Commun Biol. 2024 Apr 5;7(1):414. doi: 10.1038/s42003-024-06096-7.

ABSTRACT

Understanding the genetic architecture of brain structure is challenging, partly due to difficulties in designing robust, non-biased descriptors of brain morphology. Until recently, brain measures for genome-wide association studies (GWAS) consisted of traditionally expert-defined or software-derived image-derived phenotypes (IDPs) that are often based on theoretical preconceptions or computed from limited amounts of data. Here, we present an approach to derive brain imaging phenotypes using unsupervised deep representation learning. We train a 3-D convolutional autoencoder model with reconstruction loss on 6130 UK Biobank (UKBB) participants' T1 or T2-FLAIR (T2) brain MRIs to create a 128-dimensional representation known as Unsupervised Deep learning derived Imaging Phenotypes (UDIPs). GWAS of these UDIPs in held-out UKBB subjects (n = 22,880 discovery and n = 12,359/11,265 replication cohorts for T1/T2) identified 9457 significant SNPs organized into 97 independent genetic loci of which 60 loci were replicated. Twenty-six loci were not reported in earlier T1 and T2 IDP-based UK Biobank GWAS. We developed a perturbation-based decoder interpretation approach to show that these loci are associated with UDIPs mapped to multiple relevant brain regions. Our results established unsupervised deep learning can derive robust, unbiased, heritable, and interpretable brain imaging phenotypes.

PMID:38580839 | DOI:10.1038/s42003-024-06096-7

Sci Rep. 2024 Apr 5;14(1):8004. doi: 10.1038/s41598-024-58810-4.

ABSTRACT

Pelvic fractures pose significant challenges in medical diagnosis due to the complex structure of the pelvic bones. Timely diagnosis of pelvic fractures is critical to reduce complications and mortality rates. While computed tomography (CT) is highly accurate in detecting pelvic fractures, the initial diagnostic procedure usually involves pelvic X-rays (PXR). In recent years, many deep learning-based methods have been developed utilizing ImageNet-based transfer learning for diagnosing hip and pelvic fractures. However, the ImageNet dataset contains natural RGB images which are different than PXR. In this study, we proposed a two-step transfer learning approach that improved the diagnosis of pelvic fractures in PXR images. The first step involved training a deep convolutional neural network (DCNN) using synthesized PXR images derived from 3D-CT by digitally reconstructed radiographs (DRR). In the second step, the classification layers of the DCNN were fine-tuned using acquired PXR images. The performance of the proposed method was compared with the conventional ImageNet-based transfer learning method. Experimental results demonstrated that the proposed DRR-based method, using 20 synthesized PXR images for each CT, achieved superior performance with the area under the receiver operating characteristic curves (AUROCs) of 0.9327 and 0.8014 for visible and invisible fractures, respectively. The ImageNet-based method yields AUROCs of 0.8908 and 0.7308 for visible and invisible fractures, respectively.

PMID:38580737 | DOI:10.1038/s41598-024-58810-4

Sci Rep. 2024 Apr 5;14(1):8012. doi: 10.1038/s41598-024-58146-z.

ABSTRACT

The objective of human pose estimation (HPE) derived from deep learning aims to accurately estimate and predict the human body posture in images or videos via the utilization of deep neural networks. However, the accuracy of real-time HPE tasks is still to be improved due to factors such as partial occlusion of body parts and limited receptive field of the model. To alleviate the accuracy loss caused by these issues, this paper proposes a real-time HPE model called CCAM - Person based on the YOLOv8 framework. Specifically, we have improved the backbone and neck of the YOLOv8x-pose real-time HPE model to alleviate the feature loss and receptive field constraints. Secondly, we introduce the context coordinate attention module (CCAM) to augment the model's focus on salient features, reduce background noise interference, alleviate key point regression failure caused by limb occlusion, and improve the accuracy of pose estimation. Our approach attains competitive results on multiple metrics of two open-source datasets, MS COCO 2017 and CrowdPose. Compared with the baseline model YOLOv8x-pose, CCAM-Person improves the average precision by 2.8% and 3.5% on the two datasets, respectively.

