Lumbar Ultrasound Image Feature Extraction and Classification with Support Vector Machine

Needle entry site localization remains a challenge for procedures that involve lumbar puncture, for example, epidural anesthesia. To solve the problem, we have developed an image classification algorithm that can automatically identify the bone/interspinous region for ultrasound images obtained from lumbar spine of pregnant patients in the transverse plane. The proposed algorithm consists of feature extraction, feature selection and machine learning procedures. A set of features, including matching values, positions and the appearance of black pixels within pre-defined windows along the midline, were extracted from the ultrasound images using template matching and midline detection methods. A support vector machine was then used to classify the bone images and interspinous images. The support vector machine model was trained with 1,040 images from 26 pregnant subjects and tested on 800 images from a separate set of 20 pregnant patients. A success rate of 95.0% on training set and 93.2% on test set was achieved with the proposed method. The trained support vector machine model was further tested on 46 off-line collected videos, and successfully identified the proper needle insertion site (interspinous region) in 45 of the cases. Therefore, the proposed method is able to process the ultrasound images of lumbar spine in an automatic manner, so as to facilitate the anesthetists' work of identifying the needle entry site.     

Three-Dimensional Ultrasound-Guided Real-Time Midline Epidural Needle Placement with Epiguide: A Prospective Feasibility Study

Current 2-D ultrasound technology is unable to perform a midline neuraxial needle insertion under real-time ultrasound guidance using a standard needle and without an assistant. The aim of the work described here was to determine the feasibility of a new technology providing such capability, starting with a study evaluating the selected puncture site. A novel 3-D ultrasound imaging technique was designed using thick-slice rendering in conjunction with a custom needle guide (3DUS + Epiguide). A clinical feasibility study evaluated the ability of 3DUS + Epiguide to identify the epidural needle puncture site for a midline insertion in the lumbar spine. We hypothesized that (i) the puncture site identified by 3DUS + Epiguide was within a 5-mm radius from the site chosen by standard palpation, and (ii) the difference between the two puncture sites was not correlated to the patient characteristics age, weight, height, body mass index and gestational age. The mean (±standard deviation) distances between puncture sites determined by 3DUS + Epiguide and palpation were 3.1 (±1.7) mm and 2.8 (±1.3) mm, for the L2–3 and L3–4 interspaces of 20 patients, respectively. Distances were comparable to intra-observer variability, indicating the potential for a thick-slice rendering of 3-D ultrasound along the Epiguide trajectory to select the puncture site of a midline neuraxial needle insertion. The long-term potential benefits of this system include increased efficiency and use of anesthesia, and a reduction in the frequency and severity of the complications from incorrect needle insertions. Epidural success in the most difficult cases (e.g., the obese) will be the focus of future work.     

Enhancement of needle visibility in ultrasound-guided percutaneous procedures

The visibility of a needle in ultrasound (US)-guided percutaneous procedures is often limited by dispersion of the needle's reflections away from the probe. A needle enhancement algorithm is developed that maximizes the received reflections by steering the US beam precisely perpendicular to the needle. The resulting image clearly depicts the needle as a bright line. The key is to automatically detect the needle in the image and maintain the appropriate beam steering angle during the procedure. The brightened needle in the steered image is then fused with the original image to produce an improved image. Implementation of the needle-enhancement algorithm was made possible by using the programmable ultrasound platform and interface library (PUPIL). PUPIL is capable of acquiring digital images, processing the images in real-time, controlling the beam angle, and displaying the fused image. The results demonstrate significantly improved needle visibility in both breast phantoms and turkey breast samples.     

Automatic Midline Identification in Transverse 2-D Ultrasound Images of the Spine

Effective epidural needle placement and injection involves accurate identification of the midline of the spine. Ultrasound, as a safe pre-procedural imaging modality, is suitable for epidural guidance because it offers adequate visibility of the vertebral anatomy. However, image interpretation remains a key challenge, especially for novices. A deep neural network is proposed to automatically classify the transverse ultrasound images of the vertebrae and identify the midline. To distinguish midline images from off-center frames, the proposed network detects the left–right symmetric anatomic landmarks. To assess the feasibility of the proposed method for midline detection, a data set of ultrasound images was collected from 20 volunteers, whose body mass indices were less than 30. The data were split into two segments, for training and test. The performance of the proposed method was further evaluated using fourfold cross validation. Moreover, it was compared against a state-of-the-art deep neural network. Compared with the gold standard provided by an expert sonographer, the proposed trained network correctly classified 88% of the transverse planes from unseen test patients. This capability supports the first step of guiding the placement of an epidural needle.     

