Automated Shoulder Implant Manufacturer Detection using Encoder Decoder based Classifier from X-ray Images.

Total shoulder arthroplasty is the process of replacing the damaged ball and socket joint in the shoulder with a prosthesis made with polyethylene and metal components. The prosthesis helps to restore the normal range of motion and reduce pain, enabling the patient to return to their daily activities. These implants may need to be replaced over the years due to damage or wear and tear. It is a tedious and time-consuming process to identify the type of implant if medical records are not properly maintained. Artificial intelligence systems can speed up the treatment process by classifying the manufacturer and model of the prosthesis. We have proposed an encoder-decoder based classifier along with the supervised contrastive loss function that can identify the implant manufacturer effectively with increased accuracy of 92% from X-ray images overcoming the class imbalance problem.

Cardiac Anomaly Detection from Cine MRI Images using Physiological Features and Random Forest Classifier.

Computer-aided diagnosis (CAD) with cine MRI is a foremost research topic to enable improved, faster, and more accurate diagnosis of cardiovascular diseases (CVD). However, current approaches that use manual visualization or conventional clinical indices can lack accuracy for borderline cases. Also, manual visualization of 3D/4D MR data is time-consuming and expert-dependent. We try to simplify this process by creating an end-to-end automated CAD system that segments the critical substructures of the heart. The new domain-related physiological features are then calculated from the segmented regions. These features are fed to a random forest classifier that identifies the anomaly. We have obtained a very high accuracy when testing this end-to-end approach on the Automated Cardiac Diagnosis challenge (ACDC) dataset (4 pathologies, 1 normal). To prove the generalizability of the method we have blind-tested this approach on M&Ms-2 dataset which is a multi-center, multi-vendor, and multi-disease dataset with better than 90% accuracy.

3D Cardiac Substructures Segmentation from CMRI using Generative Adversarial Network (GAN).

Cardiac magnetic resonance imaging (CMRI) improves the diagnosis of cardiovascular diseases by providing images at high spatio-temporal resolution helping physicians in providing correct treatment plans. Segmentation and identification of various substructures of the heart at different cardiac phases of end-systole and end-diastole helps in the extraction of ventricular function information such as stroke volume, ejection fraction, myocardium thickness, etc. Manual delineation of the substructures is tedious, time-consuming, and error-prone. We have implemented a 3D GAN that includes 3D contextual information capable of segmenting and identifying the substructures at different cardiac phases with improved accuracy. Our method is evaluated on the ACDC dataset (4 pathologies, 1 healthy group) to show that the proposed out-performs other methods in literature with less amount of data. Also, the proposed provided a better Dice score in segmentation surpassing other methods on a blind-tested M&Ms dataset.

Anatomical Landmark Detection using Deep Appearance-Context Network.

Accurate identification of anatomical landmarks is a crucial step in medical image analysis. While deep neural networks have shown impressive performance on computer vision tasks, they rely on a large amount of data, which is often not available. In this work, we propose an attention-driven end-to-end deep learning architecture, which learns the local appearance and global context separately that helps in stable training under limited data. The experiments conducted demonstrate the effectiveness of the proposed approach with impressive results in localizing landmarks when evaluated on cephalometric and spine X-ray image data. The predicted landmarks are further utilized in biomedical applications to demonstrate the impact.

Precise Bleeding and Red lesions localization from Capsule Endoscopy using Compact U-Net.

Wireless capsule endoscopy is a non-invasive and painless procedure to detect anomalies from the gastrointestinal tract. Single examination results in up to 8 hrs of video and requires between 45 - 180 mins for diagnosis depending on the complexity. Image and video computational methods are needed to increase both efficiency and accuracy of the diagnosis. In this paper, a compact U-Net with lesser encoder-decoder pairs is presented, to detect and precisely segment bleeding and red lesions from endoscopy data. The proposed compact U-Net is compared with the original U-Net and also with other methods reported in the literature. The results show the proposed compact network performs on par with the original network but with faster training and lesser memory consumption. Also, the proposed model provided a dice score of 91% outperforming other methods reported on a blind tested WCE dataset with no images from this set used for training.

Heart Region Segmentation using Dense VNet from Multimodality Images.

Cardiovascular diseases (CVD) have been identified as one of the most common causes of death in the world. Advanced development of imaging techniques is allowing timely detection of CVD and helping physicians in providing correct treatment plans in saving lives. Segmentation and Identification of various substructures of the heart are very important in modeling a digital twin of the patient-specific heart. Manual delineation of various substructures of the heart is tedious and time-consuming. Here we have implemented Dense VNet for detecting substructures of the heart from both CT and MRI multimodality data. Due to the limited availability of data we have implemented an on-the-fly elastic deformation data augmentation technique. The result of the proposed has been shown to outperform other methods reported in the literature on both CT and MRI datasets.