PMID:38580704 | DOI:10.1038/s41598-024-58146-z

Nat Commun. 2024 Apr 5;15(1):2935. doi: 10.1038/s41467-024-47065-2.

ABSTRACT

Histopathology plays a critical role in the diagnosis and surgical management of cancer. However, access to histopathology services, especially frozen section pathology during surgery, is limited in resource-constrained settings because preparing slides from resected tissue is time-consuming, labor-intensive, and requires expensive infrastructure. Here, we report a deep-learning-enabled microscope, named DeepDOF-SE, to rapidly scan intact tissue at cellular resolution without the need for physical sectioning. Three key features jointly make DeepDOF-SE practical. First, tissue specimens are stained directly with inexpensive vital fluorescent dyes and optically sectioned with ultra-violet excitation that localizes fluorescent emission to a thin surface layer. Second, a deep-learning algorithm extends the depth-of-field, allowing rapid acquisition of in-focus images from large areas of tissue even when the tissue surface is highly irregular. Finally, a semi-supervised generative adversarial network virtually stains DeepDOF-SE fluorescence images with hematoxylin-and-eosin appearance, facilitating image interpretation by pathologists without significant additional training. We developed the DeepDOF-SE platform using a data-driven approach and validated its performance by imaging surgical resections of suspected oral tumors. Our results show that DeepDOF-SE provides histological information of diagnostic importance, offering a rapid and affordable slide-free histology platform for intraoperative tumor margin assessment and in low-resource settings.

PMID:38580633 | DOI:10.1038/s41467-024-47065-2

J Shoulder Elbow Surg. 2024 Apr 3:S1058-2746(24)00231-3. doi: 10.1016/j.jse.2024.02.029. Online ahead of print.

ABSTRACT

BACKGROUND: While multiple studies have tested the ability of large language models (LLM), such as ChatGPT, to pass standardized medical exams at different levels of training, LLMs have never been tested on surgical sub-specialty examinations, such as the American Shoulder and Elbow Surgeons (ASES) Maintenance of Certification (MOC). The purpose of this study was to compare results of ChatGPT 3.5, GPT-4, and fellowship-trained surgeons on the 2023 American Shoulder and Elbow Surgeons (ASES) Maintenance of Certification (MOC) self-assessment exam.

METHODS: ChatGPT 3.5 and GPT-4 were subjected to the same set of text-only questions from the ASES MOC exam, and GPT-4 was additionally subjected to image-based MOC exam questions. Question responses from both models were compared against the correct answers. Performance of both models was compared to corresponding average human performance on the same question subsets. One sided proportional z-test were utilized to analyze data.

RESULTS: Humans performed significantly better than Chat GPT 3.5 on exclusively text-based questions (76.4% vs. 60.8%, p= .044). Humans also performed significantly better than GPT 4 on image-based questions (73.9% vs. 53.2%, p= .019). There was no significant difference between humans and GPT 4 in text-based questions (76.4% vs. 66.7%, p=0.136). Accounting for all questions, humans significantly outperformed GPT-4 (75.3% vs. 60.2%, p= .012). GPT-4 did not perform statistically significantly betterer than ChatGPT 3.5 on text-only questions (66.7% vs. 60.8%, p= .268).

DISCUSSION: Although human performance was overall superior, ChatGPT demonstrated the capacity to analyze orthopedic information and answer specialty-specific questions on the ASES MOC exam for both text and image-based questions. With continued advancements in deep learning, large language models may someday rival exam performance of fellowship-trained surgeons.

PMID:38580067 | DOI:10.1016/j.jse.2024.02.029

Am J Cardiol. 2024 Apr 3:S0002-9149(24)00234-0. doi: 10.1016/j.amjcard.2024.03.035. Online ahead of print.