Bedside ultrasound for difficult lumbar puncture

Lumbar puncture is a common procedure performed in the emergency department for evaluation of several life-threatening conditions, including meningitis and subarachnoid hemorrhage. We describe the use of bedside ultrasound to assist in performance of the lumbar puncture in situations where the standard "blind" technique of needle insertion using palpable spinal landmarks is likely to be difficult or to fail. Use of ultrasound to guide lumbar puncture needle placement was originally reported 30 years ago in the Russian literature. More recently, ultrasound has been used for guiding needle placement for epidural and spinal anesthesia by anesthesiologists and for diagnostic lumbar puncture on infants by radiologists.     

Ultrasound-assisted Lumbar Puncture in Pediatric Emergency Medicine

Background

Ultrasound-assisted lumbar puncture in the pediatric emergency medicine setting has not been well established, but ultrasound could serve as a valuable tool in this setting.

Objective

To assess whether ultrasound increases provider confidence in identifying an insertion point for lumbar puncture.

Methods

A feasibility study was conducted using a convenience sample of pediatric emergency patients requiring lumbar puncture. Provider confidence in selecting a needle insertion site for lumbar puncture using ultrasound assistance was compared to provider confidence using traditional landmarks alone. A simple technique using a linear probe is described.

Results

Nineteen patients were included in the study, with the primary end point the mean confidence score (based on a five-point Likert scale) in identifying a needle insertion site prior to and after using ultrasound. Using the Wilcoxon signed-rank test, the mean confidence score was 2.89 with the landmark procedure alone, and 4.79 with ultrasound assistance, yielding an average score difference of 1.90 (95% confidence interval 1.23–2.56; Wilcoxon p < 0.001, paired t-test p < 0.001). Thus, compared to the landmark procedure, the use of ultrasound was associated with a significantly higher average confidence score.

Conclusion

The use of ultrasound in the pediatric emergency setting can be a valuable adjunct with lumbar puncture.     

Prospective evaluation of point-of-care ultrasound for pre-procedure identification of landmarks versus traditional palpation for lumbar puncture

BACKGROUND: The objective of this study is to determine if point-of-care ultrasound (POCUS) pre-procedure identification of landmarks can decrease failure rate, reduce procedural time, and decrease the number of needle redirections and reinsertions when performing a lumbar puncture (LP).METHODS: This was a prospective, randomized controlled trial comparing POCUS pre-procedure identification of landmarks versus traditional palpation for LP in a cohort of patients in the emergency department and intensive care unit.RESULTS: A total of 158 patients were enrolled. No significant difference was found in time to completion, needle re-direction, or needle re-insertion when using POCUS when compared to the traditional method of palpation.CONCLUSION: Consistent with findings of previous studies, our data indicate that there was no observed benefit of using POCUS to identify pre-procedure landmarks when performing an LP.     

Robotic ultrasound-guided facet joint insertion

Purpose

Facet joint insertion is a common treatment of chronic pain in the back and spine. This procedure is often performed under fluoroscopic guidance, where the staff’s repetitive radiation exposure remains an unsolved problem. Robotic ultrasound (rUS) has the potential to reduce or even eliminate the use of radiation by using ultrasound with a robotic-guided needle insertion. This work presents first clinical data of rUS-based needle insertions extending previous work of our group.

Methods

Our system implements an automatic US acquisition protocol combined with a calibrated needle targeting system. This approach assists the physician by positioning the needle holder on a trajectory selected in a 3D US volume of the spine.

Results

By the time of submission, nine facets were treated with our approach as first data from an ongoing clinical study. The insertion success rate was shown to be comparable to current clinical practice. Furthermore, US imaging offers additional anatomical context for needle trajectory planning.

Conclusion

This work shows first clinical data for robotic ultrasound-assisted facet joint insertion as a promising solution that can easily be incorporated into the clinical workflow. Presented results show the clinical value of such a system.

