BMC Genomics. 2024 Feb 7;25(1):152. doi: 10.1186/s12864-023-09933-x.

ABSTRACT

BACKGROUND: The accurate prediction of genomic breeding values is central to genomic selection in both plant and animal breeding studies. Genomic prediction involves the use of thousands of molecular markers spanning the entire genome and therefore requires methods able to efficiently handle high dimensional data. Not surprisingly, machine learning methods are becoming widely advocated for and used in genomic prediction studies. These methods encompass different groups of supervised and unsupervised learning methods. Although several studies have compared the predictive performances of individual methods, studies comparing the predictive performance of different groups of methods are rare. However, such studies are crucial for identifying (i) groups of methods with superior genomic predictive performance and assessing (ii) the merits and demerits of such groups of methods relative to each other and to the established classical methods. Here, we comparatively evaluate the genomic predictive performance and informally assess the computational cost of several groups of supervised machine learning methods, specifically, regularized regression methods, deep, ensemble and instance-based learning algorithms, using one simulated animal breeding dataset and three empirical maize breeding datasets obtained from a commercial breeding program.

RESULTS: Our results show that the relative predictive performance and computational expense of the groups of machine learning methods depend upon both the data and target traits and that for classical regularized methods, increasing model complexity can incur huge computational costs but does not necessarily always improve predictive accuracy. Thus, despite their greater complexity and computational burden, neither the adaptive nor the group regularized methods clearly improved upon the results of their simple regularized counterparts. This rules out selection of one procedure among machine learning methods for routine use in genomic prediction. The results also show that, because of their competitive predictive performance, computational efficiency, simplicity and therefore relatively few tuning parameters, the classical linear mixed model and regularized regression methods are likely to remain strong contenders for genomic prediction.

CONCLUSIONS: The dependence of predictive performance and computational burden on target datasets and traits call for increasing investments in enhancing the computational efficiency of machine learning algorithms and computing resources.

PMID:38326768 | DOI:10.1186/s12864-023-09933-x

Med Biol Eng Comput. 2024 Feb 8. doi: 10.1007/s11517-024-03025-y. Online ahead of print.

ABSTRACT

Early intervention in tumors can greatly improve human survival rates. With the development of deep learning technology, automatic image segmentation has taken a prominent role in the field of medical image analysis. Manually segmenting kidneys on CT images is a tedious task, and due to the diversity of these images and varying technical skills of professionals, segmentation results can be inconsistent. To address this problem, a novel ASD-Net network is proposed in this paper for kidney and kidney tumor segmentation tasks. First, the proposed network employs newly designed Adaptive Spatial-channel Convolution Optimization (ASCO) blocks to capture anisotropic information in the images. Then, other newly designed blocks, i.e., Dense Dilated Enhancement Convolution (DDEC) blocks, are utilized to enhance feature propagation and reuse it across the network, thereby improving its segmentation accuracy. To allow the network to segment complex and small kidney tumors more effectively, the Atrous Spatial Pyramid Pooling (ASPP) module is incorporated in its middle layer. With its generalized pyramid feature, this module enables the network to better capture and understand context information at various scales within the images. In addition to this, the concurrent spatial and channel squeeze & excitation (scSE) attention mechanism is adopted to better comprehend and manage context information in the images. Additional encoding layers are also added to the base (U-Net) and connected to the original encoding layer through skip connections. The resultant enhanced U-Net structure allows for better extraction and merging of high-level and low-level features, further boosting the network's ability to restore segmentation details. In addition, the combined Binary Cross Entropy (BCE)-Dice loss is utilized as the network's loss function. Experiments, conducted on the KiTS19 dataset, demonstrate that the proposed ASD-Net network outperforms the existing segmentation networks according to all evaluation metrics used, except for recall in the case of kidney tumor segmentation, where it takes the second place after Attention-UNet.

PMID:38326677 | DOI:10.1007/s11517-024-03025-y

Eur J Nucl Med Mol Imaging. 2024 Feb 8. doi: 10.1007/s00259-024-06616-x. Online ahead of print.

ABSTRACT

PURPOSE: Total metabolic tumor volume (TMTV) segmentation has significant value enabling quantitative imaging biomarkers for lymphoma management. In this work, we tackle the challenging task of automated tumor delineation in lymphoma from PET/CT scans using a cascaded approach.

