Multitask Deep Learning Model with Efficient Encoding Layer and Enhanced Parallel Convolution Block

To develop and validate a multi-task deep learning (MTDL) model using multiphase contrast-enhanced CT (CECT) for simultaneously assessing histological subtypes, clinical stages, and anatomical complexity grades of solid malignant renal tumors. This two-center retrospective study included patients with solid malignant renal tumors and their preoperative kidney CECT images. A progressive layered extraction (PLE)-based MTDL model was trained and externally tested. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC) and decision curve analysis (DCA), compared with the results of five radiologists. Among 798 patients (mean age, 54 ± 12 years; 279 females; Center A: n = 620, Center B: n = 178), 597 (74.8%) had clear cell renal cell carcinomas (ccRCC), 150 (18.8%) were clinical staging III/IV, and 187 (23.4%) had high anatomical complexity. On the external test set, the MTDL model achieved AUCs of 0.89 (95% CI: 0.82, 0.94) for distinguishing ccRCC from non-ccRCC, 0.87 (95% CI: 0.81, 0.93) for clinical staging (I/II vs. III/IV), and 0.87 (95% CI: 0.82, 0.92) for anatomical complexity grading (low-intermediate vs. high). The MTDL model outperformed single-task deep learning (STDL) in clinical staging (AUC: 0.87 vs. 0.82, p = 0.022), showed higher net benefit on DCA, and demonstrated better diagnostic performance than junior radiologists in histological subtyping and clinical staging. Additionally, it used 68% less memory and was 60% faster than STDL models. The CECT-based MTDL model demonstrated robust performance in simultaneously predicting histological subtypes, clinical stages, and anatomical complexity grades of malignant renal tumors. Question Accurate preoperative description of the histological subtyping, clinical staging, and anatomical complexity of malignant renal tumors is crucial for treatment decision-making. Findings By sharing features, the multi-task deep learning algorithm model enhances clinical staging performance and significantly improves computational efficiency in predicting all three tasks simultaneously. Clinical relevance The multi-task deep learning algorithm model enables rapid and accurate comprehensive preoperative evaluation of renal tumors, which assists surgeons in optimizing surgical plans and promotes the advancement of renal tumor management toward precision and efficiency.

Face recognition with wearable items has been a challenging task in computer vision and involves the problem of identifying humans wearing a face mask. Masked face analysis via multi-task learning could effectively improve performance in many fields of face analysis. In this paper, we propose a unified framework for predicting the age, gender, and emotions of people wearing face masks. We first construct FGNET-MASK, a masked face dataset for the problem. Then, we propose a multi-task deep learning model to tackle the problem. In particular, the multi-task deep learning model takes the data as inputs and shares their weight to yield predictions of age, expression, and gender for the masked face. Through extensive experiments, the proposed framework has been found to provide a better performance than other existing methods.

Glaucoma detection from retinal fundus images using graph convolution based multi-task model

BackgroundThe precise prediction of epidermal growth factor receptor (EGFR) mutation status and gross tumor volume (GTV) segmentation are crucial goals in computer-aided lung adenocarcinoma brain metastasis diagnosis. However, these two tasks present continuous difficulties due to the nonuniform intensity distributions, ambiguous boundaries, and variable shapes of brain metastasis (BM) in MR images.The existing approaches for tackling these challenges mainly rely on single-task algorithms, which overlook the interdependence between these two tasks.MethodsTo comprehensively address these challenges, we propose a multi-task deep learning model that simultaneously enables GTV segmentation and EGFR subtype classification. Specifically, a multi-scale self-attention encoder that consists of a convolutional self-attention module is designed to extract the shared spatial and global information for a GTV segmentation decoder and an EGFR genotype classifier. Then, a hybrid CNN-Transformer classifier consisting of a convolutional block and a Transformer block is designed to combine the global and local information. Furthermore, the task correlation and heterogeneity issues are solved with a multi-task loss function, aiming to balance the above two tasks by incorporating segmentation and classification loss functions with learnable weights.ResultsThe experimental results demonstrate that our proposed model achieves excellent performance, surpassing that of single-task learning approaches. Our proposed model achieves a mean Dice score of 0.89 for GTV segmentation and an EGFR genotyping accuracy of 0.88 on an internal testing set, and attains an accuracy of 0.81 in the EGFR genotype prediction task and an average Dice score of 0.85 in the GTV segmentation task on the external testing set. This shows that our proposed method has outstanding performance and generalization.ConclusionWith the introduction of an efficient feature extraction module, a hybrid CNN-Transformer classifier, and a multi-task loss function, the proposed multi-task deep learning network significantly enhances the performance achieved in both GTV segmentation and EGFR genotyping tasks. Thus, the model can serve as a noninvasive tool for facilitating clinical treatment.

