Jinah Park

Matthias Eisenmann, Annika Reinke, Vivienn Weru et al.

International benchmarking competitions have become fundamental for the comparative performance assessment of image analysis methods. However, little attention has been given to investigating what can be learnt from these competitions. Do they really generate scientific progress? What are common and successful participation strategies? What makes a solution superior to a competing method? To address this gap in the literature, we performed a multi-center study with all 80 competitions that were conducted in the scope of IEEE ISBI 2021 and MICCAI 2021. Statistical analyses performed based on comprehensive descriptions of the submitted algorithms linked to their rank as well as the underlying participation strategies revealed common characteristics of winning solutions. These typically include the use of multi-task learning (63%) and/or multi-stage pipelines (61%), and a focus on augmentation (100%), image preprocessing (97%), data curation (79%), and postprocessing (66%). The "typical" lead of a winning team is a computer scientist with a doctoral degree, five years of experience in biomedical image analysis, and four years of experience in deep learning. Two core general development strategies stood out for highly-ranked teams: the reflection of the metrics in the method design and the focus on analyzing and handling failure cases. According to the organizers, 43% of the winning algorithms exceeded the state of the art but only 11% completely solved the respective domain problem. The insights of our study could help researchers (1) improve algorithm development strategies when approaching new problems, and (2) focus on open research questions revealed by this work.

Marc Aubreville, Nikolas Stathonikos, Christof A. Bertram et al.

The density of mitotic figures within tumor tissue is known to be highly correlated with tumor proliferation and thus is an important marker in tumor grading. Recognition of mitotic figures by pathologists is known to be subject to a strong inter-rater bias, which limits the prognostic value. State-of-the-art deep learning methods can support the expert in this assessment but are known to strongly deteriorate when applied in a different clinical environment than was used for training. One decisive component in the underlying domain shift has been identified as the variability caused by using different whole slide scanners. The goal of the MICCAI MIDOG 2021 challenge has been to propose and evaluate methods that counter this domain shift and derive scanner-agnostic mitosis detection algorithms. The challenge used a training set of 200 cases, split across four scanning systems. As a test set, an additional 100 cases split across four scanning systems, including two previously unseen scanners, were given. The best approaches performed on an expert level, with the winning algorithm yielding an F_1 score of 0.748 (CI95: 0.704-0.781). In this paper, we evaluate and compare the approaches that were submitted to the challenge and identify methodological factors contributing to better performance.

Soonchan Park, Jinah Park

Studies of virtual try-on (VITON) have been shown their effectiveness in utilizing the generative neural network for virtually exploring fashion products, and some of recent researches of VITON attempted to synthesize human image wearing given multiple types of garments (e.g., top and bottom clothes). However, when replacing the top and bottom clothes of the target human, numerous wearing styles are possible with a certain combination of the clothes. In this paper, we address the problem of variation in wearing style when simultaneously replacing the top and bottom clothes of the model. We introduce Wearing-Guide VITON (i.e., WG-VITON) which utilizes an additional input binary mask to control the wearing styles of the generated image. Our experiments show that WG-VITON effectively generates an image of the model wearing given top and bottom clothes, and create complicated wearing styles such as partly tucking in the top to the bottom

Jihoon Cho, Jinah Park

Medical segmentation has grown exponentially through the advent of a fully convolutional network (FCN), and we have now reached a turning point through the success of Transformer. However, the different characteristics of the modality have not been fully integrated into Transformer for medical segmentation. In this work, we propose the novel hybrid fusion Transformer (HFTrans) for multisequence MRI image segmentation. We take advantage of the differences among multimodal MRI sequences and utilize the Transformer layers to integrate the features extracted from each modality as well as the features of the early fused modalities. We validate the effectiveness of our hybrid-fusion method in three-dimensional (3D) medical segmentation. Experiments on two public datasets, BraTS2020 and MRBrainS18, show that the proposed method outperforms previous state-of-the-art methods on the task of brain tumor segmentation and brain structure segmentation.

