DistilIQA: Distilling Vision Transformers for no-reference perceptual CT image quality assessment.

No-reference image quality assessment (IQA) is a critical step in medical image analysis, with the objective of predicting perceptual image quality without the need for a pristine reference image. The application of no-reference IQA to CT scans is valuable in providing an automated and objective approach to assessing scan quality, optimizing radiation dose, and improving overall healthcare efficiency. In this paper, we introduce DistilIQA, a novel distilled Vision Transformer network designed for no-reference CT image quality assessment. DistilIQA integrates convolutional operations and multi-head self-attention mechanisms by incorporating a powerful convolutional stem at the beginning of the traditional ViT network. Additionally, we present a two-step distillation methodology aimed at improving network performance and efficiency. In the initial step, a "teacher ensemble network" is constructed by training five vision Transformer networks using a five-fold division schema. In the second step, a "student network", comprising of a single Vision Transformer, is trained using the original labeled dataset and the predictions generated by the teacher network as new labels. DistilIQA is evaluated in the task of quality score prediction from low-dose chest CT scans obtained from the LDCT and Projection data of the Cancer Imaging Archive, along with low-dose abdominal CT images from the LDCTIQAC2023 Grand Challenge. Our results demonstrate DistilIQA's remarkable performance in both benchmarks, surpassing the capabilities of various CNNs and Transformer architectures. Moreover, our comprehensive experimental analysis demonstrates the effectiveness of incorporating convolutional operations within the ViT architecture and highlights the advantages of our distillation methodology.

CrossMoDA 2021 challenge: Benchmark of cross-modality domain adaptation techniques for vestibular schwannoma and cochlea segmentation.

Domain Adaptation (DA) has recently been of strong interest in the medical imaging community. While a large variety of DA techniques have been proposed for image segmentation, most of these techniques have been validated either on private datasets or on small publicly available datasets. Moreover, these datasets mostly addressed single-class problems. To tackle these limitations, the Cross-Modality Domain Adaptation (crossMoDA) challenge was organised in conjunction with the 24th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI 2021). CrossMoDA is the first large and multi-class benchmark for unsupervised cross-modality Domain Adaptation. The goal of the challenge is to segment two key brain structures involved in the follow-up and treatment planning of vestibular schwannoma (VS): the VS and the cochleas. Currently, the diagnosis and surveillance in patients with VS are commonly performed using contrast-enhanced T1 (ceT1) MR imaging. However, there is growing interest in using non-contrast imaging sequences such as high-resolution T2 (hrT2) imaging. For this reason, we established an unsupervised cross-modality segmentation benchmark. The training dataset provides annotated ceT1 scans (N=105) and unpaired non-annotated hrT2 scans (N=105). The aim was to automatically perform unilateral VS and bilateral cochlea segmentation on hrT2 scans as provided in the testing set (N=137). This problem is particularly challenging given the large intensity distribution gap across the modalities and the small volume of the structures. A total of 55 teams from 16 countries submitted predictions to the validation leaderboard. Among them, 16 teams from 9 different countries submitted their algorithm for the evaluation phase. The level of performance reached by the top-performing teams is strikingly high (best median Dice score - VS: 88.4%; Cochleas: 85.7%) and close to full supervision (median Dice score - VS: 92.5%; Cochleas: 87.7%). All top-performing methods made use of an image-to-image translation approach to transform the source-domain images into pseudo-target-domain images. A segmentation network was then trained using these generated images and the manual annotations provided for the source image.

EMONAS-Net: Efficient multiobjective neural architecture search using surrogate-assisted evolutionary algorithm for 3D medical image segmentation.

Deep learning plays a critical role in medical image segmentation. Nevertheless, manually designing a neural network for a specific segmentation problem is a very difficult and time-consuming task due to the massive hyperparameter search space, long training time and large volumetric data. Therefore, most designed networks are highly complex, task specific and over-parametrized. Recently, multiobjective neural architecture search (NAS) methods have been proposed to automate the design of accurate and efficient segmentation architectures. However, they only search for either the micro- or macro-structure of the architecture, do not use the information produced during the optimization process to increase the efficiency of the search, or do not consider the volumetric nature of medical images. In this work, we present EMONAS-Net, an Efficient MultiObjective NAS framework for 3D medical image segmentation that optimizes both the segmentation accuracy and size of the network. EMONAS-Net has two key components, a novel search space that considers the configuration of the micro- and macro-structure of the architecture and a Surrogate-assisted Multiobjective Evolutionary based Algorithm (SaMEA algorithm) that efficiently searches for the best hyperparameter values. The SaMEA algorithm uses the information collected during the initial generations of the evolutionary process to identify the most promising subproblems and select the best performing hyperparameter values during mutation to improve the convergence speed. Furthermore, a Random Forest surrogate model is incorporated to accelerate the fitness evaluation of the candidate architectures. EMONAS-Net is tested on the tasks of prostate segmentation from the MICCAI PROMISE12 challenge, hippocampus segmentation from the Medical Segmentation Decathlon challenge, and cardiac segmentation from the MICCAI ACDC challenge. In all the benchmarks, the proposed framework finds architectures that perform better or comparable with competing state-of-the-art NAS methods while being considerably smaller and reducing the architecture search time by more than 50%.