METHODS: Our study included 1418 2-[18F]FDG PET/CT scans from four different centers. The dataset was divided into 900 scans for development/validation/testing phases and 518 for multi-center external testing. The former consisted of 450 lymphoma, lung cancer, and melanoma scans, along with 450 negative scans, while the latter consisted of lymphoma patients from different centers with diffuse large B cell, primary mediastinal large B cell, and classic Hodgkin lymphoma cases. Our approach involves resampling PET/CT images into different voxel sizes in the first step, followed by training multi-resolution 3D U-Nets on each resampled dataset using a fivefold cross-validation scheme. The models trained on different data splits were ensemble. After applying soft voting to the predicted masks, in the second step, we input the probability-averaged predictions, along with the input imaging data, into another 3D U-Net. Models were trained with semi-supervised loss. We additionally considered the effectiveness of using test time augmentation (TTA) to improve the segmentation performance after training. In addition to quantitative analysis including Dice score (DSC) and TMTV comparisons, the qualitative evaluation was also conducted by nuclear medicine physicians.

RESULTS: Our cascaded soft-voting guided approach resulted in performance with an average DSC of 0.68 ± 0.12 for the internal test data from developmental dataset, and an average DSC of 0.66 ± 0.18 on the multi-site external data (n = 518), significantly outperforming (p < 0.001) state-of-the-art (SOTA) approaches including nnU-Net and SWIN UNETR. While TTA yielded enhanced performance gains for some of the comparator methods, its impact on our cascaded approach was found to be negligible (DSC: 0.66 ± 0.16). Our approach reliably quantified TMTV, with a correlation of 0.89 with the ground truth (p < 0.001). Furthermore, in terms of visual assessment, concordance between quantitative evaluations and clinician feedback was observed in the majority of cases. The average relative error (ARE) and the absolute error (AE) in TMTV prediction on external multi-centric dataset were ARE = 0.43 ± 0.54 and AE = 157.32 ± 378.12 (mL) for all the external test data (n = 518), and ARE = 0.30 ± 0.22 and AE = 82.05 ± 99.78 (mL) when the 10% outliers (n = 53) were excluded.

CONCLUSION: TMTV-Net demonstrates strong performance and generalizability in TMTV segmentation across multi-site external datasets, encompassing various lymphoma subtypes. A negligible reduction of 2% in overall performance during testing on external data highlights robust model generalizability across different centers and cancer types, likely attributable to its training with resampled inputs. Our model is publicly available, allowing easy multi-site evaluation and generalizability analysis on datasets from different institutions.

PMID:38326655 | DOI:10.1007/s00259-024-06616-x

J Imaging Inform Med. 2024 Feb 7. doi: 10.1007/s10278-024-01001-4. Online ahead of print.

ABSTRACT

In intraoperative brain cancer procedures, real-time diagnosis is essential for ensuring safe and effective care. The prevailing workflow, which relies on histological staining with hematoxylin and eosin (H&E) for tissue processing, is resource-intensive, time-consuming, and requires considerable labor. Recently, an innovative approach combining stimulated Raman histology (SRH) and deep convolutional neural networks (CNN) has emerged, creating a new avenue for real-time cancer diagnosis during surgery. While this approach exhibits potential, there exists an opportunity for refinement in the domain of feature extraction. In this study, we employ coherent Raman scattering imaging method and a self-supervised deep learning model (VQVAE2) to enhance the speed of SRH image acquisition and feature representation, thereby enhancing the capability of automated real-time bedside diagnosis. Specifically, we propose the VQSRS network, which integrates vector quantization with a proxy task based on patch annotation for analysis of brain tumor subtypes. Training on images collected from the SRS microscopy system, our VQSRS demonstrates a significant speed enhancement over traditional techniques (e.g., 20-30 min). Comparative studies in dimensionality reduction clustering confirm the diagnostic capacity of VQSRS rivals that of CNN. By learning a hierarchical structure of recognizable histological features, VQSRS classifies major tissue pathological categories in brain tumors. Additionally, an external semantic segmentation method is applied for identifying tumor-infiltrated regions in SRH images. Collectively, these findings indicate that this automated real-time prediction technique holds the potential to streamline intraoperative cancer diagnosis, providing assistance to pathologists in simplifying the process.

PMID:38326533 | DOI:10.1007/s10278-024-01001-4

Eur Radiol Exp. 2024 Feb 8;8(1):10. doi: 10.1186/s41747-023-00411-3.