ABSTRACT

Peripheral artery disease (PAD) is associated with impaired blood flow in the lower extremities and histopathological changes of the skeletal calf muscles resulting in abnormal microvascular perfusion. We studied the use of convolutional neural networks (CNNs) to differentiate PAD patients from matched controls by utilizing perfusion pattern features from contrast-enhanced magnetic resonance imaging (CE-MRI) of the skeletal calf muscles. We acquired CE-MRI based skeletal calf muscle perfusion in 56 individuals (36 PAD patients, 20 matched controls). Microvascular perfusion imaging was performed post reactive hyperemia at the mid-calf level with a temporal resolution of 409 ms. We analyzed perfusion scans up to 2 minutes indexed from the local pre-contrast arrival time frame. Skeletal calf muscles including the anterior muscle (AM), lateral muscle (LM), deep posterior muscle group (DM), and the soleus (SM) and gastrocnemius muscles (GM) were segmented semi-automatically. Segmented muscles were represented as 3D DICOM stacks of CE-MRI perfusion scans for deep learning analysis. We tested several CNN models for the 3D CE-MRI perfusion stacks to classify PAD patients from matched controls. Two of the best performing CNNs (resNet and divNet) were selected to develop the final classification model. A peak accuracy of 75% was obtained for both resNet and divNet. Specificity was 80% and 94% for resNet and divNet, respectively. In conclusion, deep learning utilizing CNNs, and CE-MRI skeletal calf muscle perfusion can discriminate PAD patients from matched controls. Deep learning methods may be of interest for the study of PAD.

PMID:38580040 | DOI:10.1016/j.amjcard.2024.03.035

Drug Resist Updat. 2024 Mar 19;74:101080. doi: 10.1016/j.drup.2024.101080. Online ahead of print.

ABSTRACT

BACKGROUND: Gastric Cancer (GC) characteristically exhibits heterogeneous responses to treatment, particularly in relation to immuno plus chemo therapy, necessitating a precision medicine approach. This study is centered around delineating the cellular and molecular underpinnings of drug resistance in this context.

METHODS: We undertook a comprehensive multi-omics exploration of postoperative tissues from GC patients undergoing the chemo and immuno-treatment regimen. Concurrently, an image deep learning model was developed to predict treatment responsiveness.

RESULTS: Our initial findings associate apical membrane cells with resistance to fluorouracil and oxaliplatin, critical constituents of the therapy. Further investigation into this cell population shed light on substantial interactions with resident macrophages, underscoring the role of intercellular communication in shaping treatment resistance. Subsequent ligand-receptor analysis unveiled specific molecular dialogues, most notably TGFB1-HSPB1 and LTF-S100A14, offering insights into potential signaling pathways implicated in resistance. Our SVM model, incorporating these multi-omics and spatial data, demonstrated significant predictive power, with AUC values of 0.93 and 0.84 in the exploration and validation cohorts respectively. Hence, our results underscore the utility of multi-omics and spatial data in modeling treatment response.

CONCLUSION: Our integrative approach, amalgamating mIHC assays, feature extraction, and machine learning, successfully unraveled the complex cellular interplay underlying drug resistance. This robust predictive model may serve as a valuable tool for personalizing therapeutic strategies and enhancing treatment outcomes in gastric cancer.

PMID:38579635 | DOI:10.1016/j.drup.2024.101080

Soc Sci Med. 2024 Mar 15;348:116775. doi: 10.1016/j.socscimed.2024.116775. Online ahead of print.

