Search

Lysosomes are degradation and signalling centres crucial for homeostasis, development and ageing1. To meet diverse cellular demands, lysosomes remodel their morphology and function through constant fusion and fission2,3. Little is known about the molecular basis of fission. Here we identify HPO-27, a conserved HEAT repeat protein, as a lysosome scission factor in Caenorhabditis elegans. Loss of HPO-27 impairs lysosome fission and leads to an excessive tubular network that ultimately collapses. HPO-27 and its human homologue MROH1 are recruited to lysosomes by RAB-7 and enriched at scission sites. Super-resolution imaging, negative-staining electron microscopy and in vitro reconstitution assays reveal that HPO-27 and MROH1 self-assemble to mediate the constriction and scission of lysosomal tubules in worms and mammalian cells, respectively, and assemble to sever supported membrane tubes in vitro. Loss of HPO-27 affects lysosomal morphology, integrity and degradation activity, which impairs animal development and longevity. Thus, HPO-27 and MROH1 act as self-assembling scission factors to maintain lysosomal homeostasis and function.The conserved HEAT repeat protein HPO-27 is identified as a lysosome scission factor in Caenorhabditis elegans, and the human homologue MROH1 also serves the same function to maintain lysosomal homeostasis. [ABSTRACT FROM AUTHOR]

Large language models (LLMs) have shown promise in accelerating radiology reporting by summarizing clinical findings into impressions. However, automatic impression generation for whole-body PET reports presents unique challenges and has received little attention. Our study aimed to evaluate whether LLMs can create clinically useful impressions for PET reporting. To this end, we fine-tuned twelve open-source language models on a corpus of 37,370 retrospective PET reports collected from our institution. All models were trained using the teacher-forcing algorithm, with the report findings and patient information as input and the original clinical impressions as reference. An extra input token encoded the reading physician's identity, allowing models to learn physician-specific reporting styles. To compare the performances of different models, we computed various automatic evaluation metrics and benchmarked them against physician preferences, ultimately selecting PEGASUS as the top LLM. To evaluate its clinical utility, three nuclear medicine physicians assessed the PEGASUS-generated impressions and original clinical impressions across 6 quality dimensions (3-point scales) and an overall utility score (5-point scale). Each physician reviewed 12 of their own reports and 12 reports from other physicians. When physicians assessed LLM impressions generated in their own style, 89% were considered clinically acceptable, with a mean utility score of 4.08/5. On average, physicians rated these personalized impressions as comparable in overall utility to the impressions dictated by other physicians (4.03, P = 0.41). In summary, our study demonstrated that personalized impressions generated by PEGASUS were clinically useful in most cases, highlighting its potential to expedite PET reporting by automatically drafting impressions. [ABSTRACT FROM AUTHOR]

Background: Small round cell tumor (SRCT) is a group of malignancy with similar optical microscopic morphology. Despite its low incidence, SRCT has a high malignant degree and poor prognosis. Besides, atypical clinical symptoms make it difficult in preoperative diagnosis. Case report: A 67-year-old man was presented to the outpatient service with dysuria and weak urine stream lasting for 3 months. After oral treatment with tamsulosin and finasteride for 2 months, the symptoms worsen. Transurethral prostate holmium laser enucleation was operated and postoperative pathology result revealed small blue round cell malignant tumor. Further immunohistochemistry and fluorescence in situ hybridization examination indicated Ewing-like SRCT. So a Da Vinci Robotic prostatectomy was performed further and whole-genome sequencing was conducted. Several gene mutations including RAF1, ARID1A, SMARCA4, and BCL2L11 were found but no FDA-approved drug could treat specifically. Then the patient received Ewing-type therapeutic regimens treatment and has been followed up to date (over 24 months). Conclusion: Because of its non-elevated serum PSA level, prostate SRCT is often ignored as a possibility of malignant tumor and regarded as benign prostatic hyperplasia (BPH). The possibility of prostate SRCT need to be considered if dysuria symptoms could not alleviate significantly after a period of oral treatment. [ABSTRACT FROM AUTHOR]

The endoplasmic reticulum (ER), which is composed of a continuous network of tubules and sheets, forms the most widely distributed membrane system in eukaryotic cells. As a result, it engages a variety of organelles by establishing membrane contact sites (MCSs). These contacts regulate organelle positioning and remodeling, including fusion and fission, facilitate precise lipid exchange, and couple vital signaling events. Here, we systematically review recent advances and converging themes on ER-involved organellar contact. The molecular basis, cellular influence, and potential physiological functions for ER/nuclear envelope contacts with mitochondria, Golgi, endosomes, lysosomes, lipid droplets, autophagosomes, and plasma membrane are summarized. [ABSTRACT FROM AUTHOR]