Surgical tool segmentation and localization using spatio-temporal deep network.

Laparoscopic cholecystectomy surgery is a minimally invasive surgery to remove the gallbladder, where surgical instruments are inserted through small incisions in the abdomen with the help of a laparoscope. Identification of tool presence and precise segmentation of tools from the video is very important in understanding the quality of the surgery and training budding surgeons. Precise segmentation of tools is required to track the tools during real-time surgeries. In this paper, a new pixel-wise instance segmentation algorithm is proposed, which segments and localizes the surgical tool using spatio-temporal deep network. The performance of the proposed has been compared with the state-of-the-art image-based instance segmentation method using the Cholec80 dataset. It is also compared with methods in the literature using frame-level presence detection and spatial detection with good results.

Cloud solution for histopathological image analysis using region of interest based compression.

Recent technological gains have led to the adoption of innovative cloud based solutions in medical imaging field. Once the medical image is acquired, it can be viewed, modified, annotated and shared on many devices. This advancement is mainly due to the introduction of Cloud computing in medical domain. Tissue pathology images are complex and are normally collected at different focal lengths using a microscope. The single whole slide image contains many multi resolution images stored in a pyramidal structure with the highest resolution image at the base and the smallest thumbnail image at the top of the pyramid. Highest resolution image will be used for tissue pathology diagnosis and analysis. Transferring and storing such huge images is a big challenge. Compression is a very useful and effective technique to reduce the size of these images. As pathology images are used for diagnosis, no information can be lost during compression (lossless compression). A novel method of extracting the tissue region and applying lossless compression on this region and lossy compression on the empty regions has been proposed in this paper. The resulting compression ratio along with lossless compression on tissue region is in acceptable range allowing efficient storage and transmission to and from the Cloud.

A hybrid approach for nucleus stain separation in histopathological images.

Difficulties in automation of histology image analysis are caused due to varying stain colors in the histology slides and the interaction of stains. Incorrect stain separation results in incorrect nucleus segmentation. A new hybrid algorithm has been proposed combining de-staining and wedge separation algorithms, which provides better stain separation and maintains color integrity of the input image. The proposed algorithm is tested on 36 histopathological images covering varying tissues and compared with popular methods in the area with excellent results in high nuclei density category.

Automatic tumour volume delineation in respiratory-gated PET images.

INTRODUCTION: Positron emission tomography (PET) is a state-of-the-art functional imaging technique used in the accurate detection of cancer. The main problem with the tumours present in the lungs is that they are non-stationary during each respiratory cycle. Tumours in the lungs can get displaced up to 2.5 cm during respiration. Accurate detection of the tumour enables avoiding the addition of extra margin around the tumour that is usually used during radiotherapy treatment planning. METHODS: This paper presents a novel method to detect and track tumour in respiratory-gated PET images. The approach followed to achieve this task is to automatically delineate the tumour from the first frame using support vector machines. The resulting volume and position information from the first frame is used in tracking its motion in the subsequent frames with the help of level set (LS) deformable model. RESULTS: An excellent accuracy of 97% is obtained using wavelets and support vector machines. The volume calculated as a result of the machine learning (ML) stage is used as a constraint for deformable models and the tumour is tracked in the remaining seven phases of the respiratory cycle. As a result, the complete information about tumour movement during each respiratory cycle is available in relatively short time. CONCLUSIONS: The combination of the LS and ML approach accurately delineated the tumour volume from all frames, thereby providing a scope of using PET images towards planning an accurate and effective radiotherapy treatment for lung cancer.

Pulmonary tumor volume delineation in PET images using deformable models.

Lung cancer is one of the most lethal form of cancer worldwide. The tumor present in the lungs is not static and changes its shape and position during each breathing cycle. In order to segment the tumor, the physicians manually outline the tumor on each slice. Slice by slice manual segmentation is prone to errors and causes physician fatigue. A semi-automatic method to segment and track the tumor in all the frames of PET data is proposed in this paper. The tumor is segmented from each slice of the first frame using wavelet features and support vector machine classifier. This segmented tumor, after validated by the experts is used in initialization of the contour for segmentation of the tumor in subsequent frames by the level set method. Another important contribution of this paper is setting up tumor volume obtained from the first frame as the termination condition for the level set method. The results obtained from the proposed methodology are very promising and eliminates the need for manual tumor segmentation. Our proposed technique also maintains consistent segmentation and the results obtained are not dependent on the operator as is the case in manual segmentation.