Multi-task learning for predicting quality-of-life and independence in activities of daily living after stroke: a proof-of-concept study.

A health-related (HR) profile is a set of multiple health-related items recording the status of the patient at different follow-up times post-stroke. In order to support clinicians in designing rehabilitation treatment programs, we propose a novel multi-task learning (MTL) strategy for predicting post-stroke patient HR profiles. The HR profile in this study is measured by the Barthel index (BI) assessment or by the EQ-5D-3L questionnaire. Three datasets are used in this work and for each dataset six neural network architectures are developed and tested. Results indicate that an MTL architecture combining a pre-trained network for all tasks with a concatenation strategy conditioned by a task grouping method is a promising approach for predicting the HR profile of a patient with stroke at different phases of the patient journey. These models obtained a mean F1-score of 0.434 (standard deviation 0.022, confidence interval at 95% [0.428, 0.44]) calculated across all the items when predicting BI at 3 months after stroke (MaS), 0.388 (standard deviation 0.029, confidence interval at 95% [0.38, 0.397]) when predicting EQ-5D-3L at 6MaS, and 0.462 (standard deviation 0.029, confidence interval at 95% [0.454, 0.47]) when predicting the EQ-5D-3L at 18MaS. Furthermore, our MTL architecture outperforms the reference single-task learning models and the classic MTL of all tasks in 8 out of 10 tasks when predicting BI at 3MaS and has better prediction performance than the reference models on all tasks when predicting EQ-5D-3L at 6 and 18MaS. The models we present in this paper are the first models to predict the components of the BI or the EQ-5D-3L, and our results demonstrate the potential benefits of using MTL in a health context to predict patient profiles.

Automated segment-level coronary artery calcium scoring on non-contrast CT: a multi-task deep-learning approach.

OBJECTIVES: To develop and evaluate a multi-task deep-learning (DL) model for automated segment-level coronary artery calcium (CAC) scoring on non-contrast computed tomography (CT) for precise localization and quantification of calcifications in the coronary artery tree. METHODS: This study included 1514 patients (mean age, 60.0 ± 10.2 years; 56.0% female) with stable chest pain from 26 centers participating in the multicenter DISCHARGE trial (NCT02400229). The patients were randomly assigned to a training/validation set (1059) and a test set (455). We developed a multi-task neural network for performing the segmentation of calcifications on the segment level as the main task and the segmentation of coronary artery segment regions with weak annotations as an auxiliary task. Model performance was evaluated using (micro-average) sensitivity, specificity, F1-score, and weighted Cohen's κ for segment-level agreement based on the Agatston score and performing interobserver variability analysis. RESULTS: In the test set of 455 patients with 1797 calcifications, the model assigned 73.2% (1316/1797) to the correct coronary artery segment. The model achieved a micro-average sensitivity of 0.732 (95% CI: 0.710-0.754), a micro-average specificity of 0.978 (95% CI: 0.976-0.980), and a micro-average F1-score of 0.717 (95% CI: 0.695-0.739). The segment-level agreement was good with a weighted Cohen's κ of 0.808 (95% CI: 0.790-0.824), which was only slightly lower than the agreement between the first and second observer (0.809 (95% CI: 0.798-0.845)). CONCLUSION: Automated segment-level CAC scoring using a multi-task neural network approach showed good agreement on the segment level, indicating that DL has the potential for automated coronary artery calcification classification. CRITICAL RELEVANCE STATEMENT: Multi-task deep learning can perform automated coronary calcium scoring on the segment level with good agreement and may contribute to the development of new and improved calcium scoring methods. KEY POINTS: Segment-level coronary artery calcium scoring is a tedious and error-prone task. The proposed multi-task model achieved good agreement with a human observer on the segment level. Deep learning can contribute to the automation of segment-level coronary artery calcium scoring.

M3T-LM: A multi-modal multi-task learning model for jointly predicting patient length of stay and mortality.