ABSTRACT

BACKGROUND: Pretraining labeled datasets, like ImageNet, have become a technical standard in advanced medical image analysis. However, the emergence of self-supervised learning (SSL), which leverages unlabeled data to learn robust features, presents an opportunity to bypass the intensive labeling process. In this study, we explored if SSL for pretraining on non-medical images can be applied to chest radiographs and how it compares to supervised pretraining on non-medical images and on medical images.

METHODS: We utilized a vision transformer and initialized its weights based on the following: (i) SSL pretraining on non-medical images (DINOv2), (ii) supervised learning (SL) pretraining on non-medical images (ImageNet dataset), and (iii) SL pretraining on chest radiographs from the MIMIC-CXR database, the largest labeled public dataset of chest radiographs to date. We tested our approach on over 800,000 chest radiographs from 6 large global datasets, diagnosing more than 20 different imaging findings. Performance was quantified using the area under the receiver operating characteristic curve and evaluated for statistical significance using bootstrapping.

RESULTS: SSL pretraining on non-medical images not only outperformed ImageNet-based pretraining (p < 0.001 for all datasets) but, in certain cases, also exceeded SL on the MIMIC-CXR dataset. Our findings suggest that selecting the right pretraining strategy, especially with SSL, can be pivotal for improving diagnostic accuracy of artificial intelligence in medical imaging.

CONCLUSIONS: By demonstrating the promise of SSL in chest radiograph analysis, we underline a transformative shift towards more efficient and accurate AI models in medical imaging.

RELEVANCE STATEMENT: Self-supervised learning highlights a paradigm shift towards the enhancement of AI-driven accuracy and efficiency in medical imaging. Given its promise, the broader application of self-supervised learning in medical imaging calls for deeper exploration, particularly in contexts where comprehensive annotated datasets are limited.

PMID:38326501 | DOI:10.1186/s41747-023-00411-3

Nat Methods. 2024 Feb 7. doi: 10.1038/s41592-024-02174-0. Online ahead of print.

ABSTRACT

Deep learning models, such as AlphaFold2 and RosettaFold, enable high-accuracy protein structure prediction. However, large protein complexes are still challenging to predict due to their size and the complexity of interactions between multiple subunits. Here we present CombFold, a combinatorial and hierarchical assembly algorithm for predicting structures of large protein complexes utilizing pairwise interactions between subunits predicted by AlphaFold2. CombFold accurately predicted (TM-score >0.7) 72% of the complexes among the top-10 predictions in two datasets of 60 large, asymmetric assemblies. Moreover, the structural coverage of predicted complexes was 20% higher compared to corresponding Protein Data Bank entries. We applied the method on complexes from Complex Portal with known stoichiometry but without known structure and obtained high-confidence predictions. CombFold supports the integration of distance restraints based on crosslinking mass spectrometry and fast enumeration of possible complex stoichiometries. CombFold's high accuracy makes it a promising tool for expanding structural coverage beyond monomeric proteins.

PMID:38326495 | DOI:10.1038/s41592-024-02174-0

Sci Rep. 2024 Feb 7;14(1):3123. doi: 10.1038/s41598-024-53778-7.

ABSTRACT

As cardiovascular disorders are prevalent, there is a growing demand for reliable and precise diagnostic methods within this domain. Audio signal-based heart disease detection is a promising area of research that leverages sound signals generated by the heart to identify and diagnose cardiovascular disorders. Machine learning (ML) and deep learning (DL) techniques are pivotal in classifying and identifying heart disease from audio signals. This study investigates ML and DL techniques to detect heart disease by analyzing noisy sound signals. This study employed two subsets of datasets from the PASCAL CHALLENGE having real heart audios. The research process and visually depict signals using spectrograms and Mel-Frequency Cepstral Coefficients (MFCCs). We employ data augmentation to improve the model's performance by introducing synthetic noise to the heart sound signals. In addition, a feature ensembler is developed to integrate various audio feature extraction techniques. Several machine learning and deep learning classifiers are utilized for heart disease detection. Among the numerous models studied and previous study findings, the multilayer perceptron model performed best, with an accuracy rate of 95.65%. This study demonstrates the potential of this methodology in accurately detecting heart disease from sound signals. These findings present promising opportunities for enhancing medical diagnosis and patient care.

PMID:38326488 | DOI:10.1038/s41598-024-53778-7

Phys Med Biol. 2024 Feb 7. doi: 10.1088/1361-6560/ad2716. Online ahead of print.

