Deep learning model integrating cfDNA methylation and fragment size profiles for lung cancer diagnosis.

Detecting aberrant cell-free DNA (cfDNA) methylation is a promising strategy for lung cancer diagnosis. In this study, our aim is to identify methylation markers to distinguish patients with lung cancer from healthy individuals. Additionally, we sought to develop a deep learning model incorporating cfDNA methylation and fragment size profiles. To achieve this, we utilized methylation data collected from The Cancer Genome Atlas and Gene Expression Omnibus databases. Then we generated methylated DNA immunoprecipitation sequencing and genome-wide Enzymatic Methyl-seq (EM-seq) form lung cancer tissue and plasma. Using these data, we selected 366 methylation markers. A targeted EM-seq panel was designed using the selected markers, and 142 lung cancer and 56 healthy samples were produced with the panel. Additionally, cfDNA samples from healthy individuals and lung cancer patients were diluted to evaluate sensitivity. Its lung cancer detection performance reached an accuracy of 81.5% and an area under the receiver operating characteristic curve of 0.87. In the serial dilution experiment, we achieved tumor fraction detection of 1% at 98% specificity and 0.1% at 80% specificity. In conclusion, we successfully developed and validated a combination of methylation panel and a deep learning model that can distinguish between patients with lung cancer and healthy individuals.

Deep learning-based natural language processing for detecting medical symptoms and histories in emergency patient triage.

OBJECTIVE: The manual recording of electronic health records (EHRs) by clinicians in the emergency department (ED) is time-consuming and challenging. In light of recent advancements in large language models (LLMs) such as GPT and BERT, this study aimed to design and validate LLMs for automatic clinical diagnoses. The models were designed to identify 12 medical symptoms and 2 patient histories from simulated clinician-patient conversations within 6 primary symptom scenarios in emergency triage rooms. MATERIALS AND METHOD: We developed classification models by fine-tuning BERT, a transformer-based pre-trained model. We subsequently analyzed these models using eXplainable artificial intelligence (XAI) and the Shapley additive explanation (SHAP) method. A Turing test was conducted to ascertain the reliability of the XAI results by comparing them to the outcomes of tasks performed and explained by medical workers. An emergency medicine specialist assessed the results of both XAI and the medical workers. RESULTS: We fine-tuned four pre-trained LLMs and compared their classification performance. The KLUE-RoBERTa-based model demonstrated the highest performance (F1-score: 0.965, AUROC: 0.893) on human-transcribed script data. The XAI results using SHAP showed an average Jaccard similarity of 0.722 when compared with explanations of medical workers for 15 samples. The Turing test results revealed a small 6% gap, with XAI and medical workers receiving the mean scores of 3.327 and 3.52, respectively. CONCLUSION: This paper highlights the potential of LLMs for automatic EHR recording in Korean EDs. The KLUE-RoBERTa-based model demonstrated superior classification performance. Furthermore, XAI using SHAP provided reliable explanations for model outputs. The reliability of these explanations was confirmed by a Turing test.

Fully Automated Segmentation Models of Supratentorial Meningiomas Assisted by Inclusion of Normal Brain Images.

To train an automatic brain tumor segmentation model, a large amount of data is required. In this paper, we proposed a strategy to overcome the limited amount of clinically collected magnetic resonance image (MRI) data regarding meningiomas by pre-training a model using a larger public dataset of MRIs of gliomas and augmenting our meningioma training set with normal brain MRIs. Pre-operative MRIs of 91 meningioma patients (171 MRIs) and 10 non-meningioma patients (normal brains) were collected between 2016 and 2019. Three-dimensional (3D) U-Net was used as the base architecture. The model was pre-trained with BraTS 2019 data, then fine-tuned with our datasets consisting of 154 meningioma MRIs and 10 normal brain MRIs. To increase the utility of the normal brain MRIs, a novel balanced Dice loss (BDL) function was used instead of the conventional soft Dice loss function. The model performance was evaluated using the Dice scores across the remaining 17 meningioma MRIs. The segmentation performance of the model was sequentially improved via the pre-training and inclusion of normal brain images. The Dice scores improved from 0.72 to 0.76 when the model was pre-trained. The inclusion of normal brain MRIs to fine-tune the model improved the Dice score; it increased to 0.79. When employing BDL as the loss function, the Dice score reached 0.84. The proposed learning strategy for U-net showed potential for use in segmenting meningioma lesions.