Ensuring accurate predictions of inpatient length of stay (LoS) and mortality rates is essential for enhancing hospital service efficiency, particularly in light of the constraints posed by limited healthcare resources. Integrative analysis of heterogeneous clinic record data from different sources can hold great promise for improving the prognosis and diagnosis level of LoS and mortality. Currently, most existing studies solely focus on single data modality or tend to single-task learning, i.e., training LoS and mortality tasks separately. This limits the utilization of available multi-modal data and prevents the sharing of feature representations that could capture correlations between different tasks, ultimately hindering the model's performance. To address the challenge, this study proposes a novel Multi-Modal Multi-Task learning model, termed as M3T-LM, to integrate clinic records to predict inpatients' LoS and mortality simultaneously. The M3T-LM framework incorporates multiple data modalities by constructing sub-models tailored to each modality. Specifically, a novel attention-embedded one-dimensional (1D) convolutional neural network (CNN) is designed to handle numerical data. For clinical notes, they are converted into sequence data, and then two long short-term memory (LSTM) networks are exploited to model on textual sequence data. A two-dimensional (2D) CNN architecture, noted as CRXMDL, is designed to extract high-level features from chest X-ray (CXR) images. Subsequently, multiple sub-models are integrated to formulate the M3T-LM to capture the correlations between patient LoS and modality prediction tasks. The efficiency of the proposed method is validated on the MIMIC-IV dataset. The proposed method attained a test MAE of 5.54 for LoS prediction and a test F1 of 0.876 for mortality prediction. The experimental results demonstrate that our approach outperforms state-of-the-art (SOTA) methods in tackling mixed regression and classification tasks.

SwinUNeCCt: bidirectional hash-based agent transformer for cervical cancer MRI image multi-task learning.

Cervical cancer is the fourth most common malignant tumor among women globally, posing a significant threat to women's health. In 2022, approximately 600,000 new cases were reported, and 340,000 deaths occurred due to cervical cancer. Magnetic resonance imaging (MRI) is the preferred imaging method for diagnosing, staging, and evaluating cervical cancer. However, manual segmentation of MRI images is time-consuming and subjective. Therefore, there is an urgent need for automatic segmentation models to identify cervical cancer lesions in MRI scans accurately. All MRIs in our research are from cervical cancer patients diagnosed by pathology at Tongren City People's Hospital. Strict data selection criteria and clearly defined inclusion and exclusion conditions were established to ensure data consistency and accuracy of research results. The dataset contains imaging data from 122 cervical cancer patients, with each patient having 100 pelvic dynamic contrast-enhanced MRI scans. Annotations were jointly completed by medical professionals from Universiti Putra Malaysia and the Radiology Department of Tongren City People's Hospital to ensure data accuracy and reliability. Additionally, a novel computer-aided diagnosis model named SwinUNeCCt is proposed. This model incorporates (i) A bidirectional hash-based agent multi-head self-attention mechanism, which optimizes the interaction between local and global features in MRI, aiding in more accurate lesion identification. (ii) Reduced computational complexity of the self-attention mechanism. The effectiveness of the SwinUNeCCt model has been validated through comparisons with state-of-the-art 3D medical models, including nnUnet, TransBTS, nnFormer, UnetR, UnesT, SwinUNetR, and SwinUNeLCsT. In semantic segmentation tasks without a classification module, the SwinUNeCCt model demonstrates excellent performance across multiple key metrics: achieving a 95HD of 6.25, an IoU of 0.669, and a DSC of 0.802, all of which are the best results among the compared models. Simultaneously, SwinUNeCCt strikes a good balance between computational efficiency and model complexity, requiring only 442.7 GFLOPs of computational power and 71.2 M parameters. Furthermore, in semantic segmentation tasks that include a classification module, the SwinUNeCCt model also exhibits powerful recognition capabilities. Although this slightly increases computational overhead and model complexity, its performance surpasses other comparative models. The SwinUNeCCt model demonstrates excellent performance in semantic segmentation tasks, achieving the best results among state-of-the-art 3D medical models across multiple key metrics. It balances computational efficiency and model complexity well, maintaining high performance even with the inclusion of a classification module.

Simultaneously mapping the 3D distributions of multiple heavy metals in an industrial site using deep learning and multisource auxiliary data.

Three-dimensional (3D) distributions of multiple soil pollutants in industrial site are crucial for risk assessment and remediation. Yet, their 3D prediction accuracies are often low because of the strong variability of pollutants and availability of 3D covariate data. This study proposed a patch-based multi-task convolution neural network (MT-CNN) model for simultaneously predicting the 3D distributions of Zn, Pb, Ni, and Cu at an industrial site. By integrating neighborhood patches from multisource covariates, the MT-CNN model captured both horizontal and vertical pollution information, and outperformed the widely-used methods such as random forest (RF), ordinary Kriging (OK), and inverse distance weighting (IDW) for all the 4 heavy metals, with R2 values of 0.58, 0.56, 0.29 and 0.23 for Zn, Pb, Ni and Cu, respectively. Besides, the MT-CNN model achieved more stable predictions with reasonable accuracy, in comparison with the single-task CNN model. These results highlighted the potential of the proposed MT-CNN in simultaneously mapping the 3D distributions of multiple pollutants, while balancing the model training, maintaining and accuracy for low-cost rapid assessment of soil pollution at industrial sites.