ABSTRACT

OBJECTIVE: To mitigate the potential radiation risk, low-dose Single Photon Emission Computed Tomography (SPECT) is of increasing interest. Numerous deep learning-based methods have been developed to perform low-dose imaging while maintaining image quality. However, most existing methods seldom explore the unique inner-structure inherent within sinograms. In addition, traditional supervised learning methods require large-scale labeled data, where the normal-dose data serves as annotation and is intractable to acquire in low-dose imaging. In this study, we aim to develop a novel sinogram inner-structure-aware semi-supervised framework for the task of low-dose SPECT sinogram restoration.

APPROACH: The proposed framework retains the strengths of UNet, meanwhile introducing a sinogram-structure-based non-local neighbors graph neural network (SSN-GNN) module and a window-based K-nearest neighbors GNN (W-KNN-GNN) module to effectively exploit the inherent inner-structure within SPECT sinograms. Moreover, the proposed framework employs the mean teacher semi-supervised learning approach to leverage the information available in abundant unlabeled low-dose sinograms.

MAIN RESULTS: The datasets exploited in this study were acquired from the XCAT anthropomorphic digital phantoms, which provide realistic images for imaging research of various modalities. Quantitative as well as qualitative results demonstrate that the proposed framework achieves superior performance compared to several state-of-the-art reconstruction methods. To further validate the effectiveness of the proposed framework, ablation and robustness experiments were also performed. The experimental results show that each component of the proposed framework effectively improves the model performance, and the framework exhibits superior robustness with respect to various noise levels. Besides, the proposed semi-supervised paradigm showcases the efficacy of incorporating supplementary unlabeled low-dose sinograms.

SIGNIFICANCE: The proposed framework improves the quality of low-dose SPECT reconstructed images by utilizing sinogram inner-structure and incorporating supplementary unlabeled data, which provides an important tool for dose reduction without sacrificing the image quality.

PMID:38324896 | DOI:10.1088/1361-6560/ad2716

Phys Med Biol. 2024 Feb 7. doi: 10.1088/1361-6560/ad2717. Online ahead of print.

ABSTRACT

OBJECTIVE: Medical image affine registration is a crucial basis before using deformable registration. On the one hand, the traditional affine registration methods based on step-by-step optimization are very time-consuming, so these methods are not compatible with most real-time medical applications. On the other hand, convolutional neural networks are limited in modeling long-range spatial relationships of the features due to inductive biases, such as weight sharing and locality. This is not conducive to affine registration tasks. Therefore, the evolution of real-time and high-accuracy affine medical image registration algorithms is necessary for registration applications.

APPROACH: In this paper, we propose a deep learning-based coarse-to-fine global and local feature fusion architecture for fast affine registration, and we use an unsupervised approach for end-to-end training. We use multiscale convolutional kernels as our elemental convolutional blocks to enhance feature extraction. Then, to learn the long-range spatial relationships of the features, we propose a new affine registration framework with weighted global positional attention that fuses global feature mapping and local feature mapping. Moreover, the fusion regressor is designed to generate the affine parameters.

MAIN RESULTS: The additive fusion method can be adaptive to global mapping and local mapping, which improves affine registration accuracy without the center of mass initialization. In addition, the max pooling layer and the multiscale convolutional kernel coding module increase the ability of the model in affine registration.

SIGNIFICANCE: We validate the effectiveness of our method on the OASIS dataset with 414 3D MRI brain maps. Comprehensive results demonstrate that our method achieves state-of-the-art affine registration accuracy and very efficient runtimes.

PMID:38324893 | DOI:10.1088/1361-6560/ad2717

Anal Chem. 2024 Feb 7. doi: 10.1021/acs.analchem.3c03889. Online ahead of print.

ABSTRACT

Molecular vibrational spectroscopies, including infrared absorption and Raman scattering, provide molecular fingerprint information and are powerful tools for qualitative and quantitative analysis. They benefit from the recent development of deep-learning-based algorithms to improve the spectral, spatial, and temporal resolutions. Although a variety of deep-learning-based algorithms, including those to simultaneously extract the global and local spectral features, have been developed for spectral classification, the classification accuracy is still far from satisfactory when the difference becomes very subtle. Here, we developed a lightweight algorithm named patch-based convolutional encoder (PACE), which effectively improved the accuracy of spectral classification by extracting spectral features while balancing local and global information. The local information was captured well by segmenting the spectrum into patches with an appropriate patch size. The global information was extracted by constructing the correlation between different patches with depthwise separable convolutions. In the five open-source spectral data sets, PACE achieved a state-of-the-art performance. The more difficult the classification, the better the performance of PACE, compared with that of residual neural network (ResNet), vision transformer (ViT), and other commonly used deep learning algorithms. PACE helped improve the accuracy to 92.1% in Raman identification of pathogen-derived extracellular vesicles at different physiological states, which is much better than those of ResNet (85.1%) and ViT (86.0%). In general, the precise recognition and extraction of subtle differences offered by PACE are expected to facilitate vibrational spectroscopy to be a powerful tool toward revealing the relevant chemical reaction mechanisms in surface science or realizing the early diagnosis in life science.

