Search

Deep learning applied to whole-slide histopathology images (WSIs) has the potential to enhance precision oncology and alleviate the workload of experts. However, developing these models necessitates large amounts of data with ground truth labels, which can be both time-consuming and expensive to obtain. Pathology reports are typically unstructured or poorly structured texts, and efforts to implement structured reporting templates have been unsuccessful, as these efforts lead to perceived extra workload. In this study, we hypothesised that large language models (LLMs), such as the generative pre-trained transformer 4 (GPT-4), can extract structured data from unstructured plain language reports using a zero-shot approach without requiring any re-training. We tested this hypothesis by utilising GPT-4 to extract information from histopathological reports, focusing on two extensive sets of pathology reports for colorectal cancer and glioblastoma. We found a high concordance between LLM-generated structured data and human-generated structured data. Consequently, LLMs could potentially be employed routinely to extract ground truth data for machine learning from unstructured pathology reports in the future. © 2023 The Authors. The Journal of Pathology published by John Wiley & Sons Ltd on behalf of The Pathological Society of Great Britain and Ireland., (© 2023 The Authors. The Journal of Pathology published by John Wiley & Sons Ltd on behalf of The Pathological Society of Great Britain and Ireland.)

Artificial intelligence (AI) can predict the presence of molecular alterations directly from routine histopathology slides. However, training robust AI systems requires large datasets for which data collection faces practical, ethical and legal obstacles. These obstacles could be overcome with swarm learning (SL), in which partners jointly train AI models while avoiding data transfer and monopolistic data governance. Here, we demonstrate the successful use of SL in large, multicentric datasets of gigapixel histopathology images from over 5,000 patients. We show that AI models trained using SL can predict BRAF mutational status and microsatellite instability directly from hematoxylin and eosin (H&E)-stained pathology slides of colorectal cancer. We trained AI models on three patient cohorts from Northern Ireland, Germany and the United States, and validated the prediction performance in two independent datasets from the United Kingdom. Our data show that SL-trained AI models outperform most locally trained models, and perform on par with models that are trained on the merged datasets. In addition, we show that SL-based AI models are data efficient. In the future, SL can be used to train distributed AI models for any histopathology image analysis task, eliminating the need for data transfer., (© 2022. The Author(s).)

Purpose: To evaluate in glaucoma patients the feasibility and use of remote monitoring of intraocular pressure (IOP) with an implanted telemetry sensor during the coronavirus disease 2019 (COVID-19) lockdown., Design: Cross-sectional study., Participants: Patients previously implanted with a telemetric IOP sensor (Eyemate; Implandata GmbH) were included., Methods: Intraocular pressure measurements acquired by the patients during the lockdown were collected by physicians who were located remotely. A questionnaire was sent to 10 participating study centers to evaluate the clinical impact of remote monitoring of IOP via the IOP sensor system., Main Outcome Measures: Number of patients who obtained home IOP measurements., Results: Data were available from all centers and from 37 eyes of 37 patients (16 patients with a sulcus-based sensor and 21 patients with a suprachoroidal sensor). Thirty-four patients obtained IOP measurements during the lockdown. Mean age of the patients was 69.3 ± 9.6 years, and 48.6% were women. A total of 8415 IOP measurements from 370 measurement days were obtained. Based on remote IOP measurements, treatment was changed in 5 patients. In another 5 patients, treatment change was considered when physicians received the IOP measurements after the lockdown. Nine of the 10 study centers judged remote IOP measurements to have a clinical impact., Conclusions: These results show the feasibility of patient-acquired measurement of IOP in conjunction with remote IOP monitoring by physicians with an implantable sensor. The data obtained impacted clinical decision making, including adjustment of ocular hypotensive therapy and avoiding unnecessary office visits during the COVID-19 pandemic., (Copyright © 2020 American Academy of Ophthalmology. Published by Elsevier Inc. All rights reserved.)

Background and Aims: Non-alcoholic steatohepatitis (NASH) is a chronic disease that can progress to end-stage liver disease (ESLD). A large proportion of early-stage NASH patients remain undiagnosed compared to those with advanced fibrosis, who are more likely to receive disease management interventions. This study estimated the disease burden and economic impact of diagnosed NASH in the adult population of France, Germany, Italy, Spain and the United Kingdom in 2018., Methods: The socioeconomic burden of diagnosed NASH was estimated using cost-of-illness methodology applying a prevalence approach to estimate the number of adults with NASH and the attributable economic and wellbeing costs. Given undiagnosed patients do not incur costs in the study, the probability of diagnosis is central to cost estimation. The analysis was based on a literature review, databases and consultation with clinical experts, economists and patient groups., Results: The proportion of adult NASH patients with a diagnosis ranged from 11.9% to 12.7% across countries, which increased to 38.8%-39.1% for advanced fibrosis (F3-F4 compensated cirrhosis). Total economic costs were €8548-19 546M. Of these, health system costs were €619-1292M. Total wellbeing costs were €41 536-90 379M. The majority of the undiagnosed population (87.3%-88.2% of total prevalence) was found to have early-stage NASH, which, left untreated, may progress to more resource consuming ESLD over time., Conclusions: This study found that the majority of economic and wellbeing costs of NASH are experienced in late disease stages. Earlier diagnosis and care of NASH patients could reduce future healthcare costs., (© 2021 The Authors. Liver International published by John Wiley & Sons Ltd.)

Glaucoma, a disease characterized by progressive optic nerve degeneration, can be prevented through timely diagnosis and treatment. We characterize optic nerve photographs of 67,040 UK Biobank participants and use a multitrait genetic model to identify risk loci for glaucoma. A glaucoma polygenic risk score (PRS) enables effective risk stratification in unselected glaucoma cases and modifies penetrance of the MYOC variant encoding p.Gln368Ter, the most common glaucoma-associated myocilin variant. In the unselected glaucoma population, individuals in the top PRS decile reach an absolute risk for glaucoma 10 years earlier than the bottom decile and are at 15-fold increased risk of developing advanced glaucoma (top 10% versus remaining 90%, odds ratio = 4.20). The PRS predicts glaucoma progression in prospectively monitored, early manifest glaucoma cases (P = 0.004) and surgical intervention in advanced disease (P = 3.6 × 10 - 6 ). This glaucoma PRS will facilitate the development of a personalized approach for earlier treatment of high-risk individuals, with less intensive monitoring and treatment being possible for lower-risk groups.