ABSTRACT

The primary goal of this study is to examine the association between vaccine rhetoric on Twitter and the public's uptake rates of COVID-19 vaccines in the United States, compared to the extent of an association between self-reported vaccine acceptance and the CDC's uptake rates. We downloaded vaccine-related posts on Twitter in real-time daily for 13 months, from October 2021 to September 2022, collecting over half a billion tweets. A previously validated deep-learning algorithm was then applied to (1) filter out irrelevant tweets and (2) group the remaining relevant tweets into pro-, anti-, and neutral vaccine sentiments. Our results indicate that the tweet counts (combining all three sentiments) were significantly correlated with the uptake rates of all stages of COVID-19 shots (p < 0.01). The self-reported level of vaccine acceptance was not correlated with any of the stages of COVID-19 shots (p > 0.05) but with the daily new infection counts. These results suggest that although social media posts on vaccines may not represent the public's opinions, they are aligned with the public's behaviors of accepting vaccines, which is an essential step for developing interventions to increase the uptake rates. In contrast, self-reported vaccine acceptance represents the public's opinions, but these were not correlated with the behaviors of accepting vaccines. These outcomes provide empirical support for the validity of social media analytics for gauging the public's vaccination behaviors and understanding a nuanced perspective of the public's vaccine sentiment for health emergencies.

PMID:38579627 | DOI:10.1016/j.socscimed.2024.116775

J Neurosurg. 2024 Apr 5:1-10. doi: 10.3171/2024.1.JNS232196. Online ahead of print.

ABSTRACT

OBJECTIVE: CT and MRI are synergistic in the information provided for neurosurgical planning. While obtaining both types of images lends unique data from each, doing so adds to cost and exposes patients to additional ionizing radiation after MRI has been performed. Cross-modal synthesis of high-resolution CT images from MRI sequences offers an appealing solution. The authors therefore sought to develop a deep learning conditional generative adversarial network (cGAN) which performs this synthesis.

METHODS: Preoperative paired CT and contrast-enhanced MR images were collected for patients with meningioma, pituitary tumor, vestibular schwannoma, and cerebrovascular disease. CT and MR images were denoised, field corrected, and coregistered. MR images were fed to a cGAN that exported a "synthetic" CT scan. The accuracy of synthetic CT images was assessed objectively using the quantitative similarity metrics as well as by clinical features such as sella and internal auditory canal (IAC) dimensions and mastoid/clinoid/sphenoid aeration.

RESULTS: A total of 92,981 paired CT/MR images obtained in 80 patients were used for training/testing, and 10,068 paired images from 10 patients were used for external validation. Synthetic CT images reconstructed the bony skull base and convexity with relatively high accuracy. Measurements of the sella and IAC showed a median relative error between synthetic CT scans and ground truth images of 6%, with greater variability in IAC reconstruction compared with the sella. Aerations in the mastoid, clinoid, and sphenoid regions were generally captured, although there was heterogeneity in finer air cell septations. Performance varied based on pathology studied, with the highest limitation observed in evaluating meningiomas with intratumoral calcifications or calvarial invasion.

CONCLUSIONS: The generation of high-resolution CT scans from MR images through cGAN offers promise for a wide range of applications in cranial and spinal neurosurgery, especially as an adjunct for preoperative evaluation. Optimizing cGAN performance on specific anatomical regions may increase its clinical viability.

PMID:38579358 | DOI:10.3171/2024.1.JNS232196

IEEE Trans Ultrason Ferroelectr Freq Control. 2024 Apr 5;PP. doi: 10.1109/TUFFC.2024.3385698. Online ahead of print.

ABSTRACT

Freehand 3D ultrasound imaging is emerging as a promising modality for regular spine exams due to its non-invasiveness and affordability. The laminae landmarks play a critical role in depicting the 3D shape of the spine. However, the extraction of the 3D lamina curves from transverse ultrasound sequences presents a challenging task, primarily attributed to the presence of diverse contrast variations, imaging artifacts, the complex surface of vertebral bones, and the difficulties associated with probe manipulation. This paper proposes Sequential Localization Recurrent Convolutional Networks (SL-RCN), a novel deep learning model that takes the contextual relationships into account and embeds the transformation matrix feature as a 3D knowledge base to enhance accurate ultrasound sequence analysis. The assessment involved the analysis of 3D ultrasound sequences obtained from 10 healthy adult human participants, covering both the lumbar and thoracic regions. The performance of SL-RCN is evaluated through 7-fold cross-validation, employing the leave-one-participant-out strategy. The validity of the AI model training is assessed on test data from 3 participants. Normalized Discrete Fréchet Distance (NDFD) is employed as the main metric to evaluate the disparity of the extracted 3D lamina curves. In contrast to our previous 2D image analysis method, SL-RCN generates reduced left/right mean distance errors from 1.62/1.63mm to 1.41/1.40mm, and NDFDs from 0.5910/0.6389 to 0.4276/0.4567. The increase in the mean NDFD value from 7-fold cross-validation to the test-data experiment is less than 0.05. The experiments demonstrate the SL-RCN's capability in extracting accurate paired smooth lamina landmark curves.