PMID:38324760 | DOI:10.1021/acs.analchem.3c03889

IEEE Trans Neural Netw Learn Syst. 2024 Feb 7;PP. doi: 10.1109/TNNLS.2024.3356310. Online ahead of print.

ABSTRACT

This article studies the generalization of neural networks (NNs) by examining how a network changes when trained on a training sample with or without out-of-distribution (OoD) examples. If the network's predictions are less influenced by fitting OoD examples, then the network learns attentively from the clean training set. A new notion, dataset-distraction stability, is proposed to measure the influence. Extensive CIFAR-10/100 experiments on the different VGG, ResNet, WideResNet, ViT architectures, and optimizers show a negative correlation between the dataset-distraction stability and generalizability. With the distraction stability, we decompose the learning process on the training set S into multiple learning processes on the subsets of S drawn from simpler distributions, i.e., distributions of smaller intrinsic dimensions (IDs), and furthermore, a tighter generalization bound is derived. Through attentive learning, miraculous generalization in deep learning can be explained and novel algorithms can also be designed.

PMID:38324433 | DOI:10.1109/TNNLS.2024.3356310

IEEE Trans Med Imaging. 2024 Feb 7;PP. doi: 10.1109/TMI.2024.3363416. Online ahead of print.

ABSTRACT

Optical coherence tomography (OCT) can perform non-invasive high-resolution three-dimensional (3D) imaging and has been widely used in biomedical fields, while it is inevitably affected by coherence speckle noise which degrades OCT imaging performance and restricts its applications. Here we present a novel speckle-free OCT imaging strategy, named toward-ground-truth OCT (tGT-OCT), that utilizes unsupervised 3D deep-learning processing and leverages OCT 3D imaging features to achieve speckle-free OCT imaging. Specifically, our proposed tGT-OCT utilizes an unsupervised 3D-convolution deep-learning network trained using random 3D volumetric data to distinguish and separate speckle from real structures in 3D imaging volumetric space; moreover, tGT-OCT effectively further reduces speckle noise and reveals structures that would otherwise be obscured by speckle noise while preserving spatial resolution. Results derived from different samples demonstrated the high-quality speckle-free 3D imaging performance of tGT-OCT and its advancement beyond the previous state-of-the-art. The code is available online: https://github.com/Voluntino/tGT-OCT.

PMID:38324426 | DOI:10.1109/TMI.2024.3363416

JMIR Res Protoc. 2024 Feb 7;13:e46493. doi: 10.2196/46493.

ABSTRACT

BACKGROUND: Artificial intelligence (AI)-powered digital therapies that detect methamphetamine cravings via consumer devices have the potential to reduce health care disparities by providing remote and accessible care solutions to communities with limited care solutions, such as Native Hawaiian, Filipino, and Pacific Islander communities. However, Native Hawaiian, Filipino, and Pacific Islander communities are understudied with respect to digital therapeutics and AI health sensing despite using technology at the same rates as other racial groups.

OBJECTIVE: In this study, we aimed to understand the feasibility of continuous remote digital monitoring and ecological momentary assessments in Native Hawaiian, Filipino, and Pacific Islander communities in Hawaii by curating a novel data set of longitudinal Fitbit (Fitbit Inc) biosignals with the corresponding craving and substance use labels. We also aimed to develop personalized AI models that predict methamphetamine craving events in real time using wearable sensor data.

METHODS: We will develop personalized AI and machine learning models for methamphetamine use and craving prediction in 40 individuals from Native Hawaiian, Filipino, and Pacific Islander communities by curating a novel data set of real-time Fitbit biosensor readings and the corresponding participant annotations (ie, raw self-reported substance use data) of their methamphetamine use and cravings. In the process of collecting this data set, we will gain insights into cultural and other human factors that can challenge the proper acquisition of precise annotations. With the resulting data set, we will use self-supervised learning AI approaches, which are a new family of machine learning methods that allows a neural network to be trained without labels by being optimized to make predictions about the data. The inputs to the proposed AI models are Fitbit biosensor readings, and the outputs are predictions of methamphetamine use or craving. This paradigm is gaining increased attention in AI for health care.