Detailed delineation of the fetal brain in diffusion MRI via multi-task learning.

Diffusion-weighted MRI is increasingly used to study the normal and abnormal development of fetal brain inutero. Recent studies have shown that dMRI can offer invaluable insights into the neurodevelopmental processes in the fetal stage. However, because of the low data quality and rapid brain development, reliable analysis of fetal dMRI data requires dedicated computational methods that are currently unavailable. The lack of automated methods for fast, accurate, and reproducible data analysis has seriously limited our ability to tap the potential of fetal brain dMRI for medical and scientific applications. In this work, we developed and validated a unified computational framework to (1) segment the brain tissue into white matter, cortical/subcortical gray matter, and cerebrospinal fluid, (2) segment 31 distinct white matter tracts, and (3) parcellate the brain's cortex and delineate the deep gray nuclei and white matter structures into 96 anatomically meaningful regions. We utilized a set of manual, semi-automatic, and automatic approaches to annotate 97 fetal brains. Using these labels, we developed and validated a multi-task deep learning method to perform the three computations. Our evaluations show that the new method can accurately carry out all three tasks, achieving a mean Dice similarity coefficient of 0.865 on tissue segmentation, 0.825 on white matter tract segmentation, and 0.819 on parcellation. The proposed method can greatly advance the field of fetal neuroimaging as it can lead to substantial improvements in fetal brain tractography, tract-specific analysis, and structural connectivity assessment.

A multi-view multi-label fast model for Auricularia cornea phenotype identification and classification.

The identification and classification of various phenotypic features of Auricularia cornea fruit bodies are crucial for quality grading and breeding efforts. The phenotypic features of Auricularia cornea fruit bodies encompass size, number, shape, color, pigmentation, and damage. These phenotypic features are distributed across various views of the fruit bodies, making the task of achieving both rapid and accurate identification and classification challenging. This paper proposes a novel multi-view multi-label fast network that integrates two different views of the Auricularia cornea fruiting body, enabling rapid and precise identification and classification of six phenotypic features simultaneously. Initially, a multi-view feature extraction model based on partial convolution was constructed. This model incorporates channel attention mechanisms to achieve rapid phenotypic feature extraction of the Auricularia cornea fruiting body. Subsequently, an efficient multi-task classifier was designed, based on class-specific residual attention, to ensure accurate classification of phenotypic features. Finally, task weights were dynamically adjusted based on heteroscedastic uncertainty, reducing the training complexity of the multi-task classification. The proposed network achieved a classification accuracy of 94.66% and an inference speed of 11.9 ms on an image dataset of dried Auricularia cornea fruiting bodies with three views and six labels. The results demonstrate that the proposed network can efficiently and accurately identify and classify all phenotypic features of Auricularia cornea.

Simultaneously predicting SPAD and water content in rice leaves using hyperspectral imaging with deep multi-task regression and transfer component analysis.

BACKGROUND: Water content and chlorophyll content are important indicators for monitoring rice growth status. Simultaneous detection of water content and chlorophyll content is of significance. Different varieties of rice show differences in phenotype, resulting in the difficulties of establishing a universal model. In this study, hyperspectral imaging was used to detect the Soil and Plant Analyzer Development (SPAD) values and water content of fresh rice leaves of three rice varieties (Jiahua 1, Xiushui 121 and Xiushui 134). RESULTS: Both partial least squares regression and convolutional neural networks were used to establish single-task and multi-task models. Transfer component analysis (TCA) was used as transfer learning to learn the common features to achieve an approximate identical distribution between any two varieties. Single-task and multi-task models were also built using the features of the source domain, and these models were applied to the target domain. These results indicated that for models of each rice variety the prediction accuracy of most multi-task models was close to that of single-task models. As for TCA, the results showed that the single-task model achieved good performance for all transfer learning tasks. CONCLUSION: Compared with the original model, good and differentiated results were obtained for the models using features learned by TCA for both the source domain and target domain. The multi-task models could be constructed to predict SPAD values and water content simultaneously and then transferred to another rice variety, which could improve the efficiency of model construction and realize rapid detection of rice growth indicators. © 2024 Society of Chemical Industry.