PMID:38578857 | DOI:10.1109/TUFFC.2024.3385698

IEEE Trans Neural Syst Rehabil Eng. 2024 Apr 5;PP. doi: 10.1109/TNSRE.2024.3384498. Online ahead of print.

ABSTRACT

Predicting the potential for recovery of motor function in stroke patients who undergo specific rehabilitation treatments is an important and major challenge. Recently, electroencephalography (EEG) has shown potential in helping to determine the relationship between cortical neural activity and motor recovery. EEG recorded in different states could more accurately predict motor recovery than single-state recordings. Here, we design a multi-state (combining eyes closed, EC, and eyes open, EO) fusion neural network for predicting the motor recovery of patients with stroke after EEG-brain-computer-interface (BCI) rehabilitation training and use an explainable deep learning method to identify the most important features of EEG power spectral density and functional connectivity contributing to prediction. The prediction accuracy of the multi-states fusion network was 82%, significantly improved compared with a single-state model. The neural network explanation result demonstrated the important region and frequency oscillation bands. Specifically, in those two states, power spectral density and functional connectivity were shown as the regions and bands related to motor recovery in frontal, central, and occipital. Moreover, the motor recovery relation in bands, the power spectrum density shows the bands at delta and alpha bands. The functional connectivity shows the delta, theta, and alpha bands in the EC state; delta, theta, and beta mid at the EO state are related to motor recovery. Multi-state fusion neural networks, which combine multiple states of EEG signals into a single network, can increase the accuracy of predicting motor recovery after BCI training, and reveal the underlying mechanisms of motor recovery in brain activity.

PMID:38578854 | DOI:10.1109/TNSRE.2024.3384498

PLoS One. 2024 Apr 5;19(4):e0300767. doi: 10.1371/journal.pone.0300767. eCollection 2024.

ABSTRACT

Semantic segmentation of cityscapes via deep learning is an essential and game-changing research topic that offers a more nuanced comprehension of urban landscapes. Deep learning techniques tackle urban complexity and diversity, which unlocks a broad range of applications. These include urban planning, transportation management, autonomous driving, and smart city efforts. Through rich context and insights, semantic segmentation helps decision-makers and stakeholders make educated decisions for sustainable and effective urban development. This study investigates an in-depth exploration of cityscape image segmentation using the U-Net deep learning model. The proposed U-Net architecture comprises an encoder and decoder structure. The encoder uses convolutional layers and down sampling to extract hierarchical information from input images. Each down sample step reduces spatial dimensions, and increases feature depth, aiding context acquisition. Batch normalization and dropout layers stabilize models and prevent overfitting during encoding. The decoder reconstructs higher-resolution feature maps using "UpSampling2D" layers. Through extensive experimentation and evaluation of the Cityscapes dataset, this study demonstrates the effectiveness of the U-Net model in achieving state-of-the-art results in image segmentation. The results clearly shown that, the proposed model has high accuracy, mean IOU and mean DICE compared to existing models.

PMID:38578733 | DOI:10.1371/journal.pone.0300767

Brain Inform. 2024 Apr 5;11(1):10. doi: 10.1186/s40708-024-00222-1.

