A semiautomatic segmentation algorithm for extracting the complete structure of acini from synchrotron micro-CT images

Application of Machine Learning for Segmentation of the Pulmonary Acinus Imaged by Synchrotron X-Ray Tomography

Background: To assess the effectiveness of inhalation therapy, it is important to evaluate the lungs' structure; thus, visualization of the entire lungs at the level of the alveoli is necessary. To achieve this goal, the applied visualization technique must satisfy the following two conditions simultaneously: (1) it has to obtain images of the entire lungs, since one part of the lungs is influenced by the other parts, and (2) the images have to capture the detailed structure of the alveolus/acinus in which gas exchange occurs. However, current visualization techniques do not fulfill these two conditions simultaneously. Segmentation is a process in which each pixel of the obtained high-resolution images is simplified (i.e., the representation of an image is changed by categorizing and modifying each pixel) so that we can perform three-dimensional volume rendering. One of the bottlenecks of current approaches is that the accuracy of the segmentation of each image has to be evaluated on the outcome of the process (mainly by an expert). It is a formidable task to evaluate the astronomically large numbers of images that would be required to resolve the entire lungs in high resolution. Methods: To overcome this challenge, we propose a new approach based on machine learning (ML) techniques for the validation step. Results: We demonstrate the accuracy of the segmentation process itself by comparison with previously validated images. In this ML approach, to achieve a reasonable accuracy, millions/billions of parameters used for segmentation have to be optimized. This computationally demanding new approach is achievable only due to recent dramatic increases in computation power. Conclusion: The objective of this article is to explain the advantages of ML over the classical approach for acinar imaging.

Mathematical modeling of pulmonary acinus structure: Verification of acinar shape effects on pathway structure using rat lungs

The pulmonary acinus is the gas exchange unit in the lung and has a very complex microstructure. The structure model is essential to understand the relationship between structural heterogeneity and mechanical phenomena at the acinus level with computational approaches. We propose an acinus structure model represented by a cluster of truncated octahedra in conical, double-conical, inverted conical, or chestnut-like conical confinement to accommodate recent experimental information of rodent acinar shapes. The basis of the model is the combined use of Voronoi and Delaunay tessellations and the optimization of the ductal tree assuming the number of alveoli and the mean path length as quantities related to gas exchange. Before applying the Voronoi tessellation, controlling the seed coordinates enables us to model acinus with arbitrary shapes. Depending on the acinar shape, the distribution of path length varies. The lengths are more widely spread for the cone acinus, with a bias toward higher values, while most of the lengths for the inverted cone acinus primarily take a similar value. Longer pathways have smaller tortuosity and more generations, and duct length per generation is almost constant irrespective of generation, which agrees well with available experimental data. The pathway structure of cone and chestnut-like cone acini is similar to the surface acini's features reported in experiments. According to space-filling requirements in the lung, other conical acini may also be acceptable. The mathematical acinus structure model with various conical shapes can be a platform for computational studies on regional differences in lung functions along the lung surface, underlying respiratory physiology and pathophysiology.

Micrometer-resolution X-ray tomographic full-volume reconstruction of an intact post-mortem juvenile rat lung

In this article, we present an X-ray tomographic imaging method that is well suited for pulmonary disease studies in animal models to resolve the full pathway from gas intake to gas exchange. Current state-of-the-art synchrotron-based tomographic phase-contrast imaging methods allow for three-dimensional microscopic imaging data to be acquired non-destructively in scan times of the order of seconds with good soft tissue contrast. However, when studying multi-scale hierarchically structured objects, such as the mammalian lung, the overall sample size typically exceeds the field of view illuminated by the X-rays in a single scan and the necessity for achieving a high spatial resolution conflicts with the need to image the whole sample. Several image stitching and calibration techniques to achieve extended high-resolution fields of view have been reported, but those approaches tend to fail when imaging non-stable samples, thus precluding tomographic measurements of large biological samples, which are prone to degradation and motion during extended scan times. In this work, we demonstrate a full-volume three-dimensional reconstruction of an intact rat lung under immediate post-mortem conditions and at an isotropic voxel size of (2.75 µm)³. We present the methodology for collecting multiple local tomographies with 360° extended field of view scans followed by locally non-rigid volumetric stitching. Applied to the lung, it allows to resolve the entire pulmonary structure from the trachea down to the parenchyma in a single dataset. The complete dataset is available online (https://doi.org/10.16907/7eb141d3-11f1-47a6-9d0e-76f8832ed1b2).

