Observing early metabolic changes in positron emission tomography (PET) is an essential tool to assess treatment efficiency in radiotherapy. However, for thoracic regions, the use of three-dimensional (3D) PET imaging is unfeasible because the radiotracer activity is smeared by the respiratory motion and averaged during the imaging acquisition process. This motion-induced degradation is similar in magnitude with the treatment-induced changes, and the two occurrences become indiscernible. We present a customized temporal-spatial deformable registration method for quantifying respiratory motion in a four-dimensional (4D) PET dataset. Once the motion is quantified, a motion-corrected (MC) dataset is created by tracking voxels to eliminate breathing-induced changes in the 4D imaging scan. The 4D voxel-tracking data is then summed to yield a 3D MC-PET scan containing only treatment-induced changes. This proof of concept is exemplified on both phantom and clinical data, where the proposed algorithm tracked the trajectories of individual points through the 4D datasets reducing motion to less than 4 mm in all phases. This correction approach using deformable registration can discern motion blurring from treatment-induced changes in treatment response assessment using PET imaging.
INTRODUCTION: Anatomic anterior cruciate ligament (ACL) reconstruction has proven to be a reliable method to restore knee stability. However, the risk of physeal arrest with transphyseal tunnel placement in skeletally immature patients has raised concern regarding this technique. Conservative nonoperative management also has its limitations resulting in meniscal and chondral damage that may lead to degenerative joint disease and poor return to sport. Researchers have used animal models to study the threshold of physeal damage producing growth deformity. The purpose of this study was to examine the distal femoral and proximal tibial physes and determine the damage produced by drilling transphyseal tunnels. In addition, we attempted to find a reproducible angle at which to drill the tibial tunnel for safe interference screw placement. To do this, we used a custom software module. METHODS: A custom software package designed by our team was used: Module for Adolescent ACL Reconstructive Surgery (MAARS). This module created a 3-dimensional model of the distal femur and proximal tibia. The data required for MAARS were sagittal and coronal T1 magnetic resonance imagings of at least 1.5T. Thirty-one knee magnetic resonance imaging studies from patients aged 10 to 15 years old were used. The physes were segmented out to obtain volumetric measurements. Transphyseal tunnels were simulated based on the anatomic trajectory of the native ACL. The module calculated volume of physis was removed with the use of an 8-mm tunnel and the optimum angle for trajectory. RESULTS: Average volume of the tibial and femoral physis was 12,683.1 μL and 14,708.3 μL, respectively. The volume increased linearly with age. Average volume removed from the tibial and femoral physis was 318.4 μL and 306.29 μL, respectively. This represented 2.4% of the distal femoral physis and 2.5% of the proximal tibial physis. The volume percent removed decreased linearly with age.Manipulation of the variables demonstrates graft radius is the most critical parameter affecting the volume of physeal injury. Variation of graft diameter from 6 mm to 11 mm will increase volume percent removed from 2.3% to 7.8%, which averages 1.1% for every 1 mm increase. Increasing tunnel drill angle from 45 degrees to 70 degrees will decrease volume percent removed from 4.1% to 3.1% which averages 0.2% removed for each 5 degrees increase in drill angle. The average angle to maintain a distance of 20 mm from the proximal tibial physis was 65 degrees with a range of 40 degrees to 85 degrees. DISCUSSION: Less than 3% injury occurs when drilling an 8-mm tunnel across the physis. A vertical tunnel has minimal effect, but the tunnel diameter is critical. Interference screws can be placed safely to avoid the physis but requires careful planning. The MAARS module may be helpful in preoperative planning. LEVEL OF EVIDENCE: Diagnostic, level IV.
We propose a simple method for reconstructing vascular trees from 3D images. Our algorithm extracts persistent maxima of the intensity on all axis-aligned 2D slices of the input image. The maxima concentrate along 1D intensity ridges, in particular along blood vessels. We build a forest connecting the persistent maxima with short edges. The forest tends to approximate the blood vessels present in the image, but also contains numerous spurious features and often fails to connect segments belonging to one vessel in low contrast areas. We improve the forest by applying simple geometric filters that trim short branches, fill gaps in blood vessels and remove spurious branches from the vascular tree to be extracted. Experiments show that our technique can be applied to extract coronary trees from heart CT scans.