Automated analysis of fetal heart rate baseline/acceleration/deceleration using MTU-Net3 + model.

In clinical practice, obstetricians use visual interpretation of fetal heart rate (FHR) to diagnose fetal conditions, but inconsistencies among interpretations can hinder accuracy. This study introduces MTU-Net3+, a deep learning model designed for automated, multi-task FHR analysis, aiming to improve diagnostic accuracy and efficiency. The proposed MTU-Net3 + was built upon the UNet3 + architecture, incorporating an encoder, a decoder, full-scale skip connections, and a deep supervision module, and further integrates a self-attention mechanism and bidirectional Long Short-Term Memory layers to enhance its performance. The MTU-Net3 + model accepts the preprocessed 20-minute FHR signals as input, outputting categorical probabilities and baseline values for each time point. The proposed MTU-Net3 + model was trained on a subset of a public database, and was tested on the remaining data of the public database and a private database. In the remaining public datasets, this model achieved F1 scores of 84.21% for deceleration (F1.Dec) and 61.33% for acceleration (F1.Acc), with a Root Mean Square Baseline Difference (RMSD.BL) of 3.46 bpm, 0% of points with an absolute difference exceeding 15 bpm(D15bpm), a Synthetic Inconsistency Coefficient (SI) of 44.82%, and a Morphological Analysis Discordance Index (MADI) of 7.00%. On the private dataset, the model recorded an RMSD.BL of 1.37 bpm, 0% D15bpm, F1.Dec of 100%, F1.Acc of 87.50%, an SI of 12.20% and a MADI of 2.79%. The MTU-Net3 + model proposed in this study performed well in automated FHR analysis, demonstrating its potential as an effective tool in the field of fetal health assessment.

A Novel Deep Learning Model for Breast Tumor Ultrasound Image Classification with Lesion Region Perception.

Multi-task learning (MTL) methods are widely applied in breast imaging for lesion area perception and classification to assist in breast cancer diagnosis and personalized treatment. A typical paradigm of MTL is the shared-backbone network architecture, which can lead to information sharing conflicts and result in the decline or even failure of the main task's performance. Therefore, extracting richer lesion features and alleviating information-sharing conflicts has become a significant challenge for breast cancer classification. This study proposes a novel Multi-Feature Fusion Multi-Task (MFFMT) model to effectively address this issue. Firstly, in order to better capture the local and global feature relationships of lesion areas, a Contextual Lesion Enhancement Perception (CLEP) module is designed, which integrates channel attention mechanisms with detailed spatial positional information to extract more comprehensive lesion feature information. Secondly, a novel Multi-Feature Fusion (MFF) module is presented. The MFF module effectively extracts differential features that distinguish between lesion-specific characteristics and the semantic features used for tumor classification, and enhances the common feature information of them as well. Experimental results on two public breast ultrasound imaging datasets validate the effectiveness of our proposed method. Additionally, a comprehensive study on the impact of various factors on the model's performance is conducted to gain a deeper understanding of the working mechanism of the proposed framework.

DetSegDiff: A joint periodontal landmark detection and segmentation in intraoral ultrasound using edge-enhanced diffusion-based network.

Individuals with malocclusion require an orthodontic diagnosis and treatment plan based on the severity of their condition. Assessing and monitoring changes in periodontal structures before, during, and after orthodontic procedures is crucial, and intraoral ultrasound (US) imaging has been shown a promising diagnostic tool in imaging periodontium. However, accurately delineating and analyzing periodontal structures in US videos is a challenging task for clinicians, as it is time-consuming and subject to interpretation errors. This paper introduces DetSegDiff, an edge-enhanced diffusion-based network developed to simultaneously detect the cementoenamel junction (CEJ) and segment alveolar bone structure in intraoral US videos. An edge feature encoder is designed to enhance edge and texture information for precise delineation of periodontal structures. Additionally, we employed the spatial squeeze-attention module (SSAM) to extract more representative features to perform both detection and segmentation tasks at global and local levels. This study used 169 videos from 17 orthodontic patients for training purposes and was subsequently tested on 41 videos from 4 additional patients. The proposed method achieved a mean distance difference of 0.17 ± 0.19 mm for the CEJ and an average Dice score of 90.1% for alveolar bone structure. As there is a lack of multi-task benchmark networks, thorough experiments were undertaken to assess and benchmark the proposed method against state-of-the-art (SOTA) detection and segmentation individual networks. The experimental results demonstrated that DetSegDiff outperformed SOTA approaches, confirming the feasibility of using automated diagnostic systems for orthodontists.