ABSTRACT

Explainable artificial intelligence (XAI) has gained much interest in recent years for its ability to explain the complex decision-making process of machine learning (ML) and deep learning (DL) models. The Local Interpretable Model-agnostic Explanations (LIME) and Shaply Additive exPlanation (SHAP) frameworks have grown as popular interpretive tools for ML and DL models. This article provides a systematic review of the application of LIME and SHAP in interpreting the detection of Alzheimer's disease (AD). Adhering to PRISMA and Kitchenham's guidelines, we identified 23 relevant articles and investigated these frameworks' prospective capabilities, benefits, and challenges in depth. The results emphasise XAI's crucial role in strengthening the trustworthiness of AI-based AD predictions. This review aims to provide fundamental capabilities of LIME and SHAP XAI frameworks in enhancing fidelity within clinical decision support systems for AD prognosis.

PMID:38578524 | DOI:10.1186/s40708-024-00222-1

J Med Imaging Radiat Oncol. 2024 Apr 5. doi: 10.1111/1754-9485.13649. Online ahead of print.

ABSTRACT

INTRODUCTION: Deep learning reconstruction (DLR) technologies are the latest methods attempting to solve the enduring problem of reducing MRI acquisition times without compromising image quality. The clinical utility of this reconstruction technique is yet to be fully established. This study aims to assess whether a commercially available DLR technique applied to 2D T2-weighted FLAIR brain images allows a reduction in scan time, without compromising image quality and thus diagnostic accuracy.

METHODS: 47 participants (24 male, mean age 55.9 ± 18.7 SD years, range 20-89 years) underwent routine, clinically indicated brain MRI studies in March 2022, that included a standard-of-care (SOC) T2-weighted FLAIR sequence, and an accelerated acquisition that was reconstructed using the DLR denoising product. Overall image quality, lesion conspicuity, signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and artefacts for each sequence, and preferred sequence on direct comparison, were subjectively assessed by two readers.

RESULTS: There was a strong preference for SOC FLAIR sequence for overall image quality (P = 0.01) and head-to-head comparison (P < 0.001). No difference was observed for lesion conspicuity (P = 0.49), perceived SNR (P = 1.0), and perceived CNR (P = 0.84). There was no difference in motion (P = 0.57) nor Gibbs ringing (P = 0.86) artefacts. Phase ghosting (P = 0.038) and pseudolesions were significantly more frequent (P < 0.001) on DLR images.

CONCLUSION: DLR algorithm allowed faster FLAIR acquisition times with comparable image quality and lesion conspicuity. However, an increased incidence and severity of phase ghosting artefact and presence of pseudolesions using this technique may result in a reduction in reading speed, efficiency, and diagnostic confidence.

PMID:38577926 | DOI:10.1111/1754-9485.13649

Acc Chem Res. 2024 Apr 5. doi: 10.1021/acs.accounts.4c00093. Online ahead of print.

ABSTRACT

ConspectusMolecular docking, also termed ligand docking (LD), is a pivotal element of structure-based virtual screening (SBVS) used to predict the binding conformations and affinities of protein-ligand complexes. Traditional LD methodologies rely on a search and scoring framework, utilizing heuristic algorithms to explore binding conformations and scoring functions to evaluate binding strengths. However, to meet the efficiency demands of SBVS, these algorithms and functions are often simplified, prioritizing speed over accuracy.The emergence of deep learning (DL) has exerted a profound impact on diverse fields, ranging from natural language processing to computer vision and drug discovery. DeepMind's AlphaFold2 has impressively exhibited its ability to accurately predict protein structures solely from amino acid sequences, highlighting the remarkable potential of DL in conformation prediction. This groundbreaking advancement circumvents the traditional search-scoring frameworks in LD, enhancing both accuracy and processing speed and thereby catalyzing a broader adoption of DL algorithms in binding pose prediction. Nevertheless, a consensus on certain aspects remains elusive.In this Account, we delineate the current status of employing DL to augment LD within the VS paradigm, highlighting our contributions to this domain. Furthermore, we discuss the challenges and future prospects, drawing insights from our scholarly investigations. Initially, we present an overview of VS and LD, followed by an introduction to DL paradigms, which deviate significantly from traditional search-scoring frameworks. Subsequently, we delve into the challenges associated with the development of DL-based LD (DLLD), encompassing evaluation metrics, application scenarios, and physical plausibility of the predicted conformations. In the evaluation of LD algorithms, it is essential to recognize the multifaceted nature of the metrics. While the accuracy of binding pose prediction, often measured by the success rate, is a pivotal aspect, the scoring/screening power and computational speed of these algorithms are equally important given the pivotal role of LD tools in VS. Regarding application scenarios, early methods focused on blind docking, where the binding site is unknown. However, recent studies suggest a shift toward identifying binding sites rather than solely predicting binding poses within these models. In contrast, LD with a known pocket in VS has been shown to be more practical. Physical plausibility poses another significant challenge. Although DLLD models often achieve higher success rates compared to traditional methods, they may generate poses with implausible local structures, such as incorrect bond angles or lengths, which are disadvantageous for postprocessing tasks like visualization. Finally, we discuss the future perspectives for DLLD, emphasizing the need to improve generalization ability, strike a balance between speed and accuracy, account for protein conformation flexibility, and enhance physical plausibility. Additionally, we delve into the comparison between generative and regression algorithms in this context, exploring their respective strengths and potential.