     

Standard Plane Localization in Ultrasound by Radial Component Model and Selective Search

Acquisition of the standard plane is crucial for medical ultrasound diagnosis. However, this process requires substantial experience and a thorough knowledge of human anatomy. Therefore it is very challenging for novices and even time consuming for experienced examiners. We proposed a hierarchical, supervised learning framework for automatically detecting the standard plane from consecutive 2-D ultrasound images. We tested this technique by developing a system that localizes the fetal abdominal standard plane from ultrasound video by detecting three key anatomical structures: the stomach bubble, umbilical vein and spine. We first proposed a novel radial component-based model to describe the geometric constraints of these key anatomical structures. We then introduced a novel selective search method which exploits the vessel probability algorithm to produce probable locations for the spine and umbilical vein. Next, using component classifiers trained by random forests, we detected the key anatomical structures at their probable locations within the regions constrained by the radial component-based model. Finally, a second-level classifier combined the results from the component detection to identify an ultrasound image as either a “fetal abdominal standard plane” or a “non- fetal abdominal standard plane.” Experimental results on 223 fetal abdomen videos showed that the detection accuracy of our method was as high as 85.6% and significantly outperformed both the full abdomen and the separate anatomy detection methods without geometric constraints. The experimental results demonstrated that our system shows great promise for application to clinical practice.     

Skin Abscess Model for Incision and Drainage

Lumbar puncture is an important diagnostic procedure in emergency medicine. Data have been published showing improved success rate with ultrasound assistance and the ability of emergency medicine physicians to recognize sonographic lumbar spinous anatomy. However, with educational models and the push for improved patient safety, procedural skills should be practiced on phantoms rather than the "see one, do one, teach one" of the past. There are no currently available phantoms for ultrasound-assisted lumbar puncture training. We have produced a phantom that can be used to train physicians on ultrasound-assisted lumbar puncture with respect to both imaging and procedural competency. A plastic fluid-filled bladder was immersed in gelled opacified mineral oil, a safe and easily used tissue mimic that obscures direct visualization of structures. Spinous anatomy is replicated with the use of wooden struts supporting wooden disks that mimic lumbar spinous processes. The spine analog was mounted over the plastic bladder and surrounded with gelled mineral oil. The phantom produces images similar to human lumbar anatomy. The phantom allows insertion of spinal needles into the "interspinous spaces" with inability to pass the needle outside of those locations. Fluid collection and repeated punctures can be performed on the phantom. Appearance and performance of the phantom were evaluated by physicians with expertise in ultrasound-assisted lumbar puncture. The only limitation is that external appearance is not realistic. This model performs well, is made from readily available materials, and can be used to train physicians in ultrasound-assisted lumbar puncture.     

Development and Evaluation of an Augmented Reality Ultrasound Guidance System for Spinal Anesthesia: Preliminary Results

Applications of ultrasound guidance for epidural injections are hindered by poor needle and epidural space visualization. This work presents an augmented reality (AR) ultrasound guidance system that addresses challenges in both needle visualization during navigation and epidural space identification for needle positioning. In this system, (i) B-mode ultrasound and the needle are visualized in a 3-D AR environment for improved navigation, and (ii) A-mode ultrasound, obtained from a custom-made single-element transducer housed at the needle tip, is used to identify the epidural space for improved needle positioning. Performance of the system was evaluated against ultrasound-only guidance in a phantom study with novice operators and an expert anesthesiologist. The procedure success rate was higher with the AR system (100%) than ultrasound-only guidance (57%). The AR system has the potential to improve procedure outcomes in terms of success rate, time, needle path-length and usability.     

A Model for Ultrasound-Assisted Lumbar Puncture

Clinicians may find traditional lumbar puncture (LP) attempts fail due to indistinct landmarks in morbidly obese patients necessitating ultrasound localization or fluoroscopy. We believe a readily available teaching model is needed because many emergency physicians may be unfamiliar with ultrasound-assisted LP. Review of current literature shows that there are few commercially available LP models suitable for use with ultrasound. Those on the market are expensive and have limited reusability. We have succeeded in creating a low-cost reusable model for training health care professionals to perform ultrasound-assisted LP. We believe there will be many benefits to using this model including: increase in emergency department (ED) LP success rates, decrease in number of radiology consults for fluoroscopy, increase in patient satisfaction, decreased waiting time in the ED, and fewer complications due to fewer needle passes. This model effectively reproduces the sonographic appearance of the lumbar spine and overlying soft tissue and aids in teaching bedside ultrasound-assisted LP. The model has an opaque "skin" overlying a gel wax mold containing a lumbar spine. A catheter containing water is imbedded in the spine to simulate the spinal canal and cerebrospinal fluid. The skin allows for a more realistic procedure and can be removed for visual confirmation of a successful LP. In addition, successful needle placement will result in return of clear fluid. Construction of the model requires a commercially available lumbar spine and items found in craft stores with a total cost of approximately $100.     