Deep learning analysis of electrocardiography (ECG) may predict cardiovascular outcomes. We present a novel multi-task deep learning model, the ECG-MACE, which predicts the one-year first-ever major adverse cardiovascular events (MACE) using 2,821,889 standard 12-lead ECGs, including training (n = 984,895), validation (n = 422,061), and test (n = 1,414,933) sets, from Chang Gung Memorial Hospital database in Taiwan. Data from another independent medical center (n = 113,224) was retrieved for external validation. The model’s performance achieves AUROCs of 0.90 for heart failure (HF), 0.85 for myocardial infarction (MI), 0.76 for ischemic stroke (IS), and 0.89 for mortality. Furthermore, it outperforms the Framingham risk score at 5-year MACEs and 10-year mortality prediction. Over 10-year follow-ups, the model-predicted-positive group exhibits significantly higher MACE incidences than the model-predicted-negative group (relative incidence ratio: HF: 15.28; MI: 7.87; IS: 4.74; mortality: 13.18). Using solely ECGs, ECG-MACE effectively predicts one-year events and exhibits long-term anticipation. It provides potential applications in preventive medicine.

Accurate detection of the Southwestern Atlantic Frontal Zones (SAFZ) requires high-precision detection of both its intensity and position, which is essential for understanding climate change and the thermohaline balance of the global ocean. The intensity of Frontal Zones (FZ) serves as the foundation for determining their position, making precise detection crucial. However, existing methods often overlook intensity detection and treat FZ intensity and position detection as separate tasks, leading to reduced computational efficiency and a lack of coupling between intensity and position detection results, which ultimately compromises accuracy. Additionally, traditional methods typically represent FZ position using frontal lines, which fail to capture the full width of the FZ, resulting in incomplete representations. To address these limitations, we propose a multi-task deep learning semantic segmentation model, named Multi-Task Attention D-LinkNet (MTAD-LinkNet), which utilizes D-LinkNet as the backbone. This model enables the simultaneous detection of both SAFZ intensity and position. We construct an intensity-based SAFZ dataset using daily sea surface temperature (SST) data from satellite observations spanning 2010-2019, ensuring comprehensive FZ representations. The dataset is then used to train and evaluate the model. Experimental results show that MTAD-LinkNet accurately detects both SAFZ intensity and position while significantly improving computational efficiency. Furthermore, a seasonal detection experiment confirms that the model maintains coupling between intensity and position detection, achieving high-precision results.

Maize is an essential cereal for humans and animals worldwide, and it is one of the staple food in Kenya. One of the main challenges facing the maize crop in Kenya is the presence of diseases spreading quickly. Early recognition of maize pathogen and disease help at preventing the disease from spreading throughout the field. This paper proposes a regularized Multitask learning (MTL)–Convolutional Neural Networks (CNN) model for simultaneously identifying maize disease and its pathogen from diseased maize images. MTL allows training one model for multiple tasks at a time, which may improve the accuracy of each task by taking advantage of their commonalities. Our baseline is made of two CNN classification models, one of them being overfitting. We then build an MTL based on the two models, which increases the test accuracy of the overfitting model from 60.08% to 74.48%. The results show that the accuracy rises to 77.44% while combining MTL to the Early stopping method. However, the test accuracy goes up to 85.22 percent when MTL is combined with Early Stopping and Transfer Learning. The model is deployed to an android mobile application for maize farmers as end-users which is very important for costs reduction and time saving.

ADMET property prediction via multi-task graph learning under adaptive auxiliary task selection

In machine learning, the relatedness across multiple tasks is usually complex and entangled. Due to dataset bias, the relatedness among tasks might be distorted and mislead the training of the models with solid learning ability, such as the multi-task neural networks. In this paper, we propose the idea of Relatedness Refinement Multi-Task Learning (RRMTDL) by introducing adversarial learning in the multi-task deep neural network to tackle the problem. The RRMTDL deep learning model restrains the misleading relatedness task by adversarial training and extracts information sharing across tasks with valuable relatedness. With RRMTDL, multi-task deep learning can enhance the task-specific representation for the major tasks by excluding the misleading relatedness. We design tests with various combinations of task-relatedness to validate the proposed model. Experimental results show that the RRMTDL model can effectively refine the task relatedness and prominently outperform other multi-task deep learning models in datasets with entangled task labels.