Inha Kang, Jinah Park

Obtaining ground truth data in medical imaging has difficulties due to the fact that it requires a lot of annotating time from the experts in the field. Also, when trained with supervised learning, it detects only the cases included in the labels. In real practice, we want to also open to other possibilities than the named cases while examining the medical images. As a solution, the need for anomaly detection that can detect and localize abnormalities by learning the normal characteristics using only normal images is emerging. With medical image data, we can design either 2D or 3D networks of self-supervised learning for anomaly detection task. Although 3D networks, which learns 3D structures of the human body, show good performance in 3D medical image anomaly detection, they cannot be stacked in deeper layers due to memory problems. While 2D networks have advantage in feature detection, they lack 3D context information. In this paper, we develop a method for combining the strength of the 3D network and the strength of the 2D network through joint embedding. We also propose the pretask of self-supervised learning to make it possible for the networks to learn efficiently. Through the experiments, we show that the proposed method achieves better performance in both classification and segmentation tasks compared to the SoTA method.

Suhyun Ahn, Wonjung Park, Jihoon Cho et al.

Spatial control methods using additional modules on pretrained diffusion models have gained attention for enabling conditional generation in natural images. These methods guide the generation process with new conditions while leveraging the capabilities of large models. They could be beneficial as training strategies in the context of 3D medical imaging, where training a diffusion model from scratch is challenging due to high computational costs and data scarcity. However, the potential application of spatial control methods with additional modules to 3D medical images has not yet been explored. In this paper, we present a tailored spatial control method for 3D medical images with a novel lightweight module, Volumetric Conditioning Module (VCM). Our VCM employs an asymmetric U-Net architecture to effectively encode complex information from various levels of 3D conditions, providing detailed guidance in image synthesis. To examine the applicability of spatial control methods and the effectiveness of VCM for 3D medical data, we conduct experiments under single- and multimodal conditions scenarios across a wide range of dataset sizes, from extremely small datasets with 10 samples to large datasets with 500 samples. The experimental results show that the VCM is effective for conditional generation and efficient in terms of requiring less training data and computational resources. We further investigate the potential applications for our spatial control method through axial super-resolution for medical images. Our code is available at \url{https://github.com/Ahn-Ssu/VCM}

Taeyun Woo, Tae-Kyun Kim, Jinah Park

Estimating the poses of both a hand and an object has become an important area of research due to the growing need for advanced vision computing. The primary challenge involves understanding and reconstructing how hands and objects interact, such as contact and physical plausibility. Existing approaches often adopt a graph neural network to incorporate spatial information of hand and object meshes. However, these approaches have not fully exploited the potential of graphs without modification of edges within and between hand- and object-graphs. We propose a graph-based refinement method that incorporates an interaction-aware graph-attention mechanism to account for hand-object interactions. Using edges, we establish connections among closely correlated nodes, both within individual graphs and across different graphs. Experiments demonstrate the effectiveness of our proposed method with notable improvements in the realm of physical plausibility.

Wonjung Park, Suhyun Ahn, Jinah Park

The deformation of the lateral ventricle (LV) shape is widely studied to identify specific morphometric changes associated with diseases. Since LV enlargement is considered a relative change due to brain atrophy, local longitudinal LV deformation can indicate deformation in adjacent brain areas. However, conventional methods for LV shape analysis focus on modeling the solely segmented LV mask. In this work, we propose a novel deep learning-based approach using peripheral area projection, which is the first attempt to analyze LV considering its surrounding areas. Our approach matches the baseline LV mesh by deforming the shape of follow-up LVs, while optimizing the corresponding points of the same adjacent brain area between the baseline and follow-up LVs. Furthermore, we quantitatively evaluated the deformation of the left LV in normal (n=10) and demented subjects (n=10), and we found that each surrounding area (thalamus, caudate, hippocampus, amygdala, and right LV) projected onto the surface of LV shows noticeable differences between normal and demented subjects.

Jihoon Cho, Xiaofeng Liu, Fangxu Xing et al.

Multimodal Magnetic Resonance (MR) Imaging plays a crucial role in disease diagnosis due to its ability to provide complementary information by analyzing a relationship between multimodal images on the same subject. Acquiring all MR modalities, however, can be expensive, and, during a scanning session, certain MR images may be missed depending on the study protocol. The typical solution would be to synthesize the missing modalities from the acquired images such as using generative adversarial networks (GANs). Yet, GANs constructed with convolutional neural networks (CNNs) are likely to suffer from a lack of global relationships and mechanisms to condition the desired modality. To address this, in this work, we propose a transformer-based modality infuser designed to synthesize multimodal brain MR images. In our method, we extract modality-agnostic features from the encoder and then transform them into modality-specific features using the modality infuser. Furthermore, the modality infuser captures long-range relationships among all brain structures, leading to the generation of more realistic images. We carried out experiments on the BraTS 2018 dataset, translating between four MR modalities, and our experimental results demonstrate the superiority of our proposed method in terms of synthesis quality. In addition, we conducted experiments on a brain tumor segmentation task and different conditioning methods.