A Methodical Quantification of Needle Visibility and Echogenicity in Ultrasound Images

During ultrasound-guided percutaneous interventions, needle localization can be a challenge. To increase needle visibility, enhancements of both the imaging methods and the needle surface properties have been investigated. However, a methodical approach to compare potential solutions is currently unavailable. The work described here involves automated image acquisition, analysis and reporting techniques to collect large amounts of data efficiently, delineate relevant factors and communicate effects. Data processing included filtering, line fitting and image intensity analysis steps. Foreground and background image samples were used to compute a contrast-to-noise ratio or a signal ratio. The approach was evaluated in a comparative study of commercially available and custom-made needles. Varied parameters included needle material, diameter and surface roughness. The shafts with kerfed patterns and the trocar and chiba tips performed best. The approach enabled an intuitive polar depiction of needle visibility in ultrasound images for a large range of insertion angles.     

Single-Camera Closed-Form Real-Time Needle Tracking for Ultrasound-Guided Needle Insertion

Many common needle intervention procedures are performed with ultrasound guidance because it is a flexible, cost-effective and widely available intra-operative imaging modality. In a needle insertion procedure with ultrasound guidance, real-time calculation and visualization of the needle trajectory can help to guide the choice of puncture site and needle angle to reach the target depicted in the ultrasound image. We found that it is feasible to calculate the needle trajectory with a single camera mounted directly on the ultrasound transducer by using the needle markings. Higher accuracy is achieved compared with other similar transducer-mounted needle trackers. We used an inexpensive, real-time and easy-to-use tracking method based on an automatic feature extraction algorithm and a closed-form method for pose estimation of the needle. The overall accuracy was 0.94 ± 0.46 mm.     

Disc herniation diagnosis in MRI using a CAD framework and a two-level classifier

Purpose

Disc herniation in the lumbar spine is a common condition, so an automated method for diagnosis could be helpful in clinical applications. A computer-aided framework for disk herniation diagnosis was developed for use in magnetic resonance imaging (MRI).

Materials and Method

A computer-aided diagnosis framework for lumbar spine with a two-level classification scheme for disc herniation diagnosis was developed using heterogeneous classifiers: a perceptron classifier, a least mean square classifier, a support vector machine classifier, and a k-Means classifier. Each classifier makes a diagnosis based on a feature set generated from regions of interest that contain vertebrae, a disc, and the spinal cord. Then, an ensemble classifier makes a final decision using score values of each classifier. We used clinical MR image data from 70 subjects in T1-weighted sagittal view and T2-weighted sagittal view for evaluation of the system.

Results

MR images of 70 subjects were processed using the proposed framework resulting in successful detection of disc herniation with 99% accuracy, achieving a speedup factor of 30 in comparison with radiologist??s diagnosis.

Conclusion

The computer-aided framework works well to diagnose herniated discs in MRI scans. We expect the framework can be adapted to effectively diagnose a variety of abnormalities in the lumbar spine.     

Morphologic Evaluation of Cervical and Lumbar Facet Joints: Intra‐Articular Facet Block Considerations

Study Design: Needle orientations for lumbar and cervical facet injection were measured in cadavers and compared with facet angles measured on magnetic resonance images (MRIs). Objectives: To establish facet orientation relative to clinical procedures of a facet joint block in the cervical and lumbar spine. Methods: Needle orientation angles were measured from 20 unembalmed human cadaveric specimens (13 cervical and 7 lumbar). Spinal needles were inserted into the midpoints of the facet joint spaces from C3 to C7 and L1 to L5. Needle trajectories were measured with an optical tracking system. For comparison, facet angles from 100 clinical MRIs of lumbar spines were also measured. Facet orientations on MRIs were measured at their intersection with the transverse plane, and angles were quantified using image analysis software. Results: Typical angles for insertion of the needle into the cervical facets were oriented closer to the coronal plane, whereas insertion angles for lumbar needles were oriented closer to the sagittal plane. Relative to the sagittal plane, the mean cervical angle was 72 degrees and the mean lumbar angle was 33 degrees. The insertion points of the cervical facets were a mean of 29 mm from the midsagittal plane compared with a mean of 22 mm for the lumbar facets. MRI‐based facet joint angles correlated poorly with actual injection angles, which were overestimated 5 to 23 degrees, depending on the lumbar level. Conclusions: Knowledge of the quantitative anatomy of the facets may help improve clinical diagnosis and treatment. These data also may aid in constructing more realistic computer simulations.     