Multi-task learning (MTL) is a machine learning paradigm that improves the performance of each task by exploiting useful information contained in multiple related tasks. However, the relatedness of tasks can be exploited by attackers to launch data poisoning attacks, which has been demonstrated a big threat to single-task learning. In this paper, we provide the first study on the vulnerability of MTL. Specifically, we focus on multi-task relationship learning (MTRL) models, a popular subclass of MTL models where task relationships are quantized and are learned directly from training data. We formulate the problem of computing optimal poisoning attacks on MTRL as a bilevel program that is adaptive to arbitrary choice of target tasks and attacking tasks. We propose an efficient algorithm called PATOM for computing optimal attack strategies. PATOM leverages the optimality conditions of the subproblem of MTRL to compute the implicit gradients of the upper level objective function. Experimental results on real-world datasets show that MTRL models are very sensitive to poisoning attacks and the attacker can significantly degrade the performance of target tasks, by either directly poisoning the target tasks or indirectly poisoning the related tasks exploiting the task relatedness. We also found that the tasks being attacked are always strongly correlated, which provides a clue for defending against such attacks.

The aim of this study was to determine the invasiveness of ground-glass nodules (GGNs) using a 3D multi-task deep learning network. We propose a novel architecture based on 3D multi-task learning to determine the invasiveness of GGNs. In total, 770 patients with 909 GGNs who underwent lung CT scans were enrolled. The patients were divided into the training (n = 626) and test sets (n = 144). In the test set, invasiveness was classified using deep learning into three categories: atypical adenomatous hyperplasia (AAH) and adenocarcinoma in situ (AIS), minimally invasive adenocarcinoma (MIA), and invasive pulmonary adenocarcinoma (IA). Furthermore, binary classifications (AAH/AIS/MIA vs. IA) were made by two thoracic radiologists and compared with the deep learning results. In the three-category classification task, the sensitivity, specificity, and accuracy were 65.41%, 82.21%, and 64.9%, respectively. In the binary classification task, the sensitivity, specificity, accuracy, and area under the ROC curve (AUC) values were 69.57%, 95.24%, 87.42%, and 0.89, respectively. In the visual assessment of GGN invasiveness of binary classification by the two thoracic radiologists, the sensitivity, specificity, and accuracy of the senior and junior radiologists were 58.93%, 90.51%, and 81.35% and 76.79%, 55.47%, and 61.66%, respectively. The proposed multi-task deep learning model achieved good classification results in determining the invasiveness of GGNs. This model may help to select patients with invasive lesions who need surgery and the proper surgical methods. • The proposed multi-task model has achieved good classification results for the invasiveness of GGNs. • The proposed network includes a classification and segmentation branch to learn global and regional features, respectively. • The multi-task model could assist doctors in selecting patients with invasive lesions who need surgery and choosing appropriate surgical methods.

