Nikhil Sindhwani

Immediate postoperative changes in synthetic meshes – In vivo measurements

BACKGROUND AND OBJECTIVE Immediate post-operative structural changes in implanted synthetic meshes are believed to contribute to graft related complications. Our aim was to observe in vivo dimensional changes at the pore level. METHOD Two different polyvinylidine fluoride (PVDF) meshes, CICAT and ENDOLAP (Dynamesh, FEG Textiltechnik) were implanted in 18 female Sprague Dawley (n=9/group). The meshes (30×25mm(2)) were overlaid on a full thickness incision (2×1cm(2)) and sutured on the abdominal wall. All animals underwent microCT imaging (res. 35µm/px) at day 1 and 15 postsurgery. A customized procedure was developed to semi-automatically detect the pore centers from the microCT dataset. Horizontal (transverse) and vertical (cranio-caudal) inter-pore distances were then recorded. The overall mesh dimensions were also noted from 3D models generated from in vivo microCT datasets. Inter-pore distances and the overall dimensions from microCT images of the meshes set in agarose gel phantom were used as controls. Mann-Whitney U test was done to check for significant differences. RESULTS Number of measurable vertical and horizontal inter-pore distances was 56.5(10.5) and 54.5(14.5) [median (IQR)] per animal. At day 1, we observed a 4.3% (CICAT) and 4.6% (ENDOLAP) increase in vertical inter-pore distance when compared to controls (p<0.001, p=0.003, respectively). Measurements fell back to phantom values by day 15 (3.7% and 4.9% decrease compared to day 1, p<0.001 for both). The horizontal inter-pore distances for ENDOLAP increased by 1.4% (p=0.003) during the two weeks period. The overall mesh dimensions did not change significantly day 1 and day 15. The in vivo measurement of the overall mesh dimensions demonstrated a 15.9% reduction in mesh area as compared to that in phantom controls. CONCLUSION We report for the first time, in vivo changes in pore dimensions of a textile implant. This study clearly demonstrates the dynamic nature of a textile implant during the tissue integration process. For studied PVDF meshes, the process of tissue integration leads to limited but significant reduction over time as observed at the pore level. Remarkably the extent of this reduction does not account for the change in overall mesh dimensions.

Cell-based secondary prevention of childbirth-induced pelvic floor trauma

With advancing population age, pelvic-floor dysfunction (PFD) will affect an increasing number of women. Many of these women wish to maintain active lifestyles, indicating an urgent need for effective strategies to treat or, preferably, prevent the occurrence of PFD. Childbirth and pregnancy have both long been recognized as crucial contributing factors in the pathophysiology of PFD. Vaginal delivery of a child is a serious traumatic event, causing anatomical and functional changes in the pelvic floor. Similar changes to those experienced during childbirth can be found in symptomatic women, often many years after delivery. Thus, women with such PFD symptoms might have incompletely recovered from the trauma caused by vaginal delivery. This hypothesis creates the possibility that preventive measures can be initiated around the time of delivery. Secondary prevention has been shown to be beneficial in patients with many other chronic conditions. The current general consensus is that clinicians should aim to minimize the extent of damage during delivery, and aim to optimize healing processes after delivery, therefore preventing later dysfunction. A substantial amount of research investigating the potential of stem-cell injections as a therapeutic strategy for achieving this purpose is currently ongoing. Data from small animal models have demonstrated positive effects of mesenchymal stem-cell injections on the healing process following simulated vaginal birth injury.

In vivo evidence of significant levator ani muscle stretch on MR images of a live childbirth