Automatic initialization of an active shape model of the prostate

In this work is reported a new method for automatic segmentation of the boundary of the prostate, in transurethral ultrasound images. The scheme is based on a robust automatic initialization of an active shape model (ASM) of the prostate, which is subsequently fitted to the boundary of the gland. The initialization of the ASM is based on pixel classification to estimate the prostate region in an ultrasound image, followed by automatic adjustment – using a multipopulation genetic algorithm (MPGA) – of the initial pose of the ASM to the binary image produced by the classifier. The initial pose is next adjusted to the gray level ultrasound image, using the MPGA. After automatic initialization, the ASM is adjusted to the gray level ultrasound image to produce the final prostate contour. The method provides fast and robust segmentation of the prostate boundary. Validation results on 22 ultrasound images are reported with 1.74 mm of mean boundary error and an estimated processing time of 66 s per image. Our automatic initialization method can be applied with the ASMs of different organs in various imaging modalities.     

三维超声成像诊断胎儿畸形的初步研究

目的 探讨三维超声成像诊断胎儿畸形的价值。方法 对７５名具有高危妊娠因素或在常规二维超声检查中拟诊胎儿畸形的孕妇进行三维超声成像。结果 ７５名妊娠妇女中共检出胎儿畸形２４例,三维超声成像所提供的畸形信息丰富,图像与原形逼真,形态自然、完整,微细结构显示清晰。结果 三维超声成像是一种良好的产前检查方法。     

Phase Grouping-Based Needle Segmentation in 3-D Trans-rectal Ultrasound-Guided Prostate Trans-perineal Therapy

A robust and efficient needle segmentation method used to localize and track the needle in 3-D trans-rectal ultrasound (TRUS)-guided prostate therapy is proposed. The algorithmic procedure begins by cropping the 3-D US image containing a needle; then all voxels in the cropped 3-D image are grouped into different line support regions (LSRs) based on the outer product of the adjacent voxels' gradient vector. Two different needle axis extraction methods in the candidate LSR are presented: least-squares fitting and 3-D randomized Hough transform. Subsequent local optimization refines the position of the needle axis. Finally, the needle endpoint is localized by finding an intensity drop along the needle axis. The proposed methods were validated with 3-D TRUS tissue-mimicking agar phantom images, chicken breast phantom images and patient images obtained during prostate cryotherapy. The results of the in vivo test indicate that our method can localize the needle accurately and robustly with a needle endpoint localization accuracy <1.43 mm and detection accuracy >84%, which are favorable for 3-D TRUS-guided prostate trans-perineal therapy.     

Acoustic Radiation Force Impulse (ARFI) imaging-based needle visualization

Ultrasound-guided needle placement is widely used in the clinical setting, particularly for central venous catheter placement, tissue biopsy and regional anesthesia. Difficulties with ultrasound guidance in these areas often result from steep needle insertion angles and spatial offsets between the imaging plane and the needle. Acoustic Radiation Force Impulse (ARFI) imaging leads to improved needle visualization because it uses a standard diagnostic scanner to perform radiation force based elasticity imaging, creating a displacement map that displays tissue stiffness variations. The needle visualization in ARFI images is independent of needle-insertion angle and also extends needle visibility out of plane. Although ARFI images portray needles well, they often do not contain the usual B-mode landmarks. Therefore, a three-step segmentation algorithm has been developed to identify a needle in an ARFI image and overlay the needle prediction on a coregistered B-mode image. The steps are: (1) contrast enhancement by median filtration and Laplacian operator filtration, (2) noise suppression through displacement estimate correlation coefficient thresholding and (3) smoothing by removal of outliers and best-fit line prediction. The algorithm was applied to data sets from horizontal 18, 21 and 25 gauge needles between 0-4 mm offset in elevation from the transducer imaging plane and to 18G needles on the transducer axis (in plane) between 10 degrees and 35 degrees from the horizontal. Needle tips were visualized within 2 mm of their actual position for both horizontal needle orientations up to 1.5 mm offset in elevation from the transducer imaging plane and on-axis angled needles between 10 degrees-35 degrees above the horizontal orientation. We conclude that segmented ARFI images overlaid on matched B-mode images hold promise for improved needle visibility in many clinical applications.