<sec> <title>BACKGROUND</title> Personal narratives dominate health communication on social media platforms, integrating medical information with lived experiences and emotional framing. However, conventional computational systems exhibit "narrative blindness" which is an inability to process the causal logic and interpretive dimensions of health narratives, limiting their effectiveness in assessing information credibility and detecting potentially harmful health misinformation being shared in cyberspace. </sec> <sec> <title>OBJECTIVE</title> This study introduces and validates the Computational Narrative Builder, a novel multi-task deep learning model that reconstructs unstructured online health narratives into structured, machine-readable representations. By operationalizing Labovian sociolinguistic narrative theory through Agent-Action-Outcome (AAO) triplets enriched with interpretive annotations, we ask: Can transformer-based AAO models achieve superior recall and interpretability versus traditional approaches? </sec> <sec> <title>METHODS</title> We operationalized Labov and Waletzky's six-stage narrative framework into a computational AAO schema, representing narrative events as structured triplets: Agent (actor), Action (intervention), and Outcome (result). From 5253 Reddit health threads with 185,181 unique comments and 2,295,242 words collected across six health-focused subreddits (February 2019–November 2024), we created a gold-standard dataset through expert annotation of 1000 posts (2000 narrative segments), achieving substantial inter-annotator agreement (Fleiss κ=0.76 for narrative segmentation, κ=0.81 for AAO extraction at span level). We developed a multi-task DistilBERT architecture with five parallel task heads: AAO extraction (using Conditional Random Fields), narrative segmentation (six Labovian stages), emotion classification (8 classes), stance classification (5 classes), and tone classification (6 classes). The model was evaluated using 5-fold stratified cross-validation with bootstrap resampling (n=1000) against TF-IDF and single-task BERT baselines. </sec> <sec> <title>RESULTS</title> The Computational Narrative Builder AAO span extraction achieved micro-averaged F1 = 0.76 (exact-match), outperforming a single-task BERT baseline (0.71; P = .008) and a TF-IDF baseline (0.60; P&lt;.001). Overlap-tolerant scoring yielded micro-F1 = 0.87 (IoU ≥ 0.5). Component-wise F1 was 0.78 (actions), 0.75 (outcomes), and 0.74 (agents). Exact-match triplet accuracy = 0.57. Narrative stage classification reached micro-F1 = 0.84. Interpretive classification was robust: emotion (macro F1 = 0.80), stance (0.75), and tone (0.68). Multi-task learning produced significant gains vs single-task (+4 percentage points for AAO, P = .008; +3 points for segmentation, P = .008). Processing averaged 4.35 narratives/second. </sec> <sec> <title>CONCLUSIONS</title> The Computational Narrative Builder successfully addresses narrative blindness by providing a validated, efficient method for translating unstructured health narratives into structured representations. This enables systematic analysis of health narratives at scale, establishing foundations for advanced applications in credibility assessment and public health surveillance. While validated on English-language Reddit data, the framework's theoretical grounding suggests potential cross-platform generalizability, pending future validation. </sec>

Magnetotelluric (MT) forward modeling is a key technique in magnetotelluric sounding, and deep learning has been widely applied to MT forward modeling. In three-dimensional (3-D) problems, although existing methods can predict forward modeling results with high accuracy, they often use multiple networks to simulate multiple forward modeling parameters, resulting in low efficiency. We apply multi-task learning (MTL) to 3-D MT forward modeling to achieve simultaneous inference of apparent resistivity and impedance phase, effectively improving overall efficiency. Furthermore, through comparative analysis of feature map differences in various decoder layers of the network, we identify the optimal branching point for multi-task learning decoders. This enhances the feature extraction capabilities of the network and improves the prediction accuracy of forward modeling parameters. Additionally, we introduce an uncertainty-based loss function to dynamically balance the learning weights between tasks, addressing the shortcomings of traditional loss functions. Experiments demonstrate that compared with single-task networks and existing multi-task networks, the proposed network (MT-FeatureNet) achieves the best results in terms of Structural Similarity Index Measure (SSIM), Mean Relative Error (MRE), and Mean Absolute Error (MAE). The proposed multi-task learning model not only improves the efficiency and accuracy of 3-D MT forward modeling but also provides a novel approach to the design of multi-task learning network structures.

Soil acts as a foundation for human survival and social development and soil quality has a great effect on the growth of agricultural products. Visible/near-infrared spectroscopy has been acknowledged as a rapid and non-destructive method for predicting soil properties, and multi-task learning is a preferable approach to analyze the correlation between the spectroscopy data and soil properties. However, current multi-task learning models with the soft parameter sharing structure extremely rely on the task relatedness. To tackle this limitation, we proposed PS-MTL-LUCAS, a multi-task learning with a partially shared structure in this study. An additional shared layer was utilized to learn the general informative representations and interact with each task-specific layer. The partially shared structure ensured the maximum information flow between layers, thereby boosting the prediction performance. Also, the SHapley Addictive exPlanations (SHAP) algorithm was adopted to extract the feature wavelengths of each soil property. PS-MTL-LUCAS was validated on the LUCAS topsoil dataset (2009), and the experimental result suggested that PS-MTL-LUCAS dominated state-of-the-art models by achieving the determination of coefficient at 0.945, 0.936, 0.413, 0.624, 0.837, 0.952, and 0.956 for pH, N, P, K, CEC, OC, and CaCO3, respectively. In summary, this study highlighted the use of the soil spectroscopy and multi-task learning techniques in the soil property prediction task and provided a very promising approach for soil studies.

Multi-task feature selection with sparse regularization to extract common and task-specific features