by
Susan Smith;
James Bates;
S Shrestha;
Q Liu;
KC Oeffinger;
EJ Chow;
AC Gupta;
CA Owens;
LS Constine;
BS Hoppe;
WM Leisenring;
Y Qiao;
RE Weathers;
LE Court;
CC Pinnix;
SF Kry;
DA Mulrooney;
GT Armstrong;
Y Yasui;
RM Howell
Background and purpose: We previously evaluated late cardiac disease in long-term survivors in the Childhood Cancer Survivor Study (CCSS) based on heart radiation therapy (RT) doses estimated from an age-scaled phantom with a simple atlas-based heart model (HAtlas). We enhanced our phantom with a high-resolution CT-based anatomically realistic and validated age-scalable cardiac model (HHybrid). We aimed to evaluate how this update would impact our prior estimates of RT-related late cardiac disease risk in the CCSS cohort. Methods: We evaluated 24,214 survivors from the CCSS diagnosed from 1970 to 1999. RT fields were reconstructed on an age-scaled phantom with HHybrid and mean heart dose (Dm), percent volume receiving ≥ 20 Gy (V20) and ≥ 5 Gy with V20 = 0 (V5,V20=0%) were calculated. We reevaluated cumulative incidences and adjusted relative rates of grade 3–5 Common Terminology Criteria for Adverse Events outcomes for any cardiac disease, coronary artery disease (CAD), and heart failure (HF) in association with Dm, V20, and V5,V20=0% (as categorical variables). Dose-response relationships were evaluated using piecewise-exponential models, adjusting for attained age, sex, cancer diagnosis age, race/ethnicity, time-dependent smoking history, diagnosis year, and chemotherapy exposure and doses. For relative rates, Dm was also considered as a continuous variable. Results: Consistent with previous findings with HAtlas, reevaluation using HHybrid dosimetry found that, Dm ≥ 10 Gy, V20 ≥ 0.1%, and V5,V20=0% ≥ 50% were all associated with increased cumulative incidences and relative rates for any cardiac disease, CAD, and HF. While updated risk estimates were consistent with previous estimates overall without statistically significant changes, there were some important and significant (P < 0.05) increases in risk with updated dosimetry for Dm in the category of 20 to 29.9 Gy and V20 in the category of 30% to 79.9%. When changes in the linear dose–response relationship for Dm were assessed, the slopes of the dose response were steeper (P < 0.001) with updated dosimetry. Changes were primarily observed among individuals with chest-directed RT with prescribed doses ≥ 20 Gy. Conclusion: These findings present a methodological advancement in heart RT dosimetry with improved estimates of RT-related late cardiac disease risk. While results are broadly consistent with our prior study, we report that, with updated cardiac dosimetry, risks of cardiac disease are significantly higher in two dose and volume categories and slopes of the Dm-specific RT-response relationships are steeper. These data support the use of contemporary RT to achieve lower heart doses for pediatric patients, particularly those requiring chest-directed RT.
Total cavopulmonary connection is the result of a series of palliative surgical repairs performed on patients with single ventricle heart defects. The resulting anatomy has complex and unsteady hemodynamics characterized by flow mixing and flow separation. Although varying degrees of flow pulsatility have been observed in vivo, non-pulsatile (time-averaged) boundary conditions have traditionally been assumed in hemodynamic modeling, and only recently have pulsatile conditions been incorporated without completely characterizing their effect or importance. In this study, 3D numerical simulations with both pulsatile and non-pulsatile boundary conditions were performed for 24 patients with different anatomies and flow boundary conditions from Georgia Tech database. Flow structures, energy dissipation rates and pressure drops were compared under rest and simulated exercise conditions. It was found that flow pulsatility is the primary factor in determining the appropriate choice of boundary conditions, whereas the anatomic configuration and cardiac output had secondary effects. Results show that the hemodynamics can be strongly influenced by the presence of pulsatile flow. However, there was a minimum pulsatility threshold, identified by defining a weighted pulsatility index (wPI), above which the influence was significant. It was shown that when wPI < 30%, the relative error in hemodynamic predictions using time-averaged boundary conditions was less than 10% compared to pulsatile simulations. In addition, when wPI < 50, the relative error was less than 20%. A correlation was introduced to relate wPI to the relative error in predicting the flow metrics with non-pulsatile flow conditions.
This paper presents a review of deep learning (DL)-based medical image registration methods. We summarized the latest developments and applications of DL-based registration methods in the medical field. These methods were classified into seven categories according to their methods, functions and popularity. A detailed review of each category was presented, highlighting important contributions and identifying specific challenges. A short assessment was presented following the detailed review of each category to summarize its achievements and future potential. We provided a comprehensive comparison among DL-based methods for lung and brain registration using benchmark datasets. Lastly, we analyzed the statistics of all the cited works from various aspects, revealing the popularity and future trend of DL-based medical image registration.