PMID:38577892 | DOI:10.1021/acs.accounts.4c00093

Front Immunol. 2024 Mar 21;15:1342285. doi: 10.3389/fimmu.2024.1342285. eCollection 2024.

ABSTRACT

B cell receptors (BCRs) denote antigen specificity, while corresponding cell subsets indicate B cell functionality. Since each B cell uniquely encodes this combination, physical isolation and subsequent processing of individual B cells become indispensable to identify both attributes. However, this approach accompanies high costs and inevitable information loss, hindering high-throughput investigation of B cell populations. Here, we present BCR-SORT, a deep learning model that predicts cell subsets from their corresponding BCR sequences by leveraging B cell activation and maturation signatures encoded within BCR sequences. Subsequently, BCR-SORT is demonstrated to improve reconstruction of BCR phylogenetic trees, and reproduce results consistent with those verified using physical isolation-based methods or prior knowledge. Notably, when applied to BCR sequences from COVID-19 vaccine recipients, it revealed inter-individual heterogeneity of evolutionary trajectories towards Omicron-binding memory B cells. Overall, BCR-SORT offers great potential to improve our understanding of B cell responses.

PMID:38576618 | PMC:PMC10991714 | DOI:10.3389/fimmu.2024.1342285

Ann Surg. 2024 Apr 5. doi: 10.1097/SLA.0000000000006294. Online ahead of print.

ABSTRACT

OBJECTIVE: The aim of this study was to develop and test a prototype of a deep learning surgical guidance system (CASL) that can intra-operative identify peritoneal surface metastases on routine laparoscopy images.

BACKGROUND: For a number of cancer patients, operative resection with curative intent can end up in early recurrence of the cancer. Surgeons misidentifying visible peritoneal surface metastases is likely a common reason.

METHODS: CASL was developed and tested using staging laparoscopy images recorded from 132 patients with histologically-confirmed adenocarcinoma involving the gastrointestinal tract. The data included images depicting 4287 visible peritoneal surface lesions and 3650 image patches of 365 biopsied peritoneal surface lesions. The prototype's diagnostic performance was compared to results from a national survey evaluating 111 oncologic surgeons in a simulated clinical environment.

RESULTS: In a simulated environment, surgeons' accuracy of correctly recommending a biopsy for metastases while omitting a biopsy for benign lesions was only 52%. In this environment, the prototype of a deep learning surgical guidance system demonstrated improved performance in identifying peritoneal surface metastases compared to oncologic surgeons with an area under the receiver operating characteristic curve of 0.69 (oncologic surgeon) versus 0.78 (CASL) versus 0.79 (human-computer combined). A proposed model would have improved the identification of metastases by 5% while reducing the number of unnecessary biopsies by 28% compared to current standard practice.