AdaEn-Net: An ensemble of adaptive 2D-3D Fully Convolutional Networks for medical image segmentation.

Fully Convolutional Networks (FCNs) have emerged as powerful segmentation models but are usually designed manually, which requires extensive time and can result in large and complex architectures. There is a growing interest to automatically design efficient architectures that can accurately segment 3D medical images. However, most approaches either do not fully exploit volumetric information or do not optimize the model's size. To address these problems, we propose a self-adaptive 2D-3D ensemble of FCNs called AdaEn-Net for 3D medical image segmentation that incorporates volumetric data and adapts to a particular dataset by optimizing both the model's performance and size. The AdaEn-Net consists of a 2D FCN that extracts intra-slice information and a 3D FCN that exploits inter-slice information. The architecture and hyperparameters of the 2D and 3D architectures are found through a multiobjective evolutionary based algorithm that maximizes the expected segmentation accuracy and minimizes the number of parameters in the network. The main contribution of this work is a model that fully exploits volumetric information and automatically searches for a high-performing and efficient architecture. The AdaEn-Net was evaluated for prostate segmentation on the PROMISE12 Grand Challenge and for cardiac segmentation on the MICCAI ACDC challenge. In the first challenge, the AdaEn-Net ranks 9 out of 297 submissions and surpasses the performance of an automatically-generated segmentation network while producing an architecture with 13× fewer parameters. In the second challenge, the proposed model is ranked within the top 8 submissions and outperforms an architecture designed with reinforcement learning while having 1.25× fewer parameters.

A web-based knowledge management system integrating Western and Traditional Chinese Medicine for relational medical diagnosis.

This article presents the design of a web-based knowledge management system as a training and research tool for the exploration of key relationships between Western and Traditional Chinese Medicine, in order to facilitate relational medical diagnosis integrating these mainstream healing modalities. The main goal of this system is to facilitate decision-making processes, while developing skills and creating new medical knowledge. Traditional Chinese Medicine can be considered as an ancient relational knowledge-based approach, focusing on balancing interrelated human functions to reach a healthy state. Western Medicine focuses on specialties and body systems and has achieved advanced methods to evaluate the impact of a health disorder on the body functions. Identifying key relationships between Traditional Chinese and Western Medicine opens new approaches for health care practices and can increase the understanding of human medical conditions. Our knowledge management system was designed from initial datasets of symptoms, known diagnosis and treatments, collected from both medicines. The datasets were subjected to process-oriented analysis, hierarchical knowledge representation and relational database interconnection. Web technology was implemented to develop a user-friendly interface, for easy navigation, training and research. Our system was prototyped with a case study on chronic prostatitis. This trial presented the system's capability for users to learn the correlation approach, connecting knowledge in Western and Traditional Chinese Medicine by querying the database, mapping validated medical information, accessing complementary information from official sites, and creating new knowledge as part of the learning process. By addressing the challenging tasks of data acquisition and modeling, organization, storage and transfer, the proposed web-based knowledge management system is presented as a tool for users in medical training and research to explore, learn and update relational information for the practice of integrated medical diagnosis. This proposal in education has the potential to enable further creation of medical knowledge from both Traditional Chinese and Western Medicine for improved care providing. The presented system positively improves the information visualization, learning process and knowledge sharing, for training and development of new skills for diagnosis and treatment, and a better understanding of medical diseases.

MRI-based segmentation of pubic bone for evaluation of pelvic organ prolapse.

Pelvic organ prolapse (POP) is a major women's health problem. Its diagnosis through magnetic resonance imaging (MRI) has become popular due to current inaccuracies of clinical examination. The diagnosis of POP on MRI consists of identifying reference points on pelvic bone structures for measurement and evaluation. However, it is currently performed manually, making it a time-consuming and subjective procedure. We present a new segmentation approach for automating pelvic bone point identification on MRI. It consists of a multistage mechanism based on texture-based block classification, leak detection, and prior shape information. Texture-based block classification and clustering analysis using K-means algorithm are integrated to generate the initial bone segmentation and to identify leak areas. Prior shape information is incorporated to obtain the final bone segmentation. Then, the reference points are identified using morphological skeleton operation. Results demonstrate that the proposed method achieves higher bone segmentation accuracy compared to other segmentation methods. The proposed method can also automatically identify reference points faster and with more consistency compared with the manually identified point process by experts. This research aims to enable faster and consistent pelvic measurements on MRI to facilitate and improve the diagnosis of female POP.