Identification of genetic basis of brain imaging by group sparse multi-task learning leveraging summary statistics.

Brain imaging genetics is an evolving neuroscience topic aiming to identify genetic variations related to neuroimaging measurements of interest. Traditional linear regression methods have shown success, but their reliance on individual-level imaging and genetic data limits their applicability. Herein, we proposed S-GsMTLR, a group sparse multi-task linear regression method designed to harness summary statistics from genome-wide association studies (GWAS) of neuroimaging quantitative traits. S-GsMTLR directly employs GWAS summary statistics, bypassing the requirement for raw imaging genetic data, and applies multivariate multi-task sparse learning to these univariate GWAS results. It amalgamates the strengths of conventional sparse learning methods, including sophisticated modeling techniques and efficient feature selection. Additionally, we implemented a rapid optimization strategy to alleviate computational burdens by identifying genetic variants associated with phenotypes of interest across the entire chromosome. We first evaluated S-GsMTLR using summary statistics derived from the Alzheimer's Disease Neuroimaging Initiative. The results were remarkably encouraging, demonstrating its comparability to conventional methods in modeling and identification of risk loci. Furthermore, our method was evaluated with two additional GWAS summary statistics datasets: One focused on white matter microstructures and the other on whole brain imaging phenotypes, where the original individual-level data was unavailable. The results not only highlighted S-GsMTLR's ability to pinpoint significant loci but also revealed intriguing structures within genetic variations and loci that went unnoticed by GWAS. These findings suggest that S-GsMTLR is a promising multivariate sparse learning method in brain imaging genetics. It eliminates the need for original individual-level imaging and genetic data while demonstrating commendable modeling and feature selection capabilities.

Detailed delineation of the fetal brain in diffusion MRI via multi-task learning.

Diffusion-weighted MRI is increasingly used to study the normal and abnormal development of fetal brain inutero. Recent studies have shown that dMRI can offer invaluable insights into the neurodevelopmental processes in the fetal stage. However, because of the low data quality and rapid brain development, reliable analysis of fetal dMRI data requires dedicated computational methods that are currently unavailable. The lack of automated methods for fast, accurate, and reproducible data analysis has seriously limited our ability to tap the potential of fetal brain dMRI for medical and scientific applications. In this work, we developed and validated a unified computational framework to (1) segment the brain tissue into white matter, cortical/subcortical gray matter, and cerebrospinal fluid, (2) segment 31 distinct white matter tracts, and (3) parcellate the brain's cortex and delineate the deep gray nuclei and white matter structures into 96 anatomically meaningful regions. We utilized a set of manual, semi-automatic, and automatic approaches to annotate 97 fetal brains. Using these labels, we developed and validated a multi-task deep learning method to perform the three computations. Our evaluations show that the new method can accurately carry out all three tasks, achieving a mean Dice similarity coefficient of 0.865 on tissue segmentation, 0.825 on white matter tract segmentation, and 0.819 on parcellation. The proposed method can greatly advance the field of fetal neuroimaging as it can lead to substantial improvements in fetal brain tractography, tract-specific analysis, and structural connectivity assessment.

AnyFace++: Deep Multi-Task, Multi-Domain Learning for Efficient Face AI.

Accurate face detection and subsequent localization of facial landmarks are mandatory steps in many computer vision applications, such as emotion recognition, age estimation, and gender identification. Thanks to advancements in deep learning, numerous facial applications have been developed for human faces. However, most have to employ multiple models to accomplish several tasks simultaneously. As a result, they require more memory usage and increased inference time. Also, less attention is paid to other domains, such as animals and cartoon characters. To address these challenges, we propose an input-agnostic face model, AnyFace++, to perform multiple face-related tasks concurrently. The tasks are face detection and prediction of facial landmarks for human, animal, and cartoon faces, including age estimation, gender classification, and emotion recognition for human faces. We trained the model using deep multi-task, multi-domain learning with a heterogeneous cost function. The experimental results demonstrate that AnyFace++ generates outcomes comparable to cutting-edge models designed for specific domains.