The growing application of flotation reagents (e.g., xanthate collectors) exacerbates water pollution in mined sites. Semiconductor-based photocatalysis is widely recognized as an efficient and green technology for treating mineral processing wastewater. However, the on-site practical application of powdery photocatalysts is still unsatisfactory because of their poor photochemical stability and limited recyclability. In this work, a flexible CdS/mixed cellulose composite membrane (CdS/MCM) was synthesized via a simple impregnation–precipitation process, and such a material has not been reported so far. CdS nanoparticles were embedded into the holes of MCM but did not block the membrane channels, avoiding the over-agglomeration of CdS nanoparticles. Compared with CdS powder, CdS/MCM had higher photocatalytic property. The degradation efficiency for sodium ethyl xanthate, potassium butyl xanthate, and potassium amyl xanthate over the optimized 0.175CdS/MCM exceeded 96%. Moreover, the separation of CdS/MCM from the pollution solution was easy, and CdS/MCM could be reused more than five times without any deactivation. This work opens new avenues in beneficiation wastewater treatment by rationally designing high-performance, low-cost CdS/MCM materials. [ABSTRACT FROM AUTHOR]

Large language models (LLMs) have shown promise in accelerating radiology reporting by summarizing clinical findings into impressions. However, automatic impression generation for whole-body PET reports presents unique challenges and has received little attention. Our study aimed to evaluate whether LLMs can create clinically useful impressions for PET reporting. To this end, we fine-tuned twelve open-source language models on a corpus of 37,370 retrospective PET reports collected from our institution. All models were trained using the teacher-forcing algorithm, with the report findings and patient information as input and the original clinical impressions as reference. An extra input token encoded the reading physician's identity, allowing models to learn physician-specific reporting styles. To compare the performances of different models, we computed various automatic evaluation metrics and benchmarked them against physician preferences, ultimately selecting PEGASUS as the top LLM. To evaluate its clinical utility, three nuclear medicine physicians assessed the PEGASUS-generated impressions and original clinical impressions across 6 quality dimensions (3-point scales) and an overall utility score (5-point scale). Each physician reviewed 12 of their own reports and 12 reports from other physicians. When physicians assessed LLM impressions generated in their own style, 89% were considered clinically acceptable, with a mean utility score of 4.08/5. On average, physicians rated these personalized impressions as comparable in overall utility to the impressions dictated by other physicians (4.03, P = 0.41). In summary, our study demonstrated that personalized impressions generated by PEGASUS were clinically useful in most cases, highlighting its potential to expedite PET reporting by automatically drafting impressions., (© 2024. The Author(s) under exclusive licence to Society for Imaging Informatics in Medicine.)

Radiology narrative reports often describe characteristics of a patient's disease, including its location, size, and shape. Motivated by the recent success of multimodal learning, we hypothesized that this descriptive text could guide medical image analysis algorithms. We proposed a novel vision-language model, ConTEXTual Net, for the task of pneumothorax segmentation on chest radiographs. ConTEXTual Net extracts language features from physician-generated free-form radiology reports using a pre-trained language model. We then introduced cross-attention between the language features and the intermediate embeddings of an encoder-decoder convolutional neural network to enable language guidance for image analysis. ConTEXTual Net was trained on the CANDID-PTX dataset consisting of 3196 positive cases of pneumothorax with segmentation annotations from 6 different physicians as well as clinical radiology reports. Using cross-validation, ConTEXTual Net achieved a Dice score of 0.716±0.016, which was similar to the degree of inter-reader variability (0.712±0.044) computed on a subset of the data. It outperformed vision-only models (Swin UNETR: 0.670±0.015, ResNet50 U-Net: 0.677±0.015, GLoRIA: 0.686±0.014, and nnUNet 0.694±0.016) and a competing vision-language model (LAVT: 0.706±0.009). Ablation studies confirmed that it was the text information that led to the performance gains. Additionally, we show that certain augmentation methods degraded ConTEXTual Net's segmentation performance by breaking the image-text concordance. We also evaluated the effects of using different language models and activation functions in the cross-attention module, highlighting the efficacy of our chosen architectural design., (© 2024. The Author(s) under exclusive licence to Society for Imaging Informatics in Medicine.)