CONCLUSIONS: Our findings demonstrate a pathway for an artificial intelligence system for intra-operative identification of peritoneal surface metastases, but still requires additional development and future validation in a multi-institutional clinical setting.

PMID:38577794 | DOI:10.1097/SLA.0000000000006294

Proc SPIE Int Soc Opt Eng. 2023 Feb;12471:1247107. doi: 10.1117/12.2654851. Epub 2023 Apr 6.

ABSTRACT

Hyperspectral imaging is a label-free and non-invasive imaging modality that seeks to capture images in different wavelengths. In this study, we used a vision transformer that was pre-trained from video data to detect thyroid cancer on hyperspectral images. We built a dataset of 49 whole slide hyperspectral images (WS-HSI) of thyroid cancer. To improve training, we introduced 5 new data augmentation methods that transform spectra. We achieved an F-1 score of 88.1% and an accuracy of 89.64% on our test dataset. The transformer network and the whole slide hyperspectral imaging technique can have many applications in digital pathology.

PMID:38577581 | PMC:PMC10993530 | DOI:10.1117/12.2654851

Front Endocrinol (Lausanne). 2024 Mar 21;15:1293953. doi: 10.3389/fendo.2024.1293953. eCollection 2024.

ABSTRACT

BACKGROUND: The effect of surgery on advanced prostate cancer (PC) is unclear and predictive model for postoperative survival is lacking yet.

METHODS: We investigate the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) database, to collect clinical features of advanced PC patients. According to clinical experience, age, race, grade, pathology, T, N, M, stage, size, regional nodes positive, regional nodes examined, surgery, radiotherapy, chemotherapy, history of malignancy, clinical Gleason score (composed of needle core biopsy or transurethral resection of the prostate specimens), pathological Gleason score (composed of prostatectomy specimens) and prostate-specific antigen (PSA) are the potential predictive variables. All samples are divided into train cohort (70% of total, for model training) and test cohort (30% of total, for model validation) by random sampling. We then develop neural network to predict advanced PC patients' overall. Area under receiver operating characteristic curve (AUC) is used to evaluate model's performance.

RESULTS: 6380 patients, diagnosed with advanced (stage III-IV) prostate cancer and receiving surgery, have been included. The model using all collected clinical features as predictors and based on neural network algorithm performs best, which scores 0.7058 AUC (95% CIs, 0.7021-0.7068) in train cohort and 0.6925 AUC (95% CIs, 0.6906-0.6956) in test cohort. We then package it into a Windows 64-bit software.

CONCLUSION: Patients with advanced prostate cancer may benefit from surgery. In order to forecast their overall survival, we first build a clinical features-based prognostic model. This model is accuracy and may offer some reference on clinical decision making.

PMID:38577575 | PMC:PMC10991752 | DOI:10.3389/fendo.2024.1293953

Health Inf Sci Syst. 2024 Apr 2;12(1):28. doi: 10.1007/s13755-024-00289-4. eCollection 2024 Dec.

ABSTRACT

This paper considers a new method for providing a recommendation (second opinion) for a laboratory assistant in manual blood typing based on serological plates. The manual method consists of two steps: preparation and analysis. During preparation step the laboratory assistant needs to fill each well of a plate with a blood sample and a reagent mixture according to methodological guidelines. In the second step it is necessary to visually determine the result of the reactions, named agglutination. Despite the popularity of this method, it is slow and highly influenced by human factor, which cause blood typing errors. To increase the quality and performance of the analysis step, we propose a novel neural-based classification method. Our solution provides a fast way to fill the results into a laboratory system. We collected a new large dataset consisting of 3139 well images with GTs from donors' medical history and six experts' assessment for each. We showed that the proposed solution based on state-of-the-art architectures is comparable with the best expert and has 2.75 times fewer errors than the average one, with an overall accuracy equal to 98.4%. Taking into account the low-semantic nature of the task, we also considered shallow neural networks, which showed accuracy comparable with state-of-the-art models.

PMID:38577517 | PMC:PMC10987457 | DOI:10.1007/s13755-024-00289-4