A Self-Supervised Few-Shot Semantic Segmentation Method Based on Multi-Task Learning and Dense Attention Computation.

Nowadays, autonomous driving technology has become widely prevalent. The intelligent vehicles have been equipped with various sensors (e.g., vision sensors, LiDAR, depth cameras etc.). Among them, the vision systems with tailored semantic segmentation and perception algorithms play critical roles in scene understanding. However, the traditional supervised semantic segmentation needs a large number of pixel-level manual annotations to complete model training. Although few-shot methods reduce the annotation work to some extent, they are still labor intensive. In this paper, a self-supervised few-shot semantic segmentation method based on Multi-task Learning and Dense Attention Computation (dubbed MLDAC) is proposed. The salient part of an image is split into two parts; one of them serves as the support mask for few-shot segmentation, while cross-entropy losses are calculated between the other part and the entire region with the predicted results separately as multi-task learning so as to improve the model's generalization ability. Swin Transformer is used as our backbone to extract feature maps at different scales. These feature maps are then input to multiple levels of dense attention computation blocks to enhance pixel-level correspondence. The final prediction results are obtained through inter-scale mixing and feature skip connection. The experimental results indicate that MLDAC obtains 55.1% and 26.8% one-shot mIoU self-supervised few-shot segmentation on the PASCAL-5i and COCO-20i datasets, respectively. In addition, it achieves 78.1% on the FSS-1000 few-shot dataset, proving its efficacy.

Bone age assessment by multi-granularity and multi-attention feature encoding.

Background: Bone age assessment (BAA) is crucial for the diagnosis of growth disorders and the optimization of treatments. However, the random error caused by different observers' experiences and the low consistency of repeated assessments harms the quality of such assessments. Thus, automated assessment methods are needed. Methods: Previous research has sought to design localization modules in a strongly or weakly supervised fashion to aggregate part regions to better recognize subtle differences. Conversely, we sought to efficiently deliver information between multi-granularity regions for fine-grained feature learning and to directly model long-distance relationships for global understanding. The proposed method has been named the "Multi-Granularity and Multi-Attention Net (2M-Net)". Specifically, we first applied the jigsaw method to generate related tasks emphasizing regions with different granularities, and we then trained the model on these tasks using a hierarchical sharing mechanism. In effect, the training signals from the extra tasks created as an inductive bias, enabling 2M-Net to discover task relatedness without the need for annotations. Next, the self-attention mechanism acted as a plug-and-play module to effectively enhance the feature representation capabilities. Finally, multi-scale features were applied for prediction. Results: A public data set of 14,236 hand radiographs, provided by the Radiological Society of North America (RSNA), was used to develop and validate 2M-Net. In the public benchmark testing, the mean absolute error (MAE) between the bone age estimates of the model and of the reviewer was 3.98 months (3.89 months for males and 4.07 months for females). Conclusions: By using the jigsaw method to construct a multi-task learning strategy and inserting the self-attention module for efficient global modeling, we established 2M-Net, which is comparable to the previous best method in terms of performance.

Multi-task heterogeneous graph learning on electronic health records.

Learning electronic health records (EHRs) has received emerging attention because of its capability to facilitate accurate medical diagnosis. Since the EHRs contain enriched information specifying complex interactions between entities, modeling EHRs with graphs is shown to be effective in practice. The EHRs, however, present a great degree of heterogeneity, sparsity, and complexity, which hamper the performance of most of the models applied to them. Moreover, existing approaches modeling EHRs often focus on learning the representations for a single task, overlooking the multi-task nature of EHR analysis problems and resulting in limited generalizability across different tasks. In view of these limitations, we propose a novel framework for EHR modeling, namely MulT-EHR (Multi-Task EHR), which leverages a heterogeneous graph to mine the complex relations and model the heterogeneity in the EHRs. To mitigate the large degree of noise, we introduce a denoising module based on the causal inference framework to adjust for severe confounding effects and reduce noise in the EHR data. Additionally, since our model adopts a single graph neural network for simultaneous multi-task prediction, we design a multi-task learning module to leverage the inter-task knowledge to regularize the training process. Extensive empirical studies on MIMIC-III and MIMIC-IV datasets validate that the proposed method consistently outperforms the state-of-the-art designs in four popular EHR analysis tasks - drug recommendation, and predictions of the length of stay, mortality, and readmission. Thorough ablation studies demonstrate the robustness of our method upon variations to key components and hyperparameters.