Jihoon Cho, Seunghyuck Park, Jinah Park

Despite significant advancements in automatic brain tumor segmentation methods, their performance is not guaranteed when certain MR sequences are missing. Addressing this issue, it is crucial to synthesize the missing MR images that reflect the unique characteristics of the absent modality with precise tumor representation. Typically, MRI synthesis methods generate partial images rather than full-sized volumes due to computational constraints. This limitation can lead to a lack of comprehensive 3D volumetric information and result in image artifacts during the merging process. In this paper, we propose a two-stage approach that first synthesizes MR images from 2D slices using a novel intensity encoding method and then refines the synthesized MRI. The proposed intensity encoding reduces artifacts when synthesizing MRI on a 2D slice basis. Then, the \textit{Refiner}, which leverages complete 3D volume information, further improves the quality of the synthesized images and enhances their applicability to segmentation methods. Experimental results demonstrate that the intensity encoding effectively minimizes artifacts in the synthesized MRI and improves perceptual quality. Furthermore, using the \textit{Refiner} on synthesized MRI significantly improves brain tumor segmentation results, highlighting the potential of our approach in practical applications.

Taeyun Woo, Jinah Park, Tae-Kyun Kim

Deterministic models for 3D hand pose reconstruction, whether single-staged or cascaded, struggle with pose ambiguities caused by self-occlusions and complex hand articulations. Existing cascaded approaches refine predictions in a coarse-to-fine manner but remain deterministic and cannot capture pose uncertainties. Recent probabilistic methods model pose distributions yet are restricted to single-stage estimation, which often fails to produce accurate 3D reconstructions without refinement. To address these limitations, we propose a coarse-to-fine cascaded diffusion framework that combines probabilistic modeling with cascaded refinement. The first stage is a joint diffusion model that samples diverse 3D joint hypotheses, and the second stage is a Mesh Latent Diffusion Model (Mesh LDM) that reconstructs a 3D hand mesh conditioned on a joint sample. By training Mesh LDM with diverse joint hypotheses in a learned latent space, our framework learns distribution-aware joint-mesh relationships and robust hand priors. Furthermore, the cascaded design mitigates the difficulty of directly mapping 2D images to dense 3D poses, enhancing accuracy through sequential refinement. Experiments on FreiHAND and HO3Dv2 demonstrate that our method achieves state-of-the-art performance while effectively modeling pose distributions.

Jihoon Cho, Jonghye Woo, Jinah Park

Multisequence Magnetic Resonance Imaging (MRI) provides a more reliable diagnosis in clinical applications through complementary information across sequences. However, in practice, the absence of certain MR sequences is a common problem that can lead to inconsistent analysis results. In this work, we propose a novel unified framework for synthesizing multisequence MR images, called hybrid-fusion GAN (HF-GAN). The fundamental mechanism of this work is principled feature disentanglement, which aligns the design of the architecture with the complexity of the features. A powerful many-to-one stream is constructed for the extraction of complex complementary features, while utilizing parallel, one-to-one streams to process modality-specific information. These disentangled features are dynamically integrated into a common latent space by a channel attention-based fusion module (CAFF) and then transformed via a modality infuser to generate the target sequence. We validated our framework on public datasets of both healthy and pathological brain MRI. Quantitative and qualitative results show that HF-GAN achieves state-of-the-art performance, with our 2D slice-based framework notably outperforming a leading 3D volumetric model. Furthermore, the utilization of HF-GAN for data imputation substantially improves the performance of the downstream brain tumor segmentation task, demonstrating its clinical relevance.

Youjin Chung, Jihoon Cho, Jinah Park

We propose a new training scheme for domain generalization in mitotic figure detection. Mitotic figures show different characteristics for each scanner. We consider each scanner as a 'domain' and the image distribution specified for each domain as 'style'. The goal is to train our network to be robust on scanner types by using various 'style' images. To expand the style variance, we transfer a style of the training image into arbitrary styles, by defining a module based on StarGAN. Our model with the proposed training scheme shows positive performance on MIDOG Preliminary Test-Set containing scanners never seen before.