How high resolution 3-dimensional imaging changes our understanding of postnatal lung development

During the last 10 + years biologically and clinically significant questions about postnatal lung development could be answered due to the application of modern cutting-edge microscopic and quantitative histological techniques. These are in particular synchrotron radiation based X-ray tomographic microscopy (SRXTM), but also ³Helium Magnetic Resonance Imaging, as well as the stereological estimation of the number of alveoli and the length of the free septal edge. First, the most important new finding may be the following: alveolarization of the lung does not cease after the maturation of the alveolar microvasculature but continues until young adulthood and, even more important, maybe reactivated lifelong if needed to rescue structural damages of the lungs. Second, the pulmonary acinus represents the functional unit of the lung. Because the borders of the acini could not be detected in classical histological sections, any investigation of the acini requires 3-dimensional (imaging) methods. Based on SRXTM it was shown that in rat lungs the number of acini stays constant, meaning that their volume increases by a factor of ~ 11 after birth. The latter is very important for acinar ventilation and particle deposition.

Early Lung Adenocarcinoma in Mice: Micro-Computed Tomography Manifestations and Correlation with Pathology

Lung cancer is the most common fatal malignancy for both men and women and adenocarcinoma is the most common histologic type. Early diagnosis of lung cancer can significantly improve the survival rate of patients. This study aimed to investigate the micro-computed tomography (micro-CT) manifestations of early lung adenocarcinoma (LAC) in mice and to provide a new perspective for early clinical diagnosis. Early LAC models in 10 mice were established by subcutaneously injecting 1-methyl-3-nitro-1-nitrosoguanidine (MNNG) solution. Micro-CT scan and multiple planar reconstruction (MPR) were used for mouse lungs. Micro-CT features of early LAC, especially the relationships between tumor and bronchus, were analyzed and correlated with pathology. Micro-CT findings of early LAC were divided into three types: non-solid (n = 8, 6%), partly solid (n = 85, 64%) and totally solid (n = 39, 30%). Tumor-bronchus relationships, which could be observed in 110 of 132(83%) LAC, were classified into four patterns: type I (n = 16, 15%), bronchus was truncated at the margin of the tumor; type II (n = 33, 30%), bronchus penetrated into the tumor with tapered narrowing and interruption; type III (n = 38, 35%), bronchus penetrated into the tumor with a patent and intact lumen; type IV (n = 99, 90%), bronchus ran at the border of the tumor with an intact or compressed lumen. Micro-CT manifestations of early LAC correlated well with pathological findings. Micro-CT can clearly demonstrate the features of mouse early LAC and bronchus-tumor relationships, and can also provide a new tool and perspective for the study of early LAC.

Visualization and stereological characterization of individual rat lung acini by high-resolution X-ray tomographic microscopy

The small trees of gas-exchanging pulmonary airways, which are fed by the most distal purely conducting airways, are called acini and represent the functional gas-exchanging units. The three-dimensional architecture of the acini has a strong influence on ventilation and particle deposition. Due to the difficulty in identifying individual acini on microscopic lung sections, the knowledge about the number of acini and their biological parameters, like volume, surface area, and number of alveoli per acinus, are limited. We developed a method to extract individual acini from lungs imaged by high-resolution synchrotron radiation-based X-ray tomographic microscopy and estimated their volume, surface area, and number of alveoli. Rat acini were isolated by semiautomatically closing the airways at the transition from conducting to gas-exchanging airways. We estimated a mean internal acinar volume of 1.148 mm3, a mean acinar surface area of 73.9 mm2, and a mean of 8,470 alveoli/acinus. Assuming that the acini are similarly sized throughout different regions of the lung, we calculated that a rat lung contains 5,470 ± 833 acini. We conclude that our novel approach is well suited for the fast and reliable characterization of a large number of individual acini in healthy, diseased, or transgenic lungs of different species, including humans.