RESULTS: To date, more than 40 individuals have expressed interest in participating in the study, and we have successfully recruited our first 5 participants with minimal logistical challenges and proper compliance. Several logistical challenges that the research team has encountered so far and the related implications are discussed.

CONCLUSIONS: We expect to develop models that significantly outperform traditional supervised methods by finetuning according to the data of a participant. Such methods will enable AI solutions that work with the limited data available from Native Hawaiian, Filipino, and Pacific Islander populations and that are inherently unbiased owing to their personalized nature. Such models can support future AI-powered digital therapeutics for substance abuse.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): DERR1-10.2196/46493.

PMID:38324375 | DOI:10.2196/46493

J Cell Sci. 2024 Feb 1;137(3):jcs261545. doi: 10.1242/jcs.261545. Epub 2024 Feb 7.

ABSTRACT

Fluorescence microscopy is essential for studying living cells, tissues and organisms. However, the fluorescent light that switches on fluorescent molecules also harms the samples, jeopardizing the validity of results - particularly in techniques such as super-resolution microscopy, which demands extended illumination. Artificial intelligence (AI)-enabled software capable of denoising, image restoration, temporal interpolation or cross-modal style transfer has great potential to rescue live imaging data and limit photodamage. Yet we believe the focus should be on maintaining light-induced damage at levels that preserve natural cell behaviour. In this Opinion piece, we argue that a shift in role for AIs is needed - AI should be used to extract rich insights from gentle imaging rather than recover compromised data from harsh illumination. Although AI can enhance imaging, our ultimate goal should be to uncover biological truths, not just retrieve data. It is essential to prioritize minimizing photodamage over merely pushing technical limits. Our approach is aimed towards gentle acquisition and observation of undisturbed living systems, aligning with the essence of live-cell fluorescence microscopy.

PMID:38324353 | DOI:10.1242/jcs.261545

J Orthop Res. 2024 Feb 7. doi: 10.1002/jor.25800. Online ahead of print.

ABSTRACT

In this study, we investigated the discriminative capacity of knee morphology in automatic detection of osteophytes defined by the Osteoarthritis Research Society International atlas, using X-ray and magnetic resonance imaging (MRI) data. For the X-ray analysis, we developed a deep learning (DL) based model to segment femur and tibia. In case of MRIs, we utilized previously validated segmentations of femur, tibia, corresponding cartilage tissues, and menisci. Osteophyte detection was performed using DL models in four compartments: medial femur (FM), lateral femur (FL), medial tibia (TM), and lateral tibia (TL). To analyze the confounding effects of soft tissues, we investigated their morphology in combination with bones, including bones+cartilage, bones+menisci, and all the tissues. From X-ray-based 2D morphology, the models yielded balanced accuracy of 0.73, 0.69, 0.74, and 0.74 for FM, FL, TM, TL, respectively. Using 3D bone morphology from MRI, balanced accuracy was 0.80, 0.77, 0.71, and 0.76, respectively. The performance was higher than in 2D for all the compartments except for TM, with significant improvements observed for femoral compartments. Adding menisci or cartilage morphology consistently improved balanced accuracy in TM, with the greatest improvement seen for small osteophyte. Otherwise, the models performed similarly to bones-only. Our experiments demonstrated that MRI-based models show higher detection capability than X-ray based models for identifying knee osteophytes. This study highlighted the feasibility of automated osteophyte detection from X-ray and MRI data and suggested further need for development of osteophyte assessment criteria in addition to OARSI, particularly, for early osteophytic changes.

PMID:38323840 | DOI:10.1002/jor.25800

Med Phys. 2024 Feb 7. doi: 10.1002/mp.16976. Online ahead of print.

ABSTRACT

BACKGROUND: MR-only radiotherapy treatment planning is an attractive alternative to conventional workflow, reducing scan time and ionizing radiation. It is crucial to derive the electron density map or synthetic CT (sCT) from MR data to perform dose calculations to enable MR-only treatment planning. Automatic segmentation of relevant organs in MR images can accelerate the process by preventing the time-consuming manual contouring step. However, the segmentation label is available only for CT data in many cases.

