Deepm5C: A deep-learning-based hybrid framework for identifying human RNA N5-methylcytosine sites using a stacking strategy

1. Download : Download high-res image (173KB)
2. Download : Download full-size image

     

Bench to bedside: recent advances in lung cancer pathological research with implications for diagnostic clinical practice

After decades of relative stagnation lung cancer is emerging as a disease type where rapid progress is being made in diagnosis and therapy, as well as in our understanding of disease biology. Much of this progress is of immediate impact to diagnosticians, and more is likely to affect diagnostic practice in the near future. In this review we seek to briefly summarize several key areas of active research of immediate or probable imminent value to trainee and consultant pulmonary pathologists alike. We cover some major changes in tumour classification, grading, and patient stratification, as well as considering the state of the art in machine-assisted interpretation of lung cancer histology, and the use of genetically modified lung cancer models.     

Neural selectivity,efficiency, and dose equivalence in deep brain stimulation through pulse width tuning and segmented electrodes

BackgroundAchieving deep brain stimulation (DBS) dose equivalence is challenging, especially with pulse width tuning and directional contacts. Further, the precise effects of pulse width tuning are unknown, and recent reports of the effects of pulse width tuning on neural selectivity are at odds with classic biophysical studies.MethodsWe created multicompartment neuron models for two axon diameters and used finite element modeling to determine extracellular influence from standard and segmented electrodes. We analyzed axon activation profiles and calculated volumes of tissue activated.ResultsWe find that long pulse widths focus the stimulation effect on small, nearby fibers, suppressing distant white matter tract activation (responsible for some DBS side effects) and improving battery utilization when equivalent activation is maintained for small axons. Directional leads enable similar benefits to a greater degree. Reexamining previous reports of short pulse stimulation reducing side effects, we explore a possible alternate explanation: non-dose equivalent stimulation may have resulted in reduced spread of neural activation. Finally, using internal capsule avoidance as an example in the context of subthalamic stimulation, we present a patient-specific model to show how long pulse widths could help increase the biophysical therapeutic window.DiscussionWe find agreement with classic studies and predict that long pulse widths may focus the stimulation effect on small, nearby fibers and improve power consumption. While future pre-clinical and clinical work is necessary regarding pulse width tuning, it is clear that future studies must ensure dose equivalence, noting that energy- and charge-equivalent amplitudes do not result in equivalent spread of neural activation when changing pulse width.     

人工智能在耳鼻咽喉头颈外科的运用及展望

通过对大数据深度学习,近年人工智能技术已逐渐渗透到医学各个领域,实现一定程度应用。虽然耳鼻咽喉头颈外科专业近几年发表相关文献数量急剧增长,但大部分临床医生对于人工智能的研究还比较陌生。介绍人工智能的基本原理,列举、分析其在耳鼻喉科的主要研究情况,探讨目前人工智能技术实际应用的局限,展望未来人工智能技术在耳鼻喉科可能的应用。     

Localization of motor and verbal fluency effects in subthalamic DBS for Parkinson's disease

IntroductionSubthalamic nucleus deep brain stimulation (STN DBS) improves cardinal motor symptoms of Parkinson's disease (PD) but can worsen verbal fluency (VF). An optimal site of stimulation for overall motor improvement has been previously identified using an atlas-independent, fully individualized, field-modeling approach. This study examines if cardinal motor components (bradykinesia, tremor, and rigidity) share this identified optimal improvement site and if there is co-localization with a site that worsens VF.MethodsAn atlas-independent, field-modeling approach was used to identify sites of maximal STN DBS effect on overall and cardinal motor symptoms and VF in 60 patients. Anatomic coordinates were referenced to the STN midpoint. Symptom severity was assessed with the MDS-UPDRS part III and established VF scales.ResultsSites for improved bradykinesia and rigidity co-localized with each other and the overall part III site (0.09 mm lateral, 0.93 mm posterior, 1.75 mm dorsal). The optimal site for tremor was posterior to this site (0.10 mm lateral, 1.40 mm posterior, 1.93 mm dorsal). Semantic and phonemic VF sites were indistinguishable and co-localized medial to the motor sites (0.32 mm medial, 1.18 mm posterior, 1.74 mm dorsal).ConclusionThis study identifies statistically distinct, maximally effective stimulation sites for tremor improvement, VF worsening, and overall and other cardinal motor improvements in STN DBS. Current electrode sizes and voltage settings stimulate all of these sites simultaneously. However, future targeted lead placement and focused directional stimulation may avoid VF worsening while maintaining motor improvements in STN DBS.     

Deep learning with biopsy whole slide images for pretreatment prediction of pathological complete response to neoadjuvant chemotherapy in breast cancer：A multicenter study

IntroductionPredicting pathological complete response (pCR) for patients receiving neoadjuvant chemotherapy (NAC) is crucial in establishing individualized treatment. Whole-slide images (WSIs) of tumor tissues reflect the histopathologic information of the tumor, which is important for therapeutic response effectiveness. In this study, we aimed to investigate whether predictive information for pCR could be detected from WSIs.Materials and methodsWe retrospectively collected data from four cohorts of 874 patients diagnosed with biopsy-proven breast cancer. A deep learning pathological model (DLPM) was constructed to predict pCR using biopsy WSIs in the primary cohort, and it was then validated in three external cohorts. The DLPM could generate a deep learning pathological score (DLPs) for each patient; stromal tumor-infiltrating lymphocytes (TILs) were selected for comparison with DLPs.ResultsThe WSI feature-based DLPM showed good predictive performance with the highest area under the curve (AUC) of 0.72 among the cohorts. Alternatively, the combination of the DLPM and clinical characteristics offered a better prediction performance (AUC >0.70) in all cohorts. We also evaluated the performance of DLPM in three different breast subtypes with the best prediction for the triple-negative breast cancer (TNBC) subtype (AUC: 0.73). Moreover, DLPM combined with clinical characteristics and stromal TILs achieved the highest AUC in the primary cohort (AUC: 0.82) and validation cohort 1 (AUC: 0.80).ConclusionOur study suggested that WSIs integrated with deep learning could potentially predict pCR to NAC in breast cancer. The predictive performance will be improved by combining clinical characteristics. DLPs from DLPM can provide more information compared to stromal TILs for pCR prediction.