OBJECTIVE: Vaginal childbirth is believed to be a significant risk factor for the development of pelvic floor dysfunction later in life. Previous studies have explored the use of medical imaging and simulations of childbirth to determine the stretch in the levator ani muscle. A report in 2012 has recorded magnetic resonance images of a live childbirth of a 24 year old woman giving birth vaginally for the second time, using a 1.0 Tesla open, high‐field scanner. Our objective was to determine the stretch ratios in the levator muscle using these magnetic resonance images of live childbirth. STUDY DESIGN: Three‐dimensional magnetic resonance image sequences were obtained to visualize coronal and axial planes before and after the childbirth. These images were obtained before the expulsion phase without pushing and were used to reconstruct the levator muscle and the fetal head in 3 dimensions. The fetal head was approximated to be an ellipsoid, and it is assumed that its middle section is visible in dynamic magnetic resonance images. Assuming incompressibility, the full deformation field of the fetal head is then calculated. Real‐time cine magnetic resonance images were acquired for the during the expulsion phase, occurring over 2 contractions in the midsagittal plane. The levator muscle stretch is estimated using a custom program. The program calculates points of contact between the fetal head ellipsoid and the levator ani muscle model as the head descends down the birth canal and moves them orthogonal to its surface. Circumferential stretch was calculated to represent the extension needed to allow the passage of the fetal head. RESULTS: Starting from a position in the preexpulsion phase, the levator muscle experiences a maximum circumferential stretch of 248% on the posterior‐medial portion of the levator ani muscle, as shown in previously published finite element simulations. However, the maximal stretch was notably less than that predicted by finite element models. This is because our baseline 3‐dimensional model of the levator muscle is created from images taken shortly before expulsion and thus is already in a stretched state. Furthermore, the finite element models are created from images of a healthy nulliparous woman, while this study uses images from a para 2 woman. CONCLUSION: This study is the first attempt to estimate the stretch in levator ani muscle using magnetic resonance images of a live childbirth. The stretch was significant and the locations corroborate with previous findings of finite element models.

Semi‐automatic outlining of levator hiatus

To create a semi‐automated outlining tool for the levator hiatus, to reduce interobserver variability and and speed up analysis.

Automatic segmentation method of pelvic floor levator hiatus in ultrasound using a self-normalizing neural network

Abstract. Segmentation of the levator hiatus in ultrasound allows the extraction of biometrics, which are of importance for pelvic floor disorder assessment. We present a fully automatic method using a convolutional neural network (CNN) to outline the levator hiatus in a two-dimensional image extracted from a three-dimensional ultrasound volume. In particular, our method uses a recently developed scaled exponential linear unit (SELU) as a nonlinear self-normalizing activation function, which for the first time has been applied in medical imaging with CNN. SELU has important advantages such as being parameter-free and mini-batch independent, which may help to overcome memory constraints during training. A dataset with 91 images from 35 patients during Valsalva, contraction, and rest, all labeled by three operators, is used for training and evaluation in a leave-one-patient-out cross validation. Results show a median Dice similarity coefficient of 0.90 with an interquartile range of 0.08, with equivalent performance to the three operators (with a Williams’ index of 1.03), and outperforming a U-Net architecture without the need for batch normalization. We conclude that the proposed fully automatic method achieved equivalent accuracy in segmenting the pelvic floor levator hiatus compared to a previous semiautomatic approach.

Evaluating fetal head dimension changes during labor using open magnetic resonance imaging.

Abstract Aim: Fetal skull molding is important for the adaptation of the head to the birth canal during vaginal delivery. Importantly, the fetal head must rotate around the maternal symphysis pubis. The goals of this analysis were to observe a human birth in real-time using an open magnetic resonance imaging (MRI) scanner and describe the fetal head configuration during expulsion. Methods: Real-time cinematic MRI series (TSE single-shot sequence, TR 1600 ms, TE 150 ms) were acquired from the midsagittal plane of the maternal pelvis during the active second stage of labor at 37 weeks of gestation. Frame-by-frame analyses were performed to measure the frontooccipital diameter (FOD) and distance from the vertex to the base of the fetal skull. Results: During vaginal delivery in an occiput anterior position, the initial FOD was 10.3 cm. When expulsion began, the fetal skull was deformed and elongated, with the FOD increasing to 10.8 cm and 11.2 cm at crowning. In contrast, the distance from the vertex to the base of the skull was reduced from 6.4 cm to 5.6 cm at expulsion. Conclusions: Fetal head molding is the change in the fetal head due to the forces of labor. The biomechanics of this process are poorly understood. Our visualization of the normal mechanism of late second-stage labor shows that MRI technology can for the first time help define the changes in the diameters of the fetal head during active labor.