PURPOSE: We propose CycleSeg, a unified framework that generates sCT and corresponding segmentation from MR images without access to MR segmentation labels METHODS: CycleSeg utilizes the CycleGAN formulation to perform unpaired synthesis of sCT and image alignment. To enable MR (sCT) segmentation, CycleSeg incorporates unsupervised domain adaptation by using a pseudo-labeling approach with feature alignment in semantic segmentation space. In contrast to previous approaches that perform segmentation on MR data, CycleSeg could perform segmentation on both MR and sCT. Experiments were performed with data from prostate cancer patients, with 78/7/10 subjects in the training/validation/test sets, respectively.

RESULTS: CycleSeg showed the best sCT generation results, with the lowest mean absolute error of 102.2 and the lowest Fréchet inception distance of 13.0. CycleSeg also performed best on MR segmentation, with the highest average dice score of 81.0 and 81.1 for MR and sCT segmentation, respectively. Ablation experiments confirmed the contribution of the proposed components of CycleSeg.

CONCLUSION: CycleSeg effectively synthesized CT and performed segmentation on MR images of prostate cancer patients. Thus, CycleSeg has the potential to expedite MR-only radiotherapy treatment planning, reducing the prescribed scans and manual segmentation effort, and increasing throughput.

PMID:38323835 | DOI:10.1002/mp.16976

Pest Manag Sci. 2024 Feb 7. doi: 10.1002/ps.7990. Online ahead of print.

ABSTRACT

BACKGROUND: Machine vision-based precision weed management is a promising solution to substantially reduce herbicide input and weed control cost. The objective of this research was to compare two different deep learning-based approaches for detecting weeds in cabbage: (1) detecting weeds directly, and (2) detecting crops by generating the bounding boxes covering the crops and any green pixels outside the bounding boxes were deemed as weeds.

RESULTS: The precision, recall, F1-score, mAP0.5 , mAP0.5:0.95 of You Only Look Once (YOLO) v5 for detecting cabbage were 0.986, 0.979, 0.982, 0.995, and 0.851, respectively, while these metrics were 0.973, 0.985, 0.979, 0.993, and 0.906 for YOLOv8, respectively. However, none of these metrics exceeded 0.891 when detecting weeds. The reduced performances for directly detecting weeds could be attributed to the diverse weed species at varying densities and growth stages with different plant morphologies. A segmentation procedure demonstrated its effectiveness for extracting weeds outside the bounding boxes covering the crops, and thereby realizing effective indirect weed detection.

CONCLUSION: The indirect weed detection approach demands less manpower as the need for constructing a large training dataset containing a variety of weed species is unnecessary. However, in a certain case, weeds are likely to remain undetected due to their growth in close proximity with crops and being situated within the predicted bounding boxes that encompass the crops. The models generated in this research can be used in conjunction with the machine vision subsystem of a smart sprayer or mechanical weeder. © 2024 Society of Chemical Industry.

PMID:38323798 | DOI:10.1002/ps.7990

AJR Am J Roentgenol. 2024 Feb 7. doi: 10.2214/AJR.23.30272. Online ahead of print.

ABSTRACT

Background: Cine cardiac MRI sequences require repeated breath-holds, which can be difficult in patients with ischemic heart disease (IHD). Objective: To compare a free-breathing accelerated cine sequence using deep-learning (DL) reconstruction and a standard breath-hold cine sequence in terms of image quality and left-ventricular (LV) measurements in patients with IHD undergoing cardiac MRI. Methods: This prospective study included patients undergoing 1.5-T or 3-T cardiac MRI for evaluation of IHD between March 15, 2023, and June 21, 2023. Examinations included an investigational free-breathing cine short-axis sequence with DL reconstruction (cine-DL). Two radiologists (R1, R2), in blinded fashion, independently assessed LV ejection fraction (LVEF), LV end-diastolic volume (LVEDV), LV end-systolic volume (LVESV), and subjective image quality, for cine-DL sequence and standard breath-hold balanced SSFP sequences; R1 assessed artifacts. Results: The analysis included 26 patients (mean age, 64.3±11.7 years; 14 men, 12 women). Acquisition was shorter for cine-DL than standard sequence (0.6±0.1 min vs 2.4±0.6 min, p<.001). Cine-DL, in comparison with standard, showed no significant difference for LVEF for R1 (51.7±14.3% vs 51.3±14.7%, p=.56) or R2 (53.4±14.9% vs 52.8±14.6%, p=.53); significantly greater LVEDV for R2 (171.9±51.9 vs 160.6±49.4 ml, p=.01) but not R1 (171.8±53.7 vs 165.5±52.4 ml, p=.16); and no significant difference in LVESV for R1 (88.1±49.3 vs 86.0±50.5 ml, p=.45) or R2 (85.2±48.1 vs 81.3±48.2 ml, p=.10). Mean bias between cine-DL and standard was: LVEF, 0.4% for R1 and 0.7% for R2; LVEDV, 6.3 ml for R1 and 11.3 ml for R2; LVESV, 2.1 ml for R1 and 3.9 ml for R2. Subjective image quality was better for cine-DL than standard for R1 (2.3±0.5 vs 1.9±0.8, p=.02) and R2 (2.2±0.4 vs 1.9±0.7; p=.02). R1 reported no significant difference between cine-DL and standard for off-resonance (3.8% vs 23.1%, p=.10), and parallel-imaging (3.8% vs 19.2%, p=.19) artifacts; blurring artifacts were more frequent for cine-DL than standard (42.3% vs 7.7%; p=.008). Conclusion: Free-breathing cine-DL sequence, in comparison with standard breath-hold cine sequence, showed very small bias for LVEF measurements and better subjective quality. Cine-DL yielded greater LV volumes. Clinical Impact: Free-breathing cine-DL may yield reliable LVEF measurements in patients with IHD unable to repeatedly breath-hold.