To conduct a patient-specific computational modeling of the aortic valve, 3-D aortic valve anatomic geometries of an individual patient need to be reconstructed from clinical 3-D cardiac images. Currently, most of computational studies involve manual heart valve geometry reconstruction and manual finite element (FE) model generation, which is both time-consuming and prone to human errors. A seamless computational modeling framework, which can automate this process based on machine learning algorithms, is desirable, as it can not only eliminate human errors and ensure the consistency of the modeling results but also allow fast feedback to clinicians and permits a future population-based probabilistic analysis of large patient cohorts. In this study, we developed a novel computational modeling method to automatically reconstruct the 3-D geometries of the aortic valve from computed tomographic images. The reconstructed valve geometries have built-in mesh correspondence, which bridges harmonically for the consequent FE modeling. The proposed method was evaluated by comparing the reconstructed geometries from 10 patients with those manually created by human experts, and a mean discrepancy of 0.69 mm was obtained. Based on these reconstructed geometries, FE models of valve leaflets were developed, and aortic valve closure from end systole to middiastole was simulated for 7 patients and validated by comparing the deformed geometries with those manually created by human experts, and a mean discrepancy of 1.57 mm was obtained. The proposed method offers great potential to streamline the computational modeling process and enables the development of a preoperative planning system for aortic valve disease diagnosis and treatment.
RATIONALE AND OBJECTIVE: Three-dimensional (3D) printed anatomic models and surgical guides have been shown to reduce operative time. The purpose of this study was to generate an economic analysis of the cost-saving potential of 3D printed anatomic models and surgical guides in orthopedic and maxillofacial surgical applications. MATERIALS AND METHODS: A targeted literature search identified operating room cost-per-minute and studies that quantified time saved using 3D printed constructs. Studies that reported operative time differences due to 3D printed anatomic models or surgical guides were reviewed and cataloged. A mean of $62 per operating room minute (range of $22–$133 per minute) was used as the reference standard for operating room time cost. Different financial scenarios were modeled with the provided cost-per-minute of operating room time (using high, mean, and low values) and mean time saved using 3D printed constructs. RESULTS: Seven studies using 3D printed anatomic models in surgical care demonstrated a mean 62 minutes ($3720/case saved from reduced time) of time saved, and 25 studies of 3D printed surgical guides demonstrated a mean 23 minutes time saved ($1488/case saved from reduced time). An estimated 63 models or guides per year (or 1.2/week) were predicted to be the minimum number to breakeven and account for annual fixed costs. CONCLUSION: Based on the literature-based financial analyses, medical 3D printing appears to reduce operating room costs secondary to shortening procedure times. While resource-intensive, 3D printed constructs used in patients’ operative care provides considerable downstream value to health systems.
Skeletal muscle has a remarkable regeneration capacity to recover its structure and function after injury, except for the traumatic loss of critical muscle volume, called volumetric muscle loss (VML). Although many extremity VML models have been conducted, craniofacial VML has not been well-studied due to unavailable in vivo assay tools. Here, this paper reports a wireless, noninvasive nanomembrane system that integrates skin-wearable printed sensors and electronics for real-time, continuous monitoring of VML on craniofacial muscles. The craniofacial VML model, using biopsy punch-induced masseter muscle injury, shows impaired muscle regeneration. To measure the electrophysiology of small and round masseter muscles of active mice during mastication, a wearable nanomembrane system with stretchable graphene sensors that can be laminated to the skin over target muscles is utilized. The noninvasive system provides highly sensitive electromyogram detection on masseter muscles with or without VML injury. Furthermore, it is demonstrated that the wireless sensor can monitor the recovery after transplantation surgery for craniofacial VML. Overall, the presented study shows the enormous potential of the masseter muscle VML injury model and wearable assay tool for the mechanism study and the therapeutic development of craniofacial VML.
The problem of deciphering how low-level patterns (action potentials in the brain, amino acids in a protein, etc.) drive high-level biological features (sensorimotor behavior, enzymatic function) represents the central challenge of quantitative biology. The lack of general methods for doing so from the size of datasets that can be collected experimentally severely limits our understanding of the biological world. For example, in neuroscience, some sensory and motor codes have been shown to consist of precisely timed multi-spike patterns. However, the combinatorial complexity of such pattern codes have precluded development of methods for their comprehensive analysis. Thus, just as it is hard to predict a protein’s function based on its sequence, we still do not understand how to accurately predict an organism’s behavior based on neural activity. Here we introduce the unsupervised Bayesian Ising Approximation (uBIA) for solving this class of problems. We demonstrate its utility in an application to neural data, detecting precisely timed spike patterns that code for specific motor behaviors in a songbird vocal system. In data recorded during singing from neurons in a vocal control region, our method detects such codewords with an arbitrary number of spikes, does so from small data sets, and accounts for dependencies in occurrences of codewords. Detecting such comprehensive motor control dictionaries can improve our understanding of skilled motor control and the neural bases of sensorimotor learning in animals. To further illustrate the utility of uBIA, used it to identify the distinct sets of activity patterns that encode vocal motor exploration versus typical song production. Crucially, our method can be used not only for analysis of neural systems, but also for understanding the structure of correlations in other biological and nonbiological datasets.