-Dimensional , Tomography, X-Ray Computed , Humans , Arthroplasty, Replacement, Shoulder/methods , Tomography, X-Ray Computed/methods , Retrospective Studies , Male , Female , Humerus/diagnostic imaging , Humerus/surgery , Aged , Middle Aged , Cadaver , Preoperative Care/methods , Machine Learning , Shoulder Joint/diagnostic imaging , Shoulder Joint/surgery , Aged, 80 and overABSTRACT
Non-traumatic thoracic aorta emergencies are acute conditions associated with substantial morbidity and mortality. In the emergency setting, timely detection of aortic injury through radiological imaging is crucial for prompt treatment planning and favorable patient outcomes. 3D cinematic rendering (CR), a novel rendering algorithm for computed tomography (CT) image processing, allows for life-like visualization of spatial details and contours of highly complex anatomic structures such as the thoracic aorta and its vessels, generating a photorealistic view that not just adds to diagnostic confidence, but is especially useful for non-radiologists, including surgeons and emergency medicine physicians. In this pictorial review, we demonstrate the utility of CR in the setting of non-traumatic thoracic aorta emergencies through 10 cases that were processed at a standalone 3D CR station at the time of presentation, including its role in diagnosis and management.