PMID:38323784 | DOI:10.2214/AJR.23.30272

J Cutan Med Surg. 2024 Feb 7:12034754241229361. doi: 10.1177/12034754241229361. Online ahead of print.

ABSTRACT

BACKGROUND: Artificial intelligence (AI) in skin cancer is a promising research field to assist physicians and to provide support to patients remotely. Physicians' awareness to new developments in AI research is important to define the best practices and scope of integrating AI-enabled technologies within a clinical setting.

OBJECTIVES: To analyze the characteristics and trends of AI skin cancer publications from dermatology journals.

METHODS: AI skin cancer publications were retrieved in June 2022 from the Web of Science. Publications were screened by title, abstract, and keywords to assess eligibility. Publications were fully reviewed. Publications were divided between nonmelanoma skin cancer (NMSC), melanoma, and skin cancer studies. The primary measured outcome was the number of citations. The secondary measured outcomes were articles' general characteristics and features related to AI.

RESULTS: A total of 168 articles were included: 25 on NMSC, 77 on melanoma, and 66 on skin cancer. The most common types of skin cancers were melanoma (134, 79.8%), basal cell carcinoma (61, 36.3%), and squamous cell carcinoma (45, 26.9%). All articles were published between 2000 and 2022, with 49 (29.2%) of them being published in 2021. Original studies that developed or assessed an algorithm predominantly used supervised learning (66, 97.0%) and deep neural networks (42, 67.7%). The most used imaging modalities were standard dermoscopy (76, 45.2%) and clinical images (39, 23.2%).

CONCLUSIONS: Most publications focused on developing or assessing screening technologies with mainly deep neural network algorithms. This indicates the eminent need for dermatologists to label or annotate images used by novel AI systems.

PMID:38323537 | DOI:10.1177/12034754241229361

Curr Comput Aided Drug Des. 2024 Feb 6. doi: 10.2174/0115734099287389240126072433. Online ahead of print.

ABSTRACT

Traditional molecular de novo generation methods, such as evolutionary algorithms, generate new molecules mainly by linking existing atomic building blocks. The challenging issues in these methods include difficulty in synthesis, failure to achieve desired properties, and structural optimization requirements. Advances in deep learning offer new ideas for rational and robust de novo drug design. Deep learning, a branch of machine learning, is more efficient than traditional methods for processing problems, such as speech, image, and translation. This study provides a comprehensive overview of the current state of research in de novo drug design based on deep learning and identifies key areas for further development. Deep learning-based de novo drug design is pivotal in four key dimensions. Molecular databases form the basis for model training, while effective molecular representations impact model performance. Common DL models (GANs, RNNs, VAEs, CNNs, DMs) generate drug molecules with desired properties. The evaluation metrics guide research directions by determining the quality and applicability of generated molecules. This abstract highlights the foundational aspects of DL-based de novo drug design, offering a concise overview of its multifaceted contributions. Consequently, deep learning in de novo molecule generation has attracted more attention from academics and industry. As a result, many deep learning-based de novo molecule generation types have been actively proposed.

PMID:38321907 | DOI:10.2174/0115734099287389240126072433