Comparative Anatomy of the Ovine and Female Pelvis

Background: Pelvic organ prolapse affects half of vaginally parous women. Several animal models are used to study its pathophysiology and treatment. Sheep are interesting because they develop spontaneously prolapse with similar risk factors as women and can be used for vaginal surgery. This study describes ovine pelvis anatomy and compares it to womens pelvis to provide anatomical tools for translational researchers. Methods: MRI, pelvic dissections, and histology were used for detailed macro- and microscopic analysis of relevant anatomical structures in 6 nulliparous ewes. Results: Although sheep are quadrupeds, the gross and microscopic anatomies are similar to the female pelvis. Principal differences are the shape and its orientation, the absence of the sacrospinous ligament and the internal obturator. The levator ani (except for the puborectalis) and the coccygeus muscle are present, yet the latter is more developed - coinciding with the tail. The dimensions and morphology of the ovine vagina is comparable. The retropubic and the rectovaginal space are accessible transvaginally. There is a wide expression of estrogen receptors with low or absent immunoreactivity in the urethral epithelium, bladder, anus and internal anal sphincter. Conclusion: The ovine pelvic floor has many anatomical and ultrastructural similarities to the female pelvic floor.

Human Amniotic Fluid Stem Cells Modulate Muscle Regeneration After Cardiotoxin Injury in Mice

Amniotic fluid stem cells (AFSc) are a very heterogeneous subtype of stem cells with a broad multi potential. They could be used to treat congenital malformations or diseases. Recently, mesoangioblasts, resident pericytes of skeletal muscles, were shown to undergo muscle differentiation in vitro and in vivo. In this study we focused on the identification of an AFS subtype with pericytic characteristics and evaluate its myogenic potential. We identified monoclonal AFSc lines expressing alkaline phosphatase activity (ALP) and the canonical pericytic markers neural-glial-2 chondroitin sulphate proteglycan (NG2), platelet derived growth factor receptor α and β (PDGFR-α, -β) and α smooth muscle actin (α-SMA). These cells were able to integrate into the newly formed myotubes when co-cultured with the C2C12 cells. To test the paracrine effects of these AFSC on muscle regeneration, we assessed their affects in a transwell assay with acutely injured myotubes. AFSc were able to modulate the expression of specific growth factors involved in muscle regeneration, such as Transforming Growth Factor β (Tgfβ), Interferon γ (Ifnγ), Hepatocyte Growth Factor (Hgf) and Matrix Metalloproteinase 2 (Mmp2). When AFSc were injected in injured muscles they ameliorated muscle repair as measured by the reduction of centronucleated fibers and fibrosis. Interestingly, the transcriptional program of growth factor response in vitro is observed in large part in the in vivo xenograft experimental model, with the extension of Myostatin and Matrix Metalloproteinase 9 (Mmp9). Our data suggest that AFSc subtype with pericytic characteristics have the ability to modulate muscle regeneration in vitro and in vivo.

Technical note: automatic segmentation method of pelvic floor levator hiatus in ultrasound using a self-normalising neural network

Segmentation of the levator hiatus in ultrasound allows to extract biometrics which are of importance for pelvic floor disorder assessment. In this work, we present a fully automatic method using a convolutional neural network (CNN) to outline the levator hiatus in a 2D image extracted from a 3D ultrasound volume. In particular, our method uses a recently developed scaled exponential linear unit (SELU) as a nonlinear self-normalising activation function. SELU has important advantages such as being parameter-free and mini-batch independent. A dataset with 91 images from 35 patients all labelled by three operators, is used for training and evaluation in a leave-one-patient-out cross-validation. Results show a median Dice similarity coefficient of 0.90 with an interquartile range of 0.08, with equivalent performance to the three operators (with a Williams’ index of 1.03), and outperforming a U-Net architecture without the need for batch normalisation. We conclude that the proposed fully automatic method achieved equivalent accuracy in segmenting the pelvic floor levator hiatus compared to a previous semi-automatic approach.