Real-time estimation of the optimal coil placement in transcranial magnetic stimulation using multi-task deep learning.

Transcranial magnetic stimulation (TMS) has emerged as a promising neuromodulation technique with both therapeutic and diagnostic applications. As accurate coil placement is known to be essential for focal stimulation, computational models have been established to help find the optimal coil positioning by maximizing electric fields at the cortical target. While these numerical simulations provide realistic and subject-specific field distributions, they are computationally demanding, precluding their use in real-time applications. In this paper, we developed a novel multi-task deep neural network which simultaneously predicts the optimal coil placement for a given cortical target as well as the associated TMS-induced electric field. Trained on large amounts of preceding numerical optimizations, the Attention U-Net-based neural surrogate provided accurate coil optimizations in only 35 ms, a fraction of time compared to the state-of-the-art numerical framework. The mean errors on the position estimates were below 2 mm, i.e., smaller than previously reported manual coil positioning errors. The predicted electric fields were also highly correlated (r> 0.97) with their numerical references. In addition to healthy subjects, we validated our approach also in glioblastoma patients. We first statistically underlined the importance of using realistic heterogeneous tumor conductivities instead of simply adopting values from the surrounding healthy tissue. Second, applying the trained neural surrogate to tumor patients yielded similar accurate positioning and electric field estimates as in healthy subjects. Our findings provide a promising framework for future real-time electric field-optimized TMS applications.

A Hierarchical Multi-Task Learning Framework for Semantic Annotation in Tabular Data.

To optimize the utilization and analysis of tables, it is essential to recognize and understand their semantics comprehensively. This requirement is especially critical given that many tables lack explicit annotations, necessitating the identification of column types and inter-column relationships. Such identification can significantly augment data quality, streamline data integration, and support data analysis and mining. Current table annotation models often address each subtask independently, which may result in the neglect of constraints and contextual information, causing relational ambiguities and inference errors. To address this issue, we propose a unified multi-task learning framework capable of concurrently handling multiple tasks within a single model, including column named entity recognition, column type identification, and inter-column relationship detection. By integrating these tasks, the framework exploits their interrelations, facilitating the exchange of shallow features and the sharing of representations. Their cooperation enables each task to leverage insights from the others, thereby improving the performance of individual subtasks and enhancing the model's overall generalization capabilities. Notably, our model is designed to employ only the internal information of tabular data, avoiding reliance on external context or knowledge graphs. This design ensures robust performance even with limited input information. Extensive experiments demonstrate the superior performance of our model across various tasks, validating the effectiveness of unified multi-task learning framework in the recognition and comprehension of table semantics.

A comprehensive multi-task deep learning approach for predicting metabolic syndrome with genetic, nutritional, and clinical data.

Metabolic syndrome (MetS) is a complex disorder characterized by a cluster of metabolic abnormalities, including abdominal obesity, hypertension, elevated triglycerides, reduced high-density lipoprotein cholesterol, and impaired glucose tolerance. It poses a significant public health concern, as individuals with MetS are at an increased risk of developing cardiovascular diseases and type 2 diabetes. Early and accurate identification of individuals at risk for MetS is essential. Various machine learning approaches have been employed to predict MetS, such as logistic regression, support vector machines, and several boosting techniques. However, these methods use MetS as a binary status and do not consider that MetS comprises five components. Therefore, a method that focuses on these characteristics of MetS is needed. In this study, we propose a multi-task deep learning model designed to predict MetS and its five components simultaneously. The benefit of multi-task learning is that it can manage multiple tasks with a single model, and learning related tasks may enhance the model's predictive performance. To assess the efficacy of our proposed method, we compared its performance with that of several single-task approaches, including logistic regression, support vector machine, CatBoost, LightGBM, XGBoost and one-dimensional convolutional neural network. For the construction of our multi-task deep learning model, we utilized data from the Korean Association Resource (KARE) project, which includes 352,228 single nucleotide polymorphisms (SNPs) from 7729 individuals. We also considered lifestyle, dietary, and socio-economic factors that affect chronic diseases, in addition to genomic data. By evaluating metrics such as accuracy, precision, F1-score, and the area under the receiver operating characteristic curve, we demonstrate that our multi-task learning model surpasses traditional single-task